The efficacy of salivary testosterone and cortisol as a biomarker of training and competition in elite athletes

Christopher Mark Gaviglio

Bachelor of Science

A thesis submitted for the degree of Doctor of Philosophy at

The University of Queensland in 2017

School of Human Movement and Nutrition Sciences Abstract

In populations of elite athletes, the short-term and long-term effects of endogenous testosterone and cortisol play an important role in both training and competition. In a season-long competition format, understanding the effects and demands of competition and training plays a pivotal role in player management. This research sought to improve the understanding of salivary testosterone and cortisol responses of elite rugby union players in response to competitive match play and training. The first study examined the association of pregame testosterone and cortisol concentrations with competitive game play and outcome. Grouped pregame testosterone concentrations were significantly higher before games won than before games lost (p <.001) and were also positively correlated with a game ranked performance score (r = .81, p <.05). Analysis by playing position further demonstrated that, for the players in the ‘backs’ position, both pregame testosterone concentrations (p <.001) and the testosterone to cortisol ratio (T/C) (p <.05) were significantly higher before a win than before a loss and were positively correlated to game-ranked performance (r = .81–.91, p <.05). Key performance indicators are routinely collected throughout training and gameplay to assess the performance of individuals and their relative capacity or ability. An overall score to summarise a player’s performance during games has merit in terms of an objective measure of performance throughout a season-long competition format. The results of the second study demonstrated that an aggressive-based performance indicator (AggPI) score could significantly discriminate between game outcomes over a season-long competition format. For a subset of six games, AggPI was significantly correlated with pregame testosterone concentrations (r = .86, p <.05), pregame T/C ratio (r = .86, p <.05) and the change from pre to post (peri) testosterone concentrations (r = −.87, p <.05). Whilst the first two studies established the relationship between acute salivary hormones concentration and performance outcomes, the purpose of the third study was to investigate the association of mid-week hormone concentrations with subsequent match performance and outcomes. The mid-week pre-training T/C ratio was significantly lower (p <.01) before a win than a loss and the increase in pre to post concentrations for both testosterone (p <.01) and T/C ratio (p <.001) was also significant. Significant relationships were also observed between pregame testosterone concentration and the relative change in pre to post mid-week concentrations for both cortisol (r = –.90, p = .01) and the T/C ratio (r = .90, p = .01). The fourth study evaluated the acute salivary testosterone and cortisol concentrations in response to four exercise protocols. The first two protocols were resistance training based. These consisted of 5 sets of 15 repetitions at 55% of 1 repetition maximum (1RM), (5 × 15–55%), and 3 sets of 5 repetitions at 85% 1RM, (3 ×

ii 5–85%) respectively. The third protocol was a strongman (STRNG) session consisting of three stations of exercises within a circuit, which was completed four times. The fourth protocol was combative (COMB) and consisted of exercises inspired by and wrestling. In groups of two, the athletes worked through six different stations and were completed twice. There was no difference in the average testosterone concentration in response to each exercise stimulus between any intervention. However, when pooled according to the protocol that demonstrated the greatest absolute increase in testosterone concentration, significant (p <.01) increases in testosterone concentrations were observed for three (5 × 15–55%, STRNG and COMB) of the four intervention protocols. This study highlighted the importance of a protocol-dependent approach based on each individual athletes response to the training stimulus also the potential usefulness of employing strongman and combative training protocols as an alternative stimulus to resistance training. The use of salivary hormones appears to be a suitable biomarker for evaluating training response and assessing competition readiness in rugby union players. The routine use of salivary hormones could be utilised to better understand not only the acute effects of a training stimulus but also the interactive factors associated between training and competition.

iii Declaration by author

This thesis is composed of my original work and contains no material previously published or written by another person except where due reference has been made in the text. I have clearly stated the contribution by others to jointly-authored works that I have included in my thesis.

I have clearly stated the contribution of others to my thesis as a whole, including statistical assistance, survey design, data analysis, significant technical procedures, professional editorial advice, financial support and any other original research work used or reported in my thesis. The content of my thesis is the result of work I have carried out since the commencement of my higher degree by research candidature and does not include a substantial part of work that has been submitted to qualify for the award of any other degree or diploma in any university or other tertiary institution. I have clearly stated which parts of my thesis, if any, have been submitted to qualify for another award.

I acknowledge that an electronic copy of my thesis must be lodged with the University Library and subject to the policy and procedures of The University of Queensland, the thesis be made available for research and study in accordance with the Copyright Act 1968 unless a period of embargo has been approved by the Dean of the Graduate School.

I acknowledge that copyright of all material contained in my thesis resides with the copyright holder(s) of that material. Where appropriate I have obtained copyright permission from the copyright holder to reproduce material in this thesis and have sought permission from co-authors for any jointly authored works included in the thesis.

iv Publications during candidature

Peer-reviewed papers Gaviglio, C. M., Crewther, B. T., Kilduff, L. P., Stokes, K. A., & Cook, C. J. (2014). Relationship between pregame concentrations of free testosterone and outcome in rugby union. International Journal of Sports Physiology and Performance, 9(2), 324−331.

Gaviglio, C. M., James, N., Crewther, B. T., Kilduff, L. P., & Cook, C. J. (2013). Relationship between match statistics, game outcome and pre-match hormonal state in professional rugby union. International Journal of Performance Analysis in Sport, 13(2), 522−534.

Gaviglio, C. M., & Cook, C. J. (2014). Relationship between mid-week training measures of testosterone and cortisol concentrations and game outcome in professional rugby union matches. Journal of Strength and Conditioning Research, 28(12), 3447−3452.

Gaviglio, C. M., Osborne, M., Kelly, V. K., Kilduff, L. P., & Cook, C. J. (2015). Salivary testosterone and cortisol responses to four different rugby training exercise protocols. European Journal of Sport Science, 15(6), 497−504.

v Publications included in this thesis Gaviglio, C. M., Crewther, B. T., Kilduff, L. P., Stokes, K. A., & Cook, C. J. (2014). Relationship between pregame concentrations of free testosterone and outcome in rugby union. International Journal of Sports Physiology and Performance, 9(2), 324−331. Incorporated as Chapter 3.

Contributor Statement of contribution Chris Gaviglio (Candidate) Conception and design (65%) Analysis and interpretation (60%) Drafting paper and editing (65%) Critical review (20%) Christian Cook Conception and design (25%) Analysis and interpretation (25%) Critical review (50%) Drafting paper and editing (20%) Blair Crewther Analysis and interpretation (5%) Critical review (10%) Drafting paper and editing (5%) Liam Kilduff Analysis and interpretation (5%) Conception and design (10%) Critical review (10%) Drafting paper and editing (5%) Keith Stokes Analysis and interpretation (5%) Critical review (10%) Drafting paper and editing (5%)

vi Gaviglio, C. M., James, N., Crewther, B. T., Kilduff, L. P., & Cook, C. J. (2013). Relationship between match statistics, game outcome and pre-match hormonal state in professional rugby union. International Journal of Performance Analysis in Sport, 13(2), 522−534. Incorporated as Chapter 4.

Contributor Statement of contribution Chris Gaviglio (Candidate) Conception and design (65%) Analysis and interpretation (60%) Drafting paper and editing (60%) Critical review (20%) Christian Cook Conception and design (25%) Analysis and interpretation (25%) Drafting paper and editing (25%) Critical review (50%) Nicholas James Analysis and interpretation (5%) Drafting paper and editing (5%) Critical review (10%) Blair Crewther Analysis and interpretation (5%) Drafting paper and editing (5%) Critical review (10%) Liam Kilduff Analysis and interpretation (5%) Drafting paper and editing (5%) Critical review (10%)

vii Gaviglio, C. M., & Cook, C. J. (2014). Relationship between mid-week training measures of testosterone and cortisol concentrations and game outcome in professional rugby union matches. Journal of Strength and Conditioning Research, 28(12), 3447−3452. Incorporated as Chapter 5.

Contributor Statement of contribution Chris Gaviglio (Candidate) Conception and design (65%) Analysis and interpretation (70%) Drafting paper and editing (70%) Critical review (70%) Christian Cook Conception and design (35%) Analysis and interpretation (30%) Drafting paper and editing (30%) Critical review (30%)

viii Gaviglio, C. M., Osborne, M., Kelly, V. K., Kilduff, L. P., & Cook, C. J. (2015). Salivary testosterone and cortisol responses to four different rugby training exercise protocols. European Journal of Sport Science, 15(6), 497−504. Incorporated as Chapter 6.

Contributor Statement of contribution Chris Gaviglio (Candidate) Conception and design (75%) Analysis and interpretation (60%) Drafting paper and editing (70%) Critical review (50%) Christian Cook Conception and design (25%) Analysis and interpretation (20%) Critical review (20%) Drafting paper and editing (20%) Mark Osborne Analysis and interpretation (10%) Critical review (20%) Drafting paper and editing (10%) Liam Kilduff Analysis and interpretation (5%) Critical review (10%) Drafting paper and editing (5%) Vince Kelly Analysis and interpretation (5%) Critical review (10%) Drafting paper and editing (5%)

ix Contributions by others to the thesis

No contribution by others.

Statement of parts of the thesis submitted to qualify for the award of another degree

None.

Research Involving Human or Animal Subjects

Ethics approval number: 09/H0101/66

Name of approving committee: University of Bath – School for Health

Cop of ethics approval letter: Appendix 1

x Acknowledgements

There have been many individuals who have contributed to this thesis. Firstly, thank you to my supervisors, Dr Mark Osborne and Dr Vince Kelly for your advice and guidance.

The involvement of the players at Bath Rugby made this research possible. To get these elite athletes to do what I asked of them is a testament to their willingness to explore avenues for continual athletic excellence. Thank you to Dr Keith Stokes who assisted with ethics approval and support during the initial phases of the project. Special mention must also go to Prof Christian Cook who was a guiding resource during this time whilst at Bath Rugby.

Thank you to my wife, Angela. This journey has taken us to Europe and back and has also seen the birth of our two beautiful children, Zara and Tristan. I thank you also for your patience, counsel and support during this time.

Professional editor Dr Leigh Findlay (TrueNature Writing & Editing) provided copyediting and proofreading services, according to the guidelines laid out in the university-endorsed national Guidelines for editing research theses.

xi Financial support

No financial support was provided to fund this research.

Keywords

Testosterone, cortisol, T/C ratio, salivary hormones, rugby union, competition readiness, competition

Australian and New Zealand Standard Research Classifications (ANZSRC)

ANZSRC code: 110602, Exercise Physiology, 60%

ANZSRC code: 110699, Human Movement and Sports Science not elsewhere classified, 40%

Fields of Research (FoR) Classification

FoR code: 1106, Human Movement and Sport Science, 100%

xii Table of Contents

Abstract ...... II Declaration by author ...... iv Publications during candidature ...... v Peer-reviewed papers ...... v Publications included in this thesis ...... vi Contributions by others to the thesis ...... x Statement of parts of the thesis submitted to qualify for the award of another degree ...... x Research Involving Human or Animal Subjects ...... x Acknowledgements ...... xi Financial support ...... xii Keywords ...... xii Australian and New Zealand Standard Research Classifications (ANZSRC) ...... xii Fields of Research (FoR) Classification ...... xii List of Tables ...... 5 List of Figures ...... 6 List of Abbreviations ...... 7 1. Introduction ...... 9 1.1 Background ...... 9 1.2 Statement of the problems ...... 12 1.3 Research aims, purpose and hypotheses ...... 12 1.4 Research design ...... 13 1.5 Significance of this research ...... 14 1.6 Thesis organisation ...... 15 2. Review of Literature ...... 16 2.1 Introduction ...... 16 2.2 Secretory control of testosterone and cortisol ...... 17 2.3 Steroid-binding proteins ...... 19 2.4 Testosterone, cortisol and the neuromuscular system ...... 20 2.5 Using saliva to measure testosterone and cortisol ...... 21 2.6 The testosterone to cortisol (T/C) ratio ...... 27 2.7 Testosterone, cortisol and athletic performance ...... 29 2.7.1 Neuromuscular-based activities: sprinting, jumping, power and strength ...... 31 2.7.2 Resistance training ...... 35 2.7.3 Testosterone and cortisol responses to concurrent training methodologies ...... 36 2.8 Competition: Acute response of testosterone and cortisol concentrations ...... 38 1 2.9 Summary and conclusions ...... 41 3. Relationship Between Pregame Concentrations of Free Testosterone and Outcome in Rugby Union ...... 45 3.1 Introduction ...... 45 3.2 Methods ...... 46 3.2.1 Subjects ...... 46 3.2.2 Design ...... 47 3.2.3 Hormone Assessment ...... 47 3.2.4 Game Day Performance ...... 47 3.2.5 Statistical Analyses ...... 48 3.3 Results ...... 49 3.4 Discussion ...... 52 3.5 Practical Applications ...... 54 4. Relationship Between Match Statistics, Game Outcome and Pre-Match Hormonal State in Professional Rugby Union ...... 55 4.1 Introduction ...... 55 4.2 Methods ...... 57 4.2.1 Participants ...... 57 4.2.2 Design ...... 57 4.2.3 Game Day Performance ...... 57 4.2.4 Hormone Assessment ...... 59 4.2.5 Statistical Analyses ...... 60 4.3 Results ...... 60 4.4 Discussion ...... 62 4.5 Conclusion ...... 65 4.6 Practical Applications ...... 65 4.8 Acknowledgements ...... 66 5. Relationship Between Midweek Training Measures of Testosterone and Cortisol Concentrations and Game Outcome in Professional Rugby Union Matches ...... 67 5.1 Introduction ...... 67 5.2 Methods ...... 68 5.2.1 Experimental Approach to the Problem ...... 68 5.2.2 Subjects ...... 68 5.2.3 Assessment Procedures ...... 69 5.2.4 Midweek Skill-Based Training Session ...... 69 5.2.6 Hormone Assessment ...... 70 5.2.7 Statistical Analyses ...... 71 5.3 Results ...... 71 2 5.3.1 Hormone Concentrations vs Wins and Losses ...... 71 5.3.2 Hormone Responsiveness vs Wins and Losses ...... 72 5.3.3 Midweek Hormones vs Pregame Testosterone ...... 72 5.3.4 Game Data - Rate of Perceived Exertion ...... 73 5.4 Discussion ...... 73 5.5 Practical Applications ...... 75 5.6 Acknowledgements ...... 76 6. Salivary Testosterone and Cortisol Response to Four Different Rugby Training Exercise Protocols 77 6.1 Introduction ...... 77 6.2 Methods ...... 79 6.2.1 Subjects ...... 79 6.2.2 Design ...... 79 6.2.3 Hormone assessment ...... 79 6.2.4 Exercise protocols ...... 80 6.2.5 Statistical analyses ...... 80 6.3 Results ...... 81 6.3.1 Testosterone concentration - group response ...... 81 6.3.2 Testosterone concentration - individual response ...... 82 6.3.3 Cortisol concentration - group response ...... 82 6.3.4 T/C ratio concentration - group response ...... 83 6.4 Discussion ...... 83 6.5 Practical applications ...... 87 7. General Discussion ...... 88 7.1 Summary and discussion ...... 88 7.1.1 Assessing the measures of salivary testosterone and cortisol concentrations according to game outcome ...... 88 7.1.2 Determining if performance indicator scores discriminate between game outcomes and correlate with game-day salivary hormone concentrations ...... 89 7.1.3 Evaluating mid-week salivary hormone concentrations and subsequent match outcome ...... 90 7.1.4 Evaluating the acute response of salivary testosterone and cortisol concentrations to four rugby training exercise protocols ...... 91 7.2 Practical applications ...... 92 7.3 Limitations and delimitations ...... 93 7.4 Directions for future research ...... 93 7.5 Conclusion ...... 94 8. Reference List ...... 96 APPENDIX 1: PARTICIPANT INFORMATION DOCUMENT ...... 118 3 Appendix 2: Participant Information Document ...... 123 Appendix 3: Consent Form ...... 126

4 List of Tables

Table 3.1 Physical Characteristics of the Rugby Union Forwards and Backs ...... 47 Table 3.2 Results and Classifications for Each of the Rugby Games Played ...... 48 Table 3.3 Salivary Hormone Concentrations of Elite Rugby Union Players on Game Day ...... 49 Table 3.4 Game-Day Information for Elite Rugby Union Players ...... 49 Table 3.5 Comparison of Hormone Concentrations According to Game Venue (Home and Away) ...... 50 Table 3.6 Comparison of Salivary Hormones According to Outcome (Win or Loss) for All Games ...... 50 Table 4.1 Comparison of match performance indicator scores of elite rugby union matches according to outcome...... 60 Table 5.1 Midweek skill training session outline and timing of events...... 70 Table 5.2 Subsequent competitive game performance after the midweek skill-based training session ...... 70 Table 5.3 Correlations between the relative (%) hormonal changes across each skill-based training session and pregame testosterone concentration ...... 73 Table 6.1 Cortisol concentration and T/C ratio group response for each protocol...... 83 Table 6.2 Percentage change in PRE to POST hormone concentrations for each protocol ...... 83

5 List of Figures

Figure 2.1 Schematic representation of the hypothalamic-pituitary-gonadal axis...... 18 Figure 2.2 Schematic representation of the hypothalamic-pituitary-adrenal axis...... 19 Figure 2.3 Illustration of the transportation mechanisms of biomolecules from blood capillaries into saliva...... 22 Figure 3.1 Relationship between pregame (Pre) salivary testosterone and game-ranked outcome...... 51 Figure 3.2 Relationship between the pregame (Pre) salivary testosterone-to-cortisol (T/C) ratio and game- ranked outcome ...... 51 Figure 5.1 Preskill and postskill training salivary testosterone to cortisol (T/C) ratio before winning and losing games (mean ± SE) ...... 71 Figure 6.1 Salivary testosterone concentrations before (PRE) and immediately after (POST) four exercise protocols (5 × 15–55%; 3 × 5–85%; strongman – STRNG; combative – COMB) ...... 81 Figure 6.2 Salivary testosterone concentrations grouped according to the protocol that produced the largest absolute change before (PRE) and immediately after (POST) four exercise protocols (5 × 15–55%: n = 8; 3 × 5–85%: n = 2; strongman - STRNG: n = 6; combative - COMB: n = 11) ...... 82

6 List of Abbreviations

°C degrees Celsius ACTH adrenocorticotropic hormone AggPI aggressive performance indicator ATP adenosine triphosphate C cortisol Ca2+ calcium CBG corticosteroid-binding globulin CK creatine kinase cm centimetres CNS central nervous system COMB combative CRH corticotropin-releasing hormone ES effect size FSH follicle-stimulating hormone GnRH gonadotropin-releasing hormone h hour HPA hypothalamic-pituitary-adrenal HPG hypothalamic-pituitary-gonadal kg kilogram LH luteinizing hormone m metre min minute min/km minutes per kilometre mL millilitres nmol⋅L-1 nanomoles per litre NSCA National Strength and Conditioning Association pg⋅mL-1 picograms per millilitre PI performance indicator PNS peripheral nervous system RE resistance exercise REM rapid eye movement

7 RM repetition maximum RPE rate of perceived exertion SD standard deviation SE standard error SHBG sex-hormone-binding globulin SkAggPI skill and aggressive performance indicator SkillPI skill performance indicator STRNG strongman T testosterone T/C testosterone to cortisol UK United Kingdom USA United States of America • VO2max maximal oxygen consumption y year µg⋅dL-1 micrograms per decilitre µmol⋅L-1 micromoles per litre

8 1. Introduction

1.1 Background

Rugby union is a sport characterised by a high degree of aggressive impacts and interactions (Beaven et al., 2008b, Duthie et al., 2003). A team of rugby union players consists of two distinct playing units, ‘forwards’ and ‘backs’. In general, the forwards are engaged in intense activity, pushing and competing for the ball whilst the backs provide support in attack and cover in defence. The backs cover a greater distance during a game and spend more time in intense running than the forwards who are involved in more contact situations (Duthie et al., 2003). Competitive game play is physiologically demanding (Duthie et al., 2003, Gill et al., 2006, Cunniffe et al., 2009), highlighting the need to monitor and understand the players’ response to training and competition (Duthie et al., 2003). Recent work understanding the demands of rugby union has focused on using endocrine status, in particular, salivary testosterone and cortisol, as a marker of physiological response and readiness for game-day performance (Argus et al., 2009, Elloumi et al., 2003, Cunniffe et al., 2011, Cook and Crewther, 2012b), longitudinal season monitoring (Elloumi et al., 2008), training (Beaven et al., 2008a, Beaven et al., 2008b) and general athleticism (Crewther et al., 2009b, Crewther et al., 2012). Therefore, some researchers have suggested that the acute effects of testosterone and cortisol could play a central role in moderating training adaptation and athletic performance (Beaven et al., 2008a, Beaven et al., 2008b, Crewther et al., 2009b, Cook and Crewther, 2012a), and have thus promoted the usefulness of monitoring these hormones.

In general, the roles and interaction of testosterone and cortisol are complex in nature. Both hormones are affected via multiple mechanisms, including, but not exclusive to, psychological (Archer, 2006, Archer, 1991), social (Archer, 1991, Mazur, 1985), physiological (Viru et al., 1992, Urhausen et al., 1995) and athletic (Booth et al., 1989, Elias, 1981, Mazur and Lamb, 1980, Suay et al., 1999) events. Further, the interplay and moderating effects of both cortisol and testosterone on each other adds to this complexity (Mehta and Josephs, 2010). Additionally, the training status of an athlete is another consideration, as recent observations suggest that recreational athletes may not display the same response as elite athletes to similar stimuli (Morton et al., 2016, Crewther et al., 2012). Increases in the concentration of endogenous testosterone across sporting competitions have been identified as a marker of success (Obmiński et al., 1998, Obmiński, 2009, Elias, 1981, Fry et al., 2011, Booth et al., 1989). Researchers have also related elements of desirable athletic qualities associated with successful competitive outcomes to elevated concentrations of endogenous

9 testosterone. These factors include neuromuscular expression of speed, power and/or strength (Crewther et al., 2009b, Cardinale and Stone, 2006); mood states such as competitiveness, drive, and persistence (Dabbs, 1993, Henry, 1992); and aggressive and dominant behaviours (Bouissou, 1983, Dixson, 1980, Salvador et al., 1999).

Most studies investigating the role of endogenous hormones before and after a competitive event typically focus on acute measures specifically at that time point. Few studies have investigated mid- week measures of these endogenous hormones and subsequent competition performance a number of days later. A recent study of a professional team demonstrated that testosterone responses to mid-week strength training workouts showed significant associations with subsequent game outcome 3 to 4 days later (Crewther et al., 2013). In a contact-based team sport, this was the first study to highlight the association between mid-week hormone concentrations and performance outcome. In rugby union, an anticipatory salivary cortisol response prior to a mid-week test discriminated between match outcomes in international rugby union competition (Crewther et al., 2017). A mid-week repeat shuttle run test was completed either 3 or 4 days before seven international rugby union matches. Salivary testosterone and cortisol concentrations were monitored in the morning and in the afternoon before and after the test. A large rise in morning to pre-test cortisol concentrations was observed prior to winning, whereas cortisol concentrations decreased before losing. Apart from the different mid-week protocols used in these two studies, differences in team culture, management strategies and game demands between the club and international environments were cited as possible reasons for the observed differences (Crewther et al., 2017). These studies highlight the role that endogenous hormones can play in the early detection of team readiness to compete (or the lack thereof) (Crewther et al., 2013). However, no study has examined skill-based training of rugby players, despite the importance of this type of training. The preparation of rugby union players involves both mental and physical components (Passos et al., 2008), hence the use of a mixed psychological-physical stressor such as a skill-based training session may be a more applicable test. If such a relationship exists, these findings will support the role of mid-week salivary endogenous hormones in the early assessment of a team’s readiness to compete.

To meet the increasing physical demands of professional rugby, the body size of elite rugby players has increased significantly over the past three decades (Duthie et al., 2003, Sedeaud et al., 2012). Strength training has played a large role in this change. The prescription of strength training workouts is based on the principle that acute elevations in endogenous hormones increase the likelihood of receptor interactions (Beaven et al., 2008a), offering both short-term benefits for training performance and long-term adaptations (e.g. muscle hypertrophy) (Ahtiainen et al., 2004).

10 West and Phillips (2012) suggested that acute changes in testosterone may reflect an athlete’s motivation in training effort, which acts as a driver to improve both performance and morphological changes. Over the course of a competitive season, the potential of elite rugby players to induce muscle hypertrophy and related performance changes through training are limited (Beaven et al., 2008a). Therefore, the specificity of, and response to training sessions becomes an important factor to maximise the physiological improvement and performance of an athlete. An investigation of rugby union players who undertook resistance workouts using several different protocols, demonstrated that the hormonal response to a series of acute exercise bouts was specific to the workout design and that no two athletes followed the same response pattern for a given exercise bout (Beaven et al., 2008b). In a subsequent study, these athletes trained according to which workout design would produce either the maximal or the minimal changes in salivary free testosterone concentrations (Beaven et al., 2008a). Using the maximal testosterone workout protocol (irrespective of the design), athletes achieved significant increases in both strength and body mass over a 3-week period, whereas training with the minimal testosterone workout protocol had either no effect upon or reduced strength in most players. Due to the importance placed on resistance exercise training in rugby union, these findings provide support for the monitoring of testosterone as a marker for selecting specific training protocols (Bhasin et al., 2001, Beaven et al., 2008a). Other popular training sessions used in rugby union are based around the concept of strongman and combative (boxing and wrestling) training. Despite their popularity, no studies to date have concurrently investigated the hormonal response to, and the effectiveness of both strongman and combative training regimens in comparison to resistance-based training.

This thesis was designed to investigate the association between salivary testosterone and cortisol over the course of a rugby union season, particularly the relationship between mid-week and game- day hormone concentrations and performance. Further investigations sought to examine practical interventions that have the potential to maximise functional gains of training stimuli via hormone monitoring. Understanding such relationships and interventions may ultimately assist with optimising the players’ training content, structure and specificity with a view to improving their training and competition performance.

11 1.2 Statement of the problems

Testosterone and cortisol have been reported to positively contribute to physical (neuromuscular) and behavioural qualities important for playing performance and success. These endogenous hormones contribute to these qualities through both short-term and long-term mechanisms and these processes are complex in nature. Although research has investigated these hormones in isolated performance and training studies, there are limited studies on a single cohort of athletes within a single competition season, in particular, within the sport of rugby union. Currently, objective physiological measures of training efficacy and player readiness are scarce. Therefore, examining the relationship between endogenous hormone concentrations and both competitive rugby union game play and various training interventions may have relevance for several reasons: to predict game outcome; for early detection of athletes’ readiness to compete; and to optimise exercise training regimes for individual athletes.

1.3 Research aims, purpose and hypotheses

The aim of this research was to assess the short-term effects of testosterone and cortisol on competitive readiness and to identify optimal training protocols in elite rugby players. Four specific aims were to:

1. Measure salivary testosterone and cortisol concentrations across a series of consecutive rugby union matches and determine if any relationship exists with the game outcome.

2. Determine if performance indicator (PI) scores generated from rugby union game statistics discriminated between match outcome and determine the PI scores relationship with game- day salivary testosterone and cortisol concentrations.

3. Evaluate the salivary testosterone and cortisol concentrations in response to selected mid- week skill-based training sessions and evaluate their relationship with the subsequent match outcome.

4. Evaluate the acute response of salivary testosterone and cortisol concentrations to four different exercise protocols for rugby training.

The following hypotheses were proposed:

1. Salivary testosterone concentration and the testosterone/cortisol (T/C) ratio will be higher before a win than before a loss and that both measures will significantly correlate with a game-ranked performance score.

12 2. An aggressive-based PI score will discriminate between game outcomes and will also be significantly correlated to pregame testosterone concentration.

3. Change in the pre to post testosterone concentrations and T/C ratios across a mid-week skill- based training session will be significantly greater when subsequent rugby union matches are won rather than lost.

4. The acute testosterone response to four different exercise protocols will differ amongst athletes, and therefore standard group data will not reflect the individual response.

1.4 Research design

This thesis reports on four individual experiments:

1. Over an in-season period of 6 weeks, a team of rugby union players provided pregame and postgame saliva samples. Salivary testosterone and cortisol concentrations were measured and compared according to competition outcome. These competition outcomes included wins and losses and a numerical game-ranked performance score from 1 to 6 (1 = a ‘bad’ loss reflected by poor skill and execution, 2 = an ‘unlucky’ loss that could have been won due to a close score or a game lost in the last few minutes, 3 = draw, 4 = a win reflected by an ‘average’ performance, 5 = a ‘good’ win where the team played well against the opposition and 6 = a ‘good +’ win reflected by a dominant performance).

2. For each game played throughout the competitive season, three PI scores were generated based upon skill actions (SkillPI), aggressive actions (AggPI) and summation of AggPI and SkillPI (SkAggPI). The PI scores for each player and game were grouped according to outcome (i.e. win and loss) and compared. Across a subset of 6 consecutive games, pregame and postgame hormone concentrations were measured to determine any consistent relationship with the PI scores.

3. Salivary testosterone and cortisol samples were collected before and after a mid-week training session over 6 consecutive in-season weeks. Mid-week concentrations were pooled according to outcome of the subsequent weekend’s game and compared. The relationship between mid-week and pregame hormone concentrations was also investigated.

4. A group of elite male rugby players performed four different exercise protocols in random order over 4 weeks during the competitive season. Saliva samples were collected before and immediately after each exercise protocol to assess the acute response of salivary testosterone

13 and cortisol concentrations. Data were also grouped and analysed according to the protocol that demonstrated the greatest absolute testosterone response for each athlete.

As this was an in-season professional rugby environment, practicalities of normal training had to be observed. Although applied research has its limitations (Cairns et al., 2005), the findings are relevant for the coach regarding the transfer and adaption of research outcomes (Bishop, 2008). Therefore, it can be argued that, if a greater number of relevant training and competition variables were monitored within a population of elite athletes, a better understanding of the components necessary for success may be gained.

1.5 Significance of this research

The primary objective was to assess the relationship between salivary hormone concentrations and competitive performance. A link between salivary hormones and performance at an elite level is not currently available in the literature. This information is not currently available at an elite level. Furthermore, pregame testosterone as a potential predictor of game outcome and performance readiness in sport has not been examined or reported in the published literature. The athletes within this series of studies were from an elite professional rugby team who competed in both the English Rugby Union Premiership and the European Championship, with a number of the subjects also playing at international level. The ability to demonstrate a significant relationship between testosterone and performance outcome would be useful for providing relevant feedback to the players and coaches. If training and psychological stimuli can be shown to acutely elevate testosterone, measuring testosterone concentrations has implications for providing appropriate and timely interventions to improve competition outcomes.

The second major objective was to gain a better understanding of the relationship between mid- week hormone levels and their responsiveness to a stimulus and subsequent competition a few days later. This analysis has potential implications for understanding the interplay between hormone concentrations and athletic performance throughout the entire training week, rather than simply immediately before or after competition. Establishing an effective method (mid-week) for determining an athlete’s readiness to compete may have implications regarding the structure and content of the remainder of the training week.

The third major objective was to examine typical rugby training regimes to better understand the effectiveness of these strategies and thus improve the training practices of elite rugby players. The challenge with training in team-based sports lies around the need to train as a group rather than address the needs of each player with an individualised program. Hence, exercise-training regimes 14 are typically prescribed to athletes under the assumption that they will be of benefit (physiologically and psychologically) for all group members. The ability to individually identify those regimes that are of maximum benefit to each individual athlete has potential benefits for coaches, who may be able to use this information within the demands of a group-based training session. A better understanding of a rugby player’s current training and competition environment through hormone assessment may also provide an impetus to investigate the use of alternative interventions (e.g. visual, psychological, nutritional) to further improve the training and competition environment.

1.6 Thesis organisation

This thesis consists of 7 chapters. Chapter 2 is a review of literature that discusses the relationships between salivary steroid hormones and the physical and psychological responses in response to, and leading in to training and competition. Chapters 3 to 6 present the experimental manuscripts of each of the four individual studies. Due to the similar methods incorporated within the four studies, there is some repetition of information within these chapters. Chapter 7 is a general discussion of the experimental findings and outlines practical applications, limitations and delimitations, and directions for future research.

15 2. Review of Literature

2.1 Introduction

Steroid hormones are familiar regulators of physiological processes. There are five groups of steroid hormones that are generally categorised according to their biological actions: mineralocorticoids, glucocorticoids, progestins, oestrogens and androgens (Miller, 1988). These classes of steroid hormones are structurally similar and arise from a common series of pathways, yet differ in their actions via one or more specific steroid hormone receptors (Miller, 1988). They are converted from cholesterol and its esters through the action of specific enzymes into active steroid hormones. Testosterone is the primary circulating androgen in human males, whereas cortisol is the main glucocorticoid. The physiological roles of testosterone can be divided into two major categories: (a) those related to reproductive function and the development of male secondary sex characteristics (i.e. androgenic effects); and (b) those that pertain more generally to stimulation of tissue growth and development (i.e. anabolic effects) (Hackney, 1996). The primary function of cortisol is to regulate the stress response through psychological and physiological allostatic mechanisms such as immune functioning, cardiovascular activation, glucose metabolism and emotional processing (Schulkin et al., 1994, Sapolsky et al., 2000). Cortisol, similar to testosterone (Dabbs, 1990), has a circadian rhythm, peaking in the morning upon wakening and gradually declining throughout the day in response to a combination of diurnal and metabolic inputs (Weitzman et al., 1981).

It has been proposed that testosterone may influence neuromuscular transmission (Blanco et al., 1997, Leslie et al., 1991) as well as receptor numbers at the neuromuscular junction, suggesting that the effects of testosterone on athletic performance also involve changes outside skeletal muscle cells (Cardinale and Stone, 2006, Bleisch et al., 1982). Animal-based studies further highlight the role of testosterone in skeletal muscle excitation-contraction coupling and in ‘phenotypisation’ of fast twitch fibres (Cardinale and Stone, 2006). These findings support the theory that testosterone concentration plays a role in developing a high rate of force in muscular activity within a short contraction time (Cardinale and Stone, 2006). Testosterone concentration has been related to the neuromuscular expression of speed, power and strength (Crewther et al., 2009b, Cardinale and Stone, 2006). Mood states such as competitiveness, drive, persistence and contribution to winning are also associated with high testosterone concentrations (Dabbs, 1993, Henry, 1992), along with aggressive and dominant behaviours (Bouissou, 1983, Dixson, 1980, Salvador et al., 1999).

16 Cortisol is known to have a protein-catabolic effect on skeletal muscle (Urhausen et al., 1995). During prolonged exercise, an increase in cortisol is important for energy supply because cortisol prevents the re-esterification of fatty acids released by the catecholamine-induced lipolysis (Galbo, 1986, Holloszy and Coyle, 1984). After exercise, it is hypothesised that cortisol also assists with muscular glycogen resynthesis (Urhausen et al., 1995). A cortisol effect on the central nervous system has also been reported to decrease the rapid eye movement (REM) phase of sleep and the time spent awake (Aronson et al., 2003). Mental stressors (in the absence of physical effort) (Herman et al., 2003) and food intake (Gibson et al., 1999) can also affect cortisol secretion.

The aim of this chapter is to provide background information on steroid hormones and to examine the evidence around the role and/or marker ability of salivary testosterone in relation to competition and athletic training. Sections 2.2 and 2.3 examine the secretory control of testosterone and cortisol and the roles of steroid binding proteins, respectively. Section 2.4 highlights the importance of the neuromuscular system to movement and the role of steroid hormones in helping to moderate activity within this system. Section 2.5 examines the measurement of steroid hormones in saliva and considerations around its use as a valid biomarker. Section 2.6 discusses the testosterone to cortisol ratio and its use within sporting studies. Section 2.7 examines the role of testosterone and cortisol across a variety of athletic performance activities and section 2.8 discusses hormone responses across competition and associated outcomes. For clarity, the research discussed during this chapter will primarily focus on male populations. Where appropriate, and to assist with further understanding the relationships between steroid hormones and their effects, the relationships will be described according to a categorisation of r values: coefficient values ≤.35 are generally considered to represent low or weak correlations, values from .36 to .67 modest or moderate correlations, values from .68 to 1.0 strong or high correlations and those values ≥.90 very high correlations (Taylor, 1990).

2.2 Secretory control of testosterone and cortisol

Testosterone is secreted primarily from the Leydig cells of the testes which accounts for 95% of total secretion (Coffey, 1988). Secretion is regulated through the positive and negative feedback control of the hypothalamic-pituitary-gonadal (HPG) axis (Eik-Nes, 1964) (Figure 2.1). The HPG axis comprises the hypothalamus and its neural connections with the rest of the brain, the pituitary and the testes or ovaries (Griffin and Wilson, 1998). Secretion of the peptide gonadotropin- releasing hormone (GnRH) within the anterior hypothalamus and preoptic area stimulates the anterior pituitary to produce and release luteinizing hormone (LH) and follicle-stimulating hormone

17 (FSH). These hormones then bind to receptors in the testes and ovaries to assist with the maintenance of testicular function (in males) and the production of testosterone (Hackney, 1996).

Figure 2.1 Schematic representation of the hypothalamic-pituitary-gonadal axis. Key: GnRH = gonadotropin-releasing hormone, FSH = follicle-stimulating hormone, LH = luteinizing hormone.

Cortisol secretion is governed via a feedback mechanism involving the hypothalamic-pituitary- adrenal (HPA) axis (Figure 2.2). The HPA axis constitutes one of the principle pathways of the stress response (Tarullo and Gunnar, 2006). In response to a stressor, the hypothalamus produces corticotropin-releasing hormone (CRH), which is carried to the anterior pituitary. Upon binding of CRH, adrenocorticotropic hormone (ACTH) is released by the pituitary into the systemic bloodstream to be made available to the adrenal glands to stimulate the release of cortisol. Negative feedback occurs when the elevation of blood-borne cortisol blunts the release of ACTH from the pituitary and/or the secretion of CRH from the hypothalamus (Kraemer and Rogol, 2008).

18

Figure 2.2 Schematic representation of the hypothalamic-pituitary-adrenal axis. Key: CRH = corticotropin-releasing hormone, ACTH = adrenocorticotropin hormone.

2.3 Steroid-binding proteins

Steroids have lipophilic properties and are mostly bound (specific and non-specific) to carrier proteins in the blood (Rosner, 1991). Steroid hormones circulate in complex with three major plasma proteins: sex-hormone-binding globulin (SHBG) with testosterone, corticosteroid-binding globulin (CBG) with cortisol and non-specific binding with albumin (Dunn et al., 1981). Specific binding to the plasma protein globulins (high affinity, low capacity) accounts for 44–66% of total testosterone and 90% of total cortisol concentrations, while non-specific binding to plasma albumin (low affinity, high capacity) accounts for 30–50% of total testosterone and 6–7% of total cortisol concentrations (Dunn et al., 1981). The remaining steroid (1–4% of total hormone concentrations) circulates freely and hence is termed the free or unbound steroid (Crewther, 2009). Free testosterone and free cortisol concentrations are thought to represent the biologically active hormones that initiate the hormonal action at target cells (Rosner, 1991, Vining and McGinley, 1986). Another method to describe biologically active hormones is the combination of free plus albumin-bound hormones (Pardridge, 1986). Bioavailable hormones bound to albumin are readily dissociable and available for uptake by target cells, unlike the hormone portions bound to SHBG and CBG (Pardridge, 1986). Therefore, bioavailable testosterone and cortisol concentrations are thought to have greater physiological importance than the total and free hormones (Crewther, 2009).

19 2.4 Testosterone, cortisol and the neuromuscular system

The neuromuscular system is essential to human movement and performance. It is composed of the central nervous system (CNS), the peripheral nervous system (PNS) and a neural circuit including sensory neurons, motor neurons and skeletal muscle fibres. The CNS comprises the brain and spinal cord, whereas the PNS refers to everything distal to the spinal cord (Enoka, 2002). The functional unit of the neuromuscular system is the motor unit and during voluntary movement, neural signals are transmitted from the brain down the spinal cord to activate motor neurons and stimulate contraction of the muscle fibres (Jeffreys and Mood, 2016). In turn, the muscles generate force to assist with normal activities (e.g. standing, walking) and sport-specific activities (e.g. jumping, running) (Enoka, 2002).

The neuromuscular system is an important endocrine target for testosterone and cortisol (Crewther et al., 2011), which are purported to moderate neuronal activity in both the CNS and the PNS (Enoka, 2002). Within the CNS this action may explain the association between steroid hormones and mood, behaviour and cognitive function (Crewther, 2009). Animal studies have demonstrated rapid (i.e. within minutes) hormone-induced behavioural changes that originate from neuronal activity (Aikey et al., 2002, Schiml and Rissman, 1999). More delayed (i.e. within hours) hormone- induced changes in behaviour and cognitive function have been demonstrated in human studies (Aleman et al., 2004, Buss et al., 2004), although this may have been a result of the methods incorporated rather than an inherent delay in the neural-behavioural transduction process (Crewther, 2009). Animal studies have also shown that brain regions accumulate sex steroids, with the potential to stimulate the growth of neuronal processes and remodel neural circuits in the adult brain and spinal cord (Stumpf, 1970, Arnold et al., 1976).

Within the PNS, intracellular messengers (e.g. Ca2+, protein kinase, G-proteins) have been identified as key mediators of the rapid steroid actions. The in-vitro application of testosterone and cortisol has been shown to produce rapid changes in intracellular Ca2+ concentrations in skeletal muscle cells (Estrada et al., 2003, Passaquin et al., 1998). The Ca2+ ion is important due to its role in muscle contraction, relaxation after the twitch, regulation of energy metabolism and maintenance of the structural integrity of the muscle fibre (Berchtold et al., 2000). Furthermore, channel proteins that release Ca2+ from the sarcoplasmic reticulum are also sensitive to testosterone (Anttila et al., 2008). Energy metabolism is an additional mechanism to explain the short-term hormonal effect upon the PNS (Crewther, 2009). Cortisol stimulates gluconeogenesis and glycogenolysis via glycogen, protein and lipid metabolism, and through its permissive actions on other hormones (Viru and Viru, 2004). The use of exogenous testosterone may enhance skeletal muscle activity by 20 augmenting the supply of energy through increased activity of the enzymes creatine phosphokinase and myokinase and increased adenosine triphosphate (ATP) concentration (Ramamani et al., 1999). The role of endogenous testosterone, however, is not as clear; the metabolic effects are more predominantly involved in anaerobic metabolism and the recruitment of fast-twitch, type II muscle fibres (Ramamani et al., 1999, Crewther, 2009).

2.5 Using saliva to measure testosterone and cortisol

Serum hormones can reach saliva by transfer mechanisms that include intracellular and extracellular pathways (Figure 2.3) (Gatti and De Palo, 2011). For steroids, the most common intracellular route is passive diffusion into the acinar cells of the secretory end piece of the salivary gland (Vining et al., 1983b). A serum molecule reaching saliva by diffusion must cross 5 barriers; the capillary wall, the interstitial space, the basal cell membrane of the acinus cell or duct cell, the cytoplasm of the acinus or duct cell and the luminal cell membrane. Active transport is a second intracellular pathway for the entry of molecules into saliva. Despite this mechanism, the role of active transport of hormones into saliva appears to be equivocal (Vining and McGinley, 1986). Ultrafiltration (an extracellular mechanism) is a third method of transportation and involves filtration of molecules from the blood stream through the spaces between acinus and ductal cells into saliva (Pfaffe et al., 2011). Serum proteins such as albumin or CBG are too large to pass through the membranes of the salivary cells. These proteins bind the majority of steroids in blood; therefore, only the free, unbound and unconjugated steroid molecules are able to pass into saliva (Vining and McGinley, 1987). Consequently, salivary concentrations of steroids are lower than those measured in blood (Gozansky et al., 2005). Salivary cortisol concentrations are only 50–60% of the corresponding serum free cortisol concentration (Wood, 2009) and salivary testosterone levels reflect between 1– 3% of the total and free serum concentrations (Wang et al., 1981, Vermeulen et al., 1971, Fiers et al., 2014).

21

Figure 2.3 Illustration of the transportation mechanisms of biomolecules from blood capillaries into saliva.

The main advantage of using saliva to analyse endogenous hormones is that sample collection is relatively non-invasive, stress free and convenient, whereas blood collection is either undesirable or can often be difficult on a regular basis in elite athlete populations (Lewis, 2006, Crewther et al., 2011). Furthermore, saliva collection requires minimal training and can be performed on the sports field or in challenging environments. For professional athletes, these advantages alone hold a valuable perspective for compliance of routine and serial hormone monitoring. Various devices can be used to collect saliva and the sample can be collected with or without stimulation (Groschl, 2008). Athletic population studies generally collect saliva either into a receptacle tube using a ‘passive drool’ method (Beaven et al., 2008b, Cadore et al., 2008, Crewther et al., 2008), or by using a cotton swab placed in the mouth in combination with a specialised saliva-collecting tube (Filaire et al., 2009, Lippi et al., 2009). A study comparing salivary testosterone concentrations and serum free testosterone concentrations using saliva collected with these two common methods demonstrated a strong correlation (r = .71) with the passive drooling method but only a moderate correlation (r = .43) with the alternative method (Fiers et al., 2014). These results suggest passive drooling as the collection method of choice for measuring testosterone in saliva.

22 Sample quality and storage are other important considerations when testing saliva. Food and blood contamination (e.g. brushing teeth, contact sports) can affect steroid measurements (Vining et al., 1983b, Davis and Green, 2009), thus highlighting the importance of saliva sample quality. Storage duration and the temperature at which the samples are stored also affect the stability of salivary steroid hormones. Salivary steroids are stable for up to 1 month at +4 °C, up to 3 months at –20 °C and up to 1 year at −80 °C (Groschl, 2008, Wood, 2009, Granger et al., 2004).

Correlation studies comparing salivary and serum hormone concentrations have reported significant and strong relationships for both testosterone and cortisol when measured in clinical populations across single and multiple time points throughout the day (Table 2.1). Cortisol in particular demonstrates very strong relationships between serum and salivary concentrations (Vining et al., 1983a, Umeda et al., 1981, Peters et al., 1982, Gallagher et al., 2006, Lippi et al., 2009). Testosterone, despite rapid daily fluctuations in an individual’s salivary concentrations, displays a good overall correlation between salivary and serum concentrations (Fiers et al., 2014, Goncharov et al., 2006, Johnson et al., 1987, Baxendale et al., 1980, Sannikka et al., 1983, Wang et al., 1981). The observed variations in hormone concentrations amongst studies are attributed to the difference in methods, subjects and time of sampling (Wood, 2009). As such, normative ranges for salivary cortisol concentrations vary between 2.8–27.5 nmol⋅L-1 and salivary testosterone concentrations between 175−870 pmol⋅L-1 in normal subjects (Walker et al., 1980).

There are significant correlations between salivary and serum based measurements of the testosterone to cortisol ratio (Obmiński and Stupnicki, 1997, Lippi et al., 2016). In a population of triathletes and athletes, Obmiński and Stupnicki (1997) observed strong correlations whereas Lippi et al. (2016) reported a moderate correlation in elite professional soccer athletes. Furthermore, Lippi et al. (2016) reported that 100% of the athletes declared that they would feel more comfortable to have saliva rather than blood periodically collected for long-term monitoring. Therefore, from a practical, financial and logistical viewpoint, collecting saliva over blood within an elite athlete population is advantageous.

23 Table 2.1 Correlation summaries between salivary and serum steroid concentrations in males. Salivary Reference Mean Saliva Correlated with r value Sample time Storage Steroid concentration points temp Cortisol Vining et al. (1983a) 5.7–9.6 nmol⋅L-1 Serum free # .97 Multiple –20 °C Serum total .92 Multiple

Umeda et al. (1981) 27.3 nmol⋅L-1 Serum free .89 Multiple –20 °C

Peters et al. (1982) 11 nmol⋅L-1 Serum free .97 Single –20 °C

Gallagher et al. 6.9 nmol⋅L-1 Serum total .85 Single –20 °C (2006)

Lippi et al. (2009) 7.2 nmol⋅L-1 Serum total .80 Single nil

Testosterone Sannikka et al. (1983) 450 pmol⋅L-1 Serum free .75 Multiple –20 °C Serum total .63 Multiple

Wang et al. (1981) 308 pmol⋅L-1 Serum total .94 Multiple –20 °C

Fiers et al. (2014) Not stated Serum free .71 Single –80 °C

Goncharov et al. 369 pmol⋅L-1 Serum free .75 Single –20 °C (2006)

Johnson et al. (1987) 226 pmol⋅L-1 Serum total # .83 Single –20 °C Serum total # .92 Single Baxendale et al. 290 pmol⋅L-1 Serum free .81 Single –20 °C (1980)

T/C Ratio Lippi et al. (2016) 12 × 103 Serum free .43 Single nil

Obmiński and 1.67 Serum total .87 Single nil Stupnicki (1997)

# included female subjects in the analysis, nil – samples analysed immediately, T/C = testosterone to cortisol ratio

The correlation of serum and salivary concentrations taken before and after exercise have demonstrated to be not as consistent as that observed during rest (Table 2.2). For example, Cadore et al. (2008) observed no significant correlation between serum and salivary testosterone at rest and after resistance exercise, whereas serum cortisol had a significant and moderate correlation with salivary cortisol concentrations. During an isokinetic lower body strength test, Paccotti et al. (2005) reported a non-linear relationship, suggesting an exponential relationship may exist between serum and salivary cortisol concentrations. Vining et al. (1983a) also observed a similar exponential relationship between salivary and serum cortisol with subjects at rest and suggested that this rapid increase in salivary concentrations was reflective of serum CBG saturation.

24 Tanner et al. (2014) examined the acute response of plasma and salivary cortisol and testosterone to three training protocols in endurance trained male athletes. The three training protocols included a circuit of bodyweight exercise and two treadmill running sessions of different intensities. All of the data was pooled for analysis. A moderate and positive correlation (r = .81) was reported for plasma and salivary cortisol. Furthermore, an exponential relationship reported the same r-value as the linear regression, validating the suitability of a linear correlation for the comparison of salivary and plasma cortisol concentrations. Plasma and salivary testosterone concentrations demonstrated a weak to moderate correlation between individuals (r = .54) with large inter-individual variations also reported.

In response to short high-intensity cycling, Crewther et al. (2010) observed weak to moderate correlations (r = .57–.61) between grouped salivary and plasma testosterone concentrations. Intra- individual analysis demonstrated a stronger relationship (r = .71–.72), suggesting an individual hormone response to exercise in both blood (Dunn et al., 1981) and saliva (Crewther et al., 2010). Crewther et al. (2010) also reported that the relative peak increase in salivary testosterone exceeded the relative increases in total and free testosterone in plasma, suggesting that salivary measures may be more sensitive for assessing the hormonal response to exercise. Strong correlations were found between the salivary and plasma cortisol concentrations for both grouped and intra-individual data.

Lane and Hackney (2015) investigated the association between salivary and serum testosterone concentrations before and after 30-minute cycling sessions at three different intensities set at 40% • (low), 60% (moderate) and 80% (high) of VO2max. A strong correlation was observed between salivary and serum testosterone concentrations when exercising at both moderate (r = .91) and high (r = .89) intensities. These findings suggest a greater validity between the saliva and serum responses to higher intensity exercise (Lane and Hackney, 2015).

25 Table 2.2 Correlation summaries between salivary and serum steroid concentrations in exercising males Salivary Reference Subjects Exercise test Correlated r value Sampling (min) Steroid with Pre Post ** Cortisol Cadore et al. 28 healthy RE 75% 1RM Serum free .52Pre 0 10 *** (2008) (25min) .62Post

*** Port (1991) 6 MD Cycle Inc Serum free .86 0 0, 10 (active) (< 20min)

Paccotti et al. 11 NC Isokinetic LB Serum free .62*** 15, 5 7–120# (2005) 20 CA (15min)

*** ^ Crewther et al. 9 healthy Cycle Wingate Serum total .81 Grp 15, 0 1–60 *** (2010) (30sec) .83 Ind *** Serum free .84Grp *** .84Ind

O’Connor and 8 healthy Serum total ** 15, 0 0, 15 Cycle 75%V• O2max .6−.93 Corrigan (1987) (30min)

*** Tanner et al. 10 ET Circuit, RunTempo Serum total .81 0 0, 15, 30,

(2014) Run Interval (30min) 60

*** Stupnicki and 78 CA Run Inc, Ergo Inc Serum total .69Pre 0 2 *** Obmiński Cycle .54Post (1992)

*** VanBruggen et 12 ET Cycle Low, Mod, High Serum total .73 0 0, 30 al. (2011) (30min)

* Testosterone Cadore et al. 28 healthy RE 75% 1RM Serum free .22Pre 0 10 * (2008) (25min) .26Post

*** Lane and 12 ET NC Cycle Low, Mod, High Serum total .52Low 0 0, 30 *** Hackney (2015) (30min) .91Mod *** .89High

*** ^ Crewther et al. 9 healthy Cycle Wingate Serum total .57 Grp 15, 0 1– 60 ** (2010) (30sec) .72 Ind *** Serum free .60Grp ** .71Ind

*** Tanner et al. 10 ET Circuit, RunTempo Serum total .57 0 0, 15, 30,

(2014) Run Interval (30min) 60

Key: MD = middle distance, NC = non competitive, CA = competitive athletes, ET = Endurance trained, 75% 1RM = 75% of one repetition maximum, Circuit = bodyweight strength circuit, Run = treadmill running, Tempo = constant speed to coincided with calculated lactate threshold, Interval = interval treadmill run @ 90% • # VO2max, Pre = pre sample, Post = post sample, Inc = Incremental test to exhaustion, = multiple sample time ^ points @ 7, 15, 30, 45, 60, 90 & 120min, = multiple samples time points @ 1, 10, 20, 30 & 60min, 75%V• O2max • = 75% of VO2max, LB = lower body, Grp = grouped data, Ind = intra-individual data, Low = low intensity @ 40% • • • * ** *** VO2max, Mod = moderate intensity @ 60% VO2max, High = high intensity @ 80% VO2max, p <.05, p <.01, p <.001.

26 With the exception of Paccotti et al. (2005) and Stupnicki and Obmiński (1992), saliva and serum correlation studies did not use elite athletes. Furthermore, the exercise tests used within each study had no direct relevance to their own training or sporting background. Evidence highlights that hormones display a high individual variability and specific response to exercise stimulus (Dunn et al., 1981, Crewther et al., 2010, Beaven et al., 2008a) and therefore it could be suggested that the lack of strong correlations between serum and salivary testosterone concentrations may be due to a lack of elite sports specific studies and relevance of the subject’s training background to the test used. Rugby union is an intermittent high-intensity sport played over two 40-minute halves. The physiological demand is primarily anaerobic in nature, although the aerobic system is utilised during rest periods to replenish energy stores (Cunniffe et al., 2009, Duthie et al., 2003). To date, no study between salivary and serum concentration during this type and duration of exercise has been undertaken. Therefore, despite the debate over the validity of using saliva as a medium to analyse endogenous hormone concentrations, the advantages of saliva collection outweigh blood sampling in an elite rugby environment.

2.6 The testosterone to cortisol (T/C) ratio

Another index of the status of protein metabolism is the ratio between free testosterone and cortisol concentrations (T/C ratio). The concentration of testosterone not bound to protein (i.e. the free fraction) is assumed to be a marker of anabolism, whereas the concentration of cortisol is assumed to be a marker of catabolism (Banfi et al., 1993). Therefore, this ratio has been proposed as a marker of the anabolic-catabolic status of the athlete, that is, a global appreciation of protein turnover (Budgett, 1998, Banfi et al., 1993). In sports, this ratio has been used primarily as a potential marker of over-training (Vervoorn et al., 1991, Elloumi et al., 2003, Maso et al., 2004) and to monitor responses throughout intensive training and competition periods (McKenzie, 1999, Passelergue and Lac, 1999, Passelergue and Lac, 2012, Fry et al., 2000, West et al., 2014).

Two approaches to using the T/C ratio in monitoring over-training were originally proposed. Based upon results from long distance runners, over-training was defined either as a T/C ratio of <0.35 x 10-3, calculated on the free testosterone in nanomoles per litre (nmol⋅L-1) and free cortisol in micromoles per litre (µmol⋅L-1), or as a decrease in the T/C ratio of 30% or more in comparison with the previous value (Adlercreutz et al., 1986). However, during longitudinal training and testing in elite male rowers (Vervoorn et al., 1991) and ice speed skaters (Banfi et al., 1993), serum T/C ratios lower than 0.35 x 10-3 were not observed and decreases in the T/C ratio of 30% or more were observed. Vervoorn et al. (1991) also reported individually determined changes in the T/C ratio and suggested that the T/C ratio rather may be indicative of ‘hormonal’ overstrain. Therefore, an 27 increase in this ratio could be interpreted as an indicator of training readiness, or following a regenerative period (e.g. after a high-intensity training period), as a ‘positive consequence of the total training’ (Urhausen et al., 1995).

The T/C ratio has also been identified as a potential marker for social behavioural tendencies (Terburg et al., 2009). Human social structure is a diverse and complex environment where individuals are part of several social hierarchies. Social acts of dominance and submissiveness are the core behaviour in the formation of a social hierarchy consisting of dominant and subordinate individuals (Terburg et al., 2009). Testosterone and cortisol are reported to modulate these behaviours. High testosterone concentrations have been associated with dominant and aggressive behaviours in men (Mazur and Booth, 1998, Archer, 2006). Low concentrations of cortisol have also been associated with aggressive social tendencies (Virkkunen, 1985), whereas high concentrations of cortisol have been linked to depressed mood (Van Honk et al., 2003) and submissive behaviours (Turan, 2015). In social studies, Terburg et al. (2009) suggested that a high testosterone/low cortisol ratio appears to predict motivational and behavioural tendencies where individuals are more likely to confront threat, which could result in aggressive behaviour. They therefore stated that a high T/C ratio could predispose an individual for socially aggressive behaviours. In these situations, cognitive decision-making is void of fear and less influenced by emotional information. Decisions are steered by approaching and reward-seeking tendencies and aggression is reactive and impulsive (Terburg et al., 2009). In elite rugby union, players decision- making, motivational and behavioural tendencies, as indicated by a higher T/C ratio in social studies, are also desirable traits for success in rugby union (Passos et al., 2008, Pope and Wilson, 2012). Furthermore, the environmental setting within a rugby union team is also modulated by theories of social hierarchy (Ponzi et al., 2016) and has been identified as an important aspect of performance and recovery (Cook and Crewther, 2014). Therefore, monitoring the T/C ratio in rugby union may provide further insight to the performance of players in training and competition under stressful conditions.

The correlation of serum and salivary concentrations of the T/C ratio observed in sport are limited and vary from moderate (r = .43, p <.05) (Lippi et al., 2016) to strong (r = .87, p <.001) (Obmiński and Stupnicki, 1997). Furthermore, the calculation of the T/C ratio is based upon the assumption that the free portion is 2% of total serum testosterone (Banfi, 1998), despite reported variances between 1–3% (Fiers et al., 2014). In overtraining studies, the use of the T/C ratio has been queried (Meeusen et al., 2013, Urhausen et al., 1995), however the marker ability of the T/C ratio according

28 to behavioural tendencies (Terburg et al., 2009) may have better application with respect to the monitoring of competition readiness and performance in rugby union players.

2.7 Testosterone, cortisol and athletic performance

Hormonal effects are mediated by steroid receptors through either up-regulating (an increase in receptor content and/or binding sensitivity) or down-regulating (a decrease in receptor content and/or binding sensitivity) in response to exercise and training (Copeland et al., 2002, Eliakim et al., 2009). Exercise can provide a stimulus to elicit a simultaneous increase of antagonistic mediators, producing either anabolic or catabolic actions (Eliakim et al., 2009). Androgen receptor expression has been demonstrated to vary between different muscles, muscle fibres and cells (Yehuda and Meyer, 1984); hence the response to different training methods may also demonstrate similar specificity (Crewther et al., 2011). Glucocorticoid receptors are similarly responsive to exercise and training (Urhausen et al., 1995). Exercise-induced increases in cortisol depend on the duration and intensity of physical exercise (Kuoppasalmi et al., 1980, Kindermann et al., 1982). To achieve a significant increase in the blood cortisol concentration, exercise durations of more than 20 minutes • at an intensity of at least 60% of maximal oxygen uptake (VO2 max) are usually required (Urhausen et al., 1995). In running-based studies (Table 2.3), increases in serum cortisol concentrations were observed in session protocols that included either sprint intervals or steady state running protocols • of a higher intensity; 30 minutes at greater than 65% VO2max (Farrell et al., 1983), 50 minutes at anaerobic threshold (Kindermann et al., 1982) and 45 minutes at 3.2–3.3 min/km (Kuoppasalmi et al., 1980). The results for the testosterone response, however, are mixed and do not demonstrate a clear relationship. Although the mechanisms resulting in the inconsistent response of testosterone to exercise were unclear, the variation might reflect the subject’s individual response to the training stimulus (Crewther et al., 2011).

Analysis of elite rugby union games via heart rate monitoring and blood lactate concentrations indicates the intense nature of game play. Heart rate monitoring can be used as an indicator of average work intensity because of its linear relationship with oxygen uptake at submaximal workloads (McArdle et al., 2010). Deutsch et al. (1998) reported that forwards spent approximately 72% of the match at a heart rate greater than 85% of their maximal competition heart rate whereas the backs spend the majority of match play (38%) in heart rate zones of 75–84% of maximum heart rate. In addition, Dubois et al. (2017) reported that rugby union players spend approximately 40% of the game above 85% of maximum heart rate with no significant difference between backs and • forwards. In elite rugby union, players were reported to exercise at approximately 80–85% VO2max

29 calculated from the player’s individual maximum heart rate (88% HRmax) during the course of the • game and VO2 relationship obtained during a base-line incremental running protocol (Cunniffe et al., 2009). Mean blood lactate concentrations between 5.1–6.7 nmol⋅L-1 taken during stoppages in play have been observed in competitive rugby union games (Deutsch et al., 1998, McLean, 1992). Blood lactate concentration is a measure of intensity from the preceding few minutes of activity (Deutsch et al., 1998). Players achieved between 70–91% of their peak blood lactate concentrations indicating a considerable contribution from anaerobic glycolysis to match performance in players (Duthie et al., 2003). Therefore, despite the intermittent nature of rugby union, heart rate and blood lactate concentration reported during games indicates the high-intensity nature of this sport and would be suggestive to be intense enough to elicit relative increases in hormone concentrations.

30 Table 2.3 Summary of hormone responses to running of different intensities and durations. Reference Subjects Activity Hormone response (Training status)

Kraemer et al. 30 athletes: M&F Sprinting & running Sprint Interval20sec ** (1989) (Recreational) ↑C: 5min PE & 15min PE

Endurance30min ↑C Combination (sprint interval & endurance) ** ↑C: 5minPE & 15minPE

Farrell et al. (1983) 6 HS: M&F Treadmill run • ↑C @ 65% VO2max (r = .18) (Recreational) • * 80% VO2max (r = .63) • * 100% VO2 max (r = .76)

Kindermann et al. 17 HS Treadmill run AE50min ANE156% (1982) (Trained) ↑C (54%)** ↑C (35%)** ↑T (22%)** ↑T (14%)**

Kuoppasalmi et al. 10 TF: 5 SR, Sprinting & running Sprint Interval40m Endurance – MI90min (1980) 5 LD ↑C (27%)** ↑C (NS) (Trained) ↔ T ↑T (7%)

Endurance – HI45min ↑C (43%)** ↑T (21%)

All hormones measured in blood unless otherwise stated. Hormonal observations are only significant if indicated. Key: M&F = male and female, HS = healthy subjects, TF = track and field athletes, SR = sprint runners, LD = long-distance runners, E = endurance, T = testosterone, C = cortisol, PE = post exercise sample time point, • VO2 max = maximal oxygen consumption, AE50min = 50 minutes steady state at anaerobic threshold of 4mmol/L, ANE156% = maximal sprint at 156% of maximum recorded speed, MI90min = 90minutes steady state at a moderate intensity of 4.2min/km, HI45min = 45minutes steady state at a high intensity of 3.2min/km, NS = not stated, % Δ = percentage change, ↑ = increase, ↔ = no difference, *p <.05, **p <.01.

2.7.1 Neuromuscular-based activities: sprinting, jumping, power and strength

Studies of soccer, handball, rugby players and sprinters have demonstrated moderate to strong relationships in which both salivary and serum testosterone and cortisol concentrations are correlated with running speed, jumping power and strength (Table 2.4). In professional male soccer players, moderate relationships between serum testosterone concentration and vertical jumping height and power output were reported (Bosco et al., 1996a, Bosco et al., 1996b). Similarly, Cardinale and Stone (2006) demonstrated moderate relationships between serum testosterone concentration and vertical jump height across a cohort of elite sprinters, soccer and handball players. In this study, serum testosterone concentrations and vertical jump height were higher in 31 sprinters than in soccer players and were higher in soccer players than in handball players. These results support the theory of the influence of testosterone on the development of type II fibres and its role in explosive neuromuscular-based activities. Furthermore, it could also be speculated that a higher baseline testosterone concentration may have contributed to aggressiveness, (Mazur and Booth, 1998, Terburg et al., 2009), competitiveness and drive (Dabbs, 1993, Henry, 1992) in order to provide the motivational effort (Cook et al., 2013) necessary to develop a high rate of muscular force during a maximal explosive jump effort (Crewther et al., 2016).

Kraemer et al. (2004) studied soccer players from one team across a season of training and competition and observed the differences between hormone concentrations and performance parameters of those players who were non-starters and starters. Significant relationships were observed only for the non-starters, whose vertical jump height correlated with both serum testosterone and cortisol, whereas sprint time only correlated with cortisol. Comparisons between the groups indicated that starters had lower levels of circulating testosterone and higher cortisol concentrations and demonstrated significant (p <.05) decrements in sprint speed (+4.3%) and vertical jump height (–13.8%) from baseline values. Despite the lack of significant correlations between hormone concentrations and neuromuscular performance, the observed trends are indicative of fatigue and physical stress from game play. Soccer is a high intensity sport that has a prevalence of maximal repeat sprints (Ekblom, 1986) and the observed increase in cortisol and decrease in testosterone concentrations of the starters are reflective of similar hormonal responses to high intensity exercise programs (Adlercreutz et al., 1986, Kraemer et al., 1989). Similarly, Arruda et al. (2015) observed no relationship between salivary testosterone concentrations and loaded CMJ performance in elite young soccer players across a one year competitive season. Elevated cortisol concentrations can lead to diminished force production due to a reduction in protein synthesis typically stimulated by testosterone interactions (Florini, 1987). Continual high intensity stress experienced through a competitive season results in a catabolic environment for the neuromuscular system negatively affecting neuromuscular performance. Therefore it could be suggested that for running-based team sports, neuromuscular-based performance tests and hormone concentration independently may be a suitable method to monitor the stress of training and competition but may not be a proxy for correlative interpretation.

In weightlifters, strength measures of Olympic lifting performance have been shown to significantly correlate with salivary testosterone concentrations (Crewther and Cook, 2010b). In elite rugby union players, significant correlations between salivary testosterone and cortisol concentrations and speed, power and strength have also been reported (Crewther et al., 2012, Crewther et al., 2009b).

32 In one study, salivary testosterone and the T/C ratio correlated with sprinting time and velocities, and cortisol with mean power output in a loaded squat jump and squat one repetition maximum (1RM) (Crewther et al., 2009b). Intra-group analysis highlighted that when the data was analysed by player positions (forwards vs backs), significant relationships were observed only within the backs and were stronger than the group correlations. In a second study, salivary testosterone concentration correlated strongly with both 10 m sprint time and squat 1RM (Crewther et al., 2012). The data were then grouped according to players’ 1RM in relation to body mass and were classified as either ‘elite’ (1RM >2.0 x body weight) or ‘strong’ (1RM <1.9 x body weight). Strong correlations were observed amongst elite squatters for both the strength and sprint measures, whereas no relationships were observed for the strong squatters. These results highlight that the grouping of athletes of mixed strength abilities or sport requirements may cause significant variance between hormone concentration and neuromuscular performance. Training background and existing strength of an athlete may influence the expression of the neuromuscular structures (i.e. type II muscle fibres, steroid receptors) that can actualise the steroid effects (Crewther et al., 2012, Staron et al., 1994). Therefore it has been suggested that individuals with different strength profiles may exhibit different patterns of bio-utilisation, although hormone availability may be the same (Crewther et al., 2012).

33 Table 2.4 Summary of hormone correlations with athletic performance tests. Reference Subjects Hormone result (correlation) (Training status) * * Bosco et al. (1996a) 32 SP tT & CMJHeight (r = .43) , SpeedAve (r = .47) (Professional) tT & Cooper’s test (r = −.40)* tC & Cooper’s test (r = −.49)**

* Bosco et al. (1996b) 16 SP Δ tT v. CMJAveP (r = .61) ** (Professional) Δ tT & CMJHeight (r = .66)

Cardinale and Stone 48 SP, HB & SR tT: SR > SP** > HB** ** (2006) (Elite) tT & CMJHeight (r = .62)

Kraemer et al. (2004) 25 SP Non-starters (n =14) * (University) T & CMJHeight (r = .77) * * C & CMJHeight (r = −.59 to −.64) , Sprint20yard (r = −.78) * C & Sprint40yard (r = −.57) * T/C & Sprint20yard (r = .56)

Starters (n =11) * T/C & CMJHeight (r = .65)

Crewther et al. 34 RP Group analysis * ** (2009b) (Elite) Sal-T & Sprint10m (r = −.48) , Sprint20m (r = −.56) * Sal-T & Sprintvel (r = .48–.61) * Sal-C & SJMP (r = .41) * Sal-C & BS1RM.kg (r = .39) ** * Sal-T/C & Sprint20m (r = −.43) , Sprintvel (r = .43−.46) * Sal-T/C & SJMP (r = −.39)

Positional analysis – backs (n = 16) * Sal-T & Sprintvel (r = .65) * Sal-C & SJMP (r = .61) * Sal-C & BS1RM.kg (r = .69) * Sal-T/C & SJNorm-MP (r = −.39)

Crewther et al. (2012) 10 RP & TF Elite squatters ** ** (Elite) Sal-T & Squat1RM (r = .92) , Sprint10m (r = −.87) Sal-C & Squat1RM (r = .35), Sprint10m (r = −.23)

Strong squatters Sal-T & Squat1RM (r = .35), Sprint10m (r = −.18) Sal-C & Squat1RM (r = .24), Sprint10m (r = −.24)

Crewther and Cook 4 WL Sal-T & Snatch 1RM (r = .70)*, Clean & Jerk 1RM (r = .62) * (2010b) (Elite)

All hormones measured in blood unless otherwise stated. Hormonal observations are only significant if indicated. Key: SP = soccer players, HB = handball players, SR = sprint runners, RP = rugby players, TF = track and field athletes, CMJ = countermovement jump, tT = total testosterone, tC = total cortisol, T/C = testosterone to cortisol ratio, Sal-T = salivary testosterone, Sal-C = salivary cortisol, Sal-T/C = salivary testosterone to cortisol ratio, SJ = squat jump, BS = back squat, MP = mean power, vel = velocities, AveP = average power, Norm

= normalised data, 1RM = one repetition maximum, 1RM.kg = one repetition maximum normalised for body mass * ** to the power of 0.67, 10m = 10 metres, 20m = 20 metres, 1RM = one repetition maximum, p <.05, p <.01.

34 2.7.2 Resistance training

It has been well documented that resistance training in both athletic and non-athletic populations improves the qualities of the neuromuscular system through morphological (i.e. muscle size) and functional (i.e. strength and power) changes (Kraemer et al., 2002, Kraemer et al., 1998). The prescription of resistance-exercise workouts is based on the principle that acute elevations in endogenous hormones (Beaven et al., 2008a) increase the likelihood of receptor interactions, offering both short-term benefits for training performance and long-term adaptations (Ahtiainen et al., 2004). The long-term changes in muscle growth and subsequent performance resulting from resistance training occur partially through the biological effects of testosterone and cortisol (Kraemer and Ratamess, 2005, Crewther et al., 2006). Equally, the suppression of both endogenous total and free testosterone decreases gains in muscle size and strength traditionally observed through strength training (Kvorning et al., 2006).

Previously, hypertrophy protocols in resistance training (i.e. high total work, moderate loads and short rest periods) were postulated to typically produce relatively acute and large increases in anabolic (i.e. testosterone, growth hormone and growth factors) and catabolic (i.e. cortisol) hormones that promote protein metabolism and muscle growth during the recovery period (Crewther et al., 2008, Smilios et al., 2003). Conversely, training sessions designed to enhance maximal strength and power (i.e. lower total work, high loads and longer rest periods) was typically thought to produce little or no hormonal change (Häkkinen and Pakarinen, 1993, Kraemer et al., 1991). However, a more recent study of rugby union players (who undertook resistance workouts using several different protocols) noted that the athlete’s acute hormonal response varied depending on the workout design and that generally no two athletes followed the same response pattern (Beaven et al., 2008b). In a subsequent study, these athletes trained according to which workout design would produce maximal and minimal changes in salivary free testosterone concentrations (Beaven et al., 2008a). Using the workout protocol for maximal testosterone change (irrespective of its design), athletes achieved significant increases in both strength and body mass over a 3 week period. Conversely, training with the workout protocol for minimal testosterone change had either no effect upon or resulted in a reduction in strength in most players.

These results suggest that testosterone can be used as a marker for selecting specific resistance training protocols to maximise functional gain (Beaven et al., 2008a, Hansen et al., 2001, Mangine et al., 2017, Rønnestad et al., 2011, Ahtiainen et al., 2003). Furthermore, understanding the effects of testosterone on maximal voluntary strength and skeletal muscle hypertrophy has important implications for the periodisation and application of resistance training workouts.

35 2.7.3 Testosterone and cortisol responses to concurrent training methodologies

Athletes often perform strength and endurance training concurrently in an attempt to achieve adaptations specific to their sport’s physiological requirements (Bell et al., 2000). Concurrent training alters the response of anabolic to catabolic hormones according to the order and type of exercises performed (Table 2.5). Aerobic training has been suggested to negatively affect the strength development that would normally occur from strength training alone (Bell et al., 2000, Häkkinen et al., 2003). Furthermore, the simultaneous performance of both aerobic and strength- training activities is thought to generate a higher metabolic demand, resulting in a greater increase in cortisol concentrations, which in in turn could negatively influence the release of testosterone during the training session (Cumming et al., 1983, Brownlee et al., 2005). Rosa et al. (2015), Cadore et al. (2008) and Taipale and Häkkinen (2013) reported that performing strength training after endurance exercise resulted in significant increases in serum testosterone when compared with the alternate training order. The serum cortisol response however, varied. Rosa et al. (2015) and Taipale and Häkkinen (2013) observed significant increases in both exercise groups, whereas Cadore et al. (2008) did not. Schumann et al. (2013) did not observe any significant change in both serum cortisol and testosterone, regardless of the intra-session exercise order. Variations in the strength training protocols have been suggested as one reason to account for the observed difference (Table 2.6). The strength training protocols of Rosa et al. (2015) and Cadore et al. (2008) incorporated a session structure of moderate intensity, multiple sets and short rest intervals. This is purported to stimulate muscular hypertrophy and higher acute levels of serum testosterone (Kraemer and Ratamess, 2005), in contrast with power and submaximal strength training methodologies (Linnamo et al., 2005) as used in the concurrent training studies of Taipale and Häkkinen (2013) and Schumann et al. (2013). The sport and training background of the subjects may also contribute to the variations observed. For example, greater acute elevations of serum testosterone following a strength training session are expected in strength trained subjects than in endurance trained subjects (Tremblay et al., 2004). In recreational athletes, acute changes in hormone concentration are not related to resistance-training strength gains (Morton et al., 2016) whereas different hormonal responses have also been observed between elite and sub-elite athletes (Crewther et al., 2012).

36 Table 2.5 Summary of acute hormone responses to concurrent training. Reference Subjects Activity Hormone response (Training status) Cadore et al. (2008) 10 HS Cycling (E) ES SE (Trained) Strength (S) ↑T (41%)* ↑T (3%) ↑C (13%) ↑C (8%)

Schumann et al. 42 HS Cycling (E) ES SE (2013) Strength (S) ↔ T, C ↔ T, C

Rosa et al. (2015) 14 HS Running (E) ES SE (Trained) Strength (S) ↑T (58%)* ↑T (15%) ↑C (170%)* ↑C (92%)*

Taipale and 12 HS, LD Running (E) ES SE Häkkinen (2013) (Trained) Strength (S) ↑T** ↔ T ↔ C ↑C** ↑% ΔC* ↑ % ΔC*

All hormones measured in blood unless otherwise stated. Hormonal observations are only significant if indicated. Key: HS = healthy subjects, TF = track and field athletes, SR = sprint runners, LD = long-distance runners, • E = endurance, S = strength, T = testosterone, C = cortisol, VO2 max = maximal oxygen consumption, ES = endurance training followed by strength, SE = strength training followed by endurance, % Δ = percentage change, ↑ = increase, ↔ = no difference, *p <.05, **p <.01.

Table 2.6 Summary of endurance and strength training protocols performed during concurrent sessions. Reference Endurance Strength Exercises Set x Rep % 1RM Rest Tempo Cadore et al. (2008) 30min cycling SQ, BP, 3x8 75% 90sec Controlled

@ 75%HRmax CP, KE

Schumann et al. 30min cycling LP 3x10 40% 3min Explosive (2013) @ 65%Wattsmax 4x3 75−90%

Rosa et al. (2015) 32min treadmillInt SQ, BP, 3x10 70% 60sec Controlled CP, FCP

Taipale and 60min runningSS LPmax, expl, 3x8−10 30−40% 2min Explosive Häkkinen (2013) SJexpl, CRmax 3x5−8 70−85%

Key: HRmax = maximal heart rate, Wattsmax = maximal watts, Int= intermittent intensities set at a ratio of 2minutes at a speed corresponding to blood lactate concentration of 2 mmol/L and 1 minute at a blood lactate concentration of 4 mmol/L, SS= steady state running set at intensity between lactate threshold and respiratory compensation threshold, SQ = squat, BP = bench press, CP = cable pulldown, FCP = front cable pulldown, KE = knee extension, LP = leg press, % 1RM = percentage of one repetition maximum, min = minutes, sec = seconds.

37 2.8 Competition: Acute response of testosterone and cortisol concentrations

Increased levels of endogenous testosterone have been shown to be an important marker of competitive athletic performance and sporting success, as well as correlating positively with desirable competitive traits such as positive mood states (e.g. competitiveness, drive and persistence) (Dabbs, 1993, Henry, 1992) and aggressive and dominant behaviours (Bouissou, 1983, Dixson, 1980, Salvador et al., 1999). Testosterone concentrations generally respond to competition in two ways: testosterone rises in anticipation of the competition and the elevated concentration in testosterone post-competition is greater in event winners than in losers (Booth et al., 1989, Elias, 1981, Mazur and Lamb, 1980, Suay et al., 1999). The anticipatory rise in pre-competition testosterone is thought to be associated with an individual’s willingness to take risks and may also improve coordination, cognitive performance and concentration. A rise in testosterone concentration following a win may be associated with the individual’s elated mood (Mazur and Booth, 1998). Mazur (1985) also proposed that rising or elevated levels of testosterone facilitate attempts to achieve or maintain high status (or dominance rank) amongst fellow competitors, further highlighting the interactive role of heightened testosterone with increased competitiveness and dominance behaviour.

In wrestlers, the absolute (Fry et al., 2011) and relative change (Elias, 1981) in the pre to post competition serum testosterone concentrations have shown to be significantly greater amongst winners than in losers. In contrast, some sports have reported no difference in hormone concentration between winners and losers (Gonzalez-Bono et al., 1999, Booth et al., 1989). In , player’s mood and positive evaluation of their own performance displayed a moderate relationship to the change in testosterone concentration amongst winners (Booth et al., 1989). In , although there was no difference between winners and losers, the salivary testosterone response was positively correlated with the ‘score to time playing’ ratio, which was an indicator of individual participation in and perception of the outcome of that game (Gonzalez-Bono et al., 1999). The lack of a significant relationship between testosterone concentration and game outcome highlights the complex relationship between mood, behaviour and hormones in each individual athlete (Gonzalez-Bono et al., 1999).

Other studies have also included the monitoring of physical, psychological and behavioural responses that contribute to performance outcome. For example, in judo contests, significant correlations between serum testosterone and offensive behaviours (e.g. threat, fighting and attack) have been observed (Salvador et al., 1999, Obmiński, 2009). In the study by Obmiński (2009), one athlete demonstrated a pre-fight serum testosterone concentration approximately 3.6 times less than 38 his pre-study base concentration recorded 4 days prior. This athlete also demonstrated a depressed mood and passive behaviours during the fight, which he unexpectedly lost. A corresponding increase in serum cortisol concentration was also reported. Neurobiological studies have suggested that cortisol may antagonise the interplay between testosterone and behaviour, thus suggesting a dual-hormone relationship between testosterone and cortisol in relation to observed variations in behaviour (Mehta and Josephs, 2010). Cortisol has been shown to suppress the activity of the HPG axis, down-regulate androgen receptors and inhibit the action of testosterone on target tissues (Burnstein et al., 1995, Chen et al., 1997). In male clinical populations, aggression was positively related to serum testosterone concentration only in individuals with a corresponding low cortisol concentration (Dabbs et al., 1991, Popma et al., 2007), whereas high concentrations of cortisol have been linked to depressed mood (Van Honk et al., 2003) and submissive behaviours (Turan, 2015). A summary of the published research investigating the relationship between cortisol and testosterone concentration and the corresponding results on sporting performance are presented in Table 2.7.

Cortisol has been documented as a marker of competition performance and is characterised by an anticipatory rise in cortisol concentrations (Elias, 1981, Salvador et al., 1987, Fry et al., 2011). In judo (Suay et al., 1999) and wrestling (Elias, 1981) contests, pre to post match salivary cortisol concentrations increased in both winners and losers but were significantly higher in winners (Salvador et al., 1987, Suay et al., 1999). In tennis match play, the salivary cortisol concentrations did not relate to winning or losing but was related to the player’s seeding (top players having lower cortisol) (Booth et al., 1989). Perception and the level of experience of the athletes could be modulating these relationships (Suay et al., 1999): higher ranked players who won may have been more relaxed or thought that they would have easier matches against lower ranked opposition. Therefore, the hormonal response to competition may not be a direct consequence of winning or losing but rather is mediated by complex psychological processes.

39 Table 2.7 Changes in hormonal concentrations of winners and losers in competitive sporting events. Subjects Sample timing Reference Hormonal response (Training status) Pre Post Booth et al. (1989) 6 TP 15 min 0 min Δ Sal-T: ↑W & ↓L (University) Δ Sal-C: ↑W & ↓L Δ Sal-C & Player seed*

Elias (1981) 15 WR 10 min 10, 35 min Δ T: ↑W & ↑L (University) Δ C: ↑W > ↑L1 ** % Δ T: ↑W >↑L1 **

Gonzalez-Bono et al. 16 BB 45 min 15 min Δ Sal-T: ↑W & ↓L (1999) (Elite) Δ Sal-C: ↑W & ↑L Δ Sal-T & STP ratio (r = .56) * ** Sal-TPRE & STP ratio (r = −.66)

Salvador et al. (1987) 14 J 10 min 45 min % Δ T: ↓W & ↓ L (Regional) % Δ C: ↑W & ↓L

Suay et al. (1999) 26 J 30 min 30 min Δ T: ↑W & ↑L (Experienced) Δ C: ↑W > ↑L* Δ T & MTW score (r = .42)*

Fry et al. (2011) 12 WR n/a n/a Δ T: ↑W > ↑L* (University) Δ C: ↑W & ↑L Δ T/C: ↔ W & L

Salvador et al. (1999) 28 J 10 min 10min No W & L stated *, ** (Trained) TPRE & BO (r = .40–.54) % Δ C & BO (r = .39–.40)*

Obmiński (2009) 7 J Morning Nil No W & L stated – BO (Well trained)

Crewther et al. (2013) 13 RL Strength session (72 h pre game) (Elite) 10min 5min Δ T: ↑W** & ↔ L % Δ T: ↑W*** & ↔ L

All hormones measured in blood unless otherwise stated. Hormonal observations are only significant if indicated. Key: TP = tennis players, WR = wrestlers, BB = basketballers, RL = rugby league players, STP = score/time playing ratio, J = judoists, n/a = not available, MTW = motivation to win, W = winners, L = losers, Δ = change, % Δ = percentage change, Sal-T = salivary testosterone, Sal-C = salivary cortisol, T = testosterone, C = cortisol, T/C = testosterone to cortisol ratio, Pre = pre training concentrations, BO = behavioural observations, 1 = 10 min post, ↑ = increase, ↓ = decrease, ↔ = no difference, * p <.05, ** p <.01, *** p <.001, significant from losses.

The measurements of mid-week hormone concentrations in response to training have also been suggested as a marker for competition success relative to subsequent weekend competition (Crewther et al., 2017, Crewther et al., 2013). In a rugby league team, the salivary testosterone response to mid-week resistance training workouts has been shown to be significantly associated 40 with subsequent game outcome 3 to 4 days later (Crewther et al., 2013). The salivary testosterone responses (absolute and relative) were elevated prior to games won, whereas there was no observed change in hormone concentrations following resistance-training workouts prior to a loss. Crewther et al. (2013) suggested that a transient increase in testosterone concentration might have a delayed effect on those adaptive systems that directly influence game-day performance, or that the increase might reflect an athletes’ recovery state from previous games. In an international rugby union team, a mid-week stress test involving seven minutes of shuttle runs (2 × 20 m) dispersed across one minute stages with increasing speed was completed 3 to 4 days prior to a match (Crewther et al., 2017). An increase in morning (8 am) to pre-shuttle run test (>7 h later) salivary cortisol concentration was observed prior to winning, whereas cortisol decreased before losing. Salivary testosterone concentrations decreased significantly across the same period, irrespective of the match outcome. Crewther et al. (2017) suggested that the observed mid-week rise in cortisol before winning might have indicated a stronger capacity to prepare for and cope with competition stress 3 to 4 days later. Apart from the differences in the physical stimulus and sample timing, team culture, management strategies and game demands between the club and international environment may explain the observed differences between these two studies. Nevertheless, these studies highlight that monitoring mid-week hormone responses to selected training sessions may help to predict team readiness for upcoming competitions.

2.9 Summary and conclusions

Testosterone is the primary circulating androgen in human males, whereas cortisol is the main glucocorticoid. The physiological roles of testosterone can be divided into two major categories: (a) those related to reproductive function and the development of male secondary sex characteristics (i.e. androgenic effects); and (b) those that pertain more generally to stimulation of tissue growth and development (i.e. anabolic effects) (Hackney, 1996). The primary function of cortisol is to regulate the stress response through psychological and physiological allostatic mechanisms such as immune functioning, cardiovascular activation, glucose metabolism and emotional processing (Schulkin et al., 1994, Sapolsky et al., 2000).

Hormonal effects are mediated by steroid receptors through either up-regulating (an increase in receptor content and/or binding sensitivity) or down-regulating (a decrease in receptor content and/or binding sensitivity) in response to exercise and training (Copeland et al., 2002, Eliakim et al., 2009). Exercise can provide a stimulus to elicit a simultaneous increase of antagonistic mediators, producing either anabolic or catabolic actions (Eliakim et al., 2009). There is a significant body of evidence documenting that steroid hormones have both short-term and long-term effects in humans 41 and on human performance. Furthermore, the neuromuscular system is an important endocrine target for testosterone and cortisol. Thus, the acute effect of testosterone and cortisol contributes to improved neuromuscular performance during training. These short-term effects assist in regulating long-term changes in skeletal muscle proteins and skeletal muscle growth.

Increased testosterone concentrations play a role in developing a high rate of force in muscular activity within a short contraction time (Cardinale and Stone, 2006) and have been related to the neuromuscular expression of speed, power and strength (Crewther et al., 2009b, Cardinale and Stone, 2006). Mood states such as competitiveness, drive, persistence and contribution to winning have also been associated with increases in testosterone concentrations (Dabbs, 1993, Henry, 1992), along with aggressive and dominant behaviours (Bouissou, 1983, Dixson, 1980, Salvador et al., 1999). Cortisol is known to have a protein-catabolic effect on skeletal muscle (Urhausen et al., 1995). During prolonged exercise, an increase in cortisol is important for energy supply (Galbo, 1986, Holloszy and Coyle, 1984) whereas post-exercise, cortisol is hypothesised to assist with muscular glycogen resynthesis (Urhausen et al., 1995).

Studies comparing salivary and serum hormone concentrations at rest have reported significant and strong correlations for both testosterone and cortisol when measured in clinical populations across single and multiple time points throughout the day. Cortisol demonstrates very strong relationships between serum and salivary concentrations (Vining et al., 1983a, Umeda et al., 1981, Peters et al., 1982, Gallagher et al., 2006, Lippi et al., 2009) whereas testosterone, despite rapid daily fluctuations in an individual’s salivary concentrations, displays a moderate to strong correlation between salivary and serum concentrations (Fiers et al., 2014, Goncharov et al., 2006, Johnson et al., 1987, Baxendale et al., 1980, Sannikka et al., 1983, Wang et al., 1981). The correlation between serum and salivary concentrations collected after exercise have demonstrated to be not as consistent as that observed during rest. The lack of a strong correlation between serum and salivary testosterone concentrations following exercise are suggested to be a result of high individual variability and the specific response to exercise stimulus (Dunn et al., 1981, Crewther et al., 2010, Beaven et al., 2008a). However stronger correlations between saliva and serum testosterone responses have been observed during higher intensity exercise protocols (Lane and Hackney, 2015). These observations highlight the importance of understanding the athlete, the type and duration of exercise undertaken in relation to any observed measurements. From a practical viewpoint, athletes are more comfortable to provide a saliva sample as opposed to a periodic blood sample when used for long-term monitoring (Lippi et al., 2016). Therefore, despite the continued debate over the

42 validity of using saliva as a medium to analyse endogenous hormone concentrations, the advantages of saliva collection outweigh blood sampling in an elite sporting environment.

Increased testosterone and cortisol concentrations have been demonstrated to be an important marker of competitive athletic performance and predictor of sporting success. The anticipatory rise in pre-competition testosterone is considered to be associated with an individual’s willingness to take risks and may also improve coordination, cognitive performance and concentration. Testosterone concentrations generally respond to competition in two ways: testosterone rises in anticipation of the competition, and the elevated concentration in testosterone post-competition is greater in event winners than in losers (Booth et al., 1989, Elias, 1981, Mazur and Lamb, 1980, Suay et al., 1999). Athlete’s mood, self-evaluation and behaviours associated with competition performance have also correlated with testosterone concentrations (Gonzalez-Bono et al., 1999, Booth et al., 1989). It has been suggested that cortisol may antagonise the interplay between testosterone and behaviour, thus suggesting a dual-hormone relationship between testosterone and cortisol in relation to observed variations in behaviour (Mehta and Josephs, 2010). Cortisol has also been documented as a marker of competition performance and is characterised by an anticipatory rise in cortisol concentrations (Elias, 1981, Salvador et al., 1987, Fry et al., 2011). In combat sporting contests (Suay et al., 1999, Elias, 1981), prematch to postmatch salivary cortisol concentrations increased in both winners and losers but were significantly higher in winners. Although the cortisol concentration may be associated with successful competition outcome, it may also reflect the physiological demand and effort from the athlete. Hence these findings suggest that the hormonal response to competition may not be a direct consequence of winning or losing but rather is mediated by complex psychological and physiological processes. The measurements of mid-week hormone concentrations in response to training have also been suggested as a marker for competition success relative to subsequent weekend competition (Crewther et al., 2017, Crewther et al., 2013). Despite the differences in the mid-week training session used and consequent hormonal response, these studies highlight that monitoring mid-week hormone responses to selected training sessions may help to predict team readiness for upcoming competitions.

Acute hormonal levels may influence the neuromuscular system, which in turn influences long-term muscular changes as a result of strength training. Traditionally hypertrophy protocols have been postulated to produce relatively acute and large increases in anabolic (i.e. testosterone, growth hormone and growth factors) and catabolic (i.e. cortisol) hormones that promote protein metabolism and muscle growth during the recovery period (Crewther et al., 2008, Smilios et al., 2003). However more recent work has demonstrated that the optimal hormonal response is individual and

43 depends on the workout design. As such training according to which workout design produces maximal changes in salivary free testosterone concentrations (Beaven et al., 2008a), is more likely to achieve significantly greater increases in both strength and body mass. Hence these results suggest that testosterone concentrations can be used as a marker for determining individualised specific resistance training protocols, which will maximise function gains.

In summary, the published literature supports the use of testosterone and cortisol concentrations as reliable markers of athletic performance, training adaptation and as a predictor of competition readiness. The results observed amongst the studies of athletes also highlight the individual responses of testosterone to different training and competition stimuli. These observations may result from athletes’ genetic disposition or from the study design. Alternatively, they may be associated with the athletes’ training practices, which are specific to the physiological demands of their sport (reflected by testosterone and cortisol concentrations). Within the sporting environment, this highlights the importance of understanding the relationship between these factors and the opportunity to optimise athlete preparation (both physically and psychologically) for competition.

44 3. Relationship Between Pregame Concentrations of Free Testosterone and Outcome in Rugby Union

Citation: Gaviglio, C. M., Crewther, B. T., Kilduff, L. P., Stokes, K. A., & Cook, C. J. (2014). Relationship between pregame concentrations of free testosterone and outcome in rugby union. International Journal of Sports Physiology and Performance, 9(2), 324−331.

3.1 Introduction

Rugby union is a sport characterised by a high degree of aggressive impacts and interactions (e.g. grappling for possession of the ball) with forwards generally involved in more body-contact situations than the backs (Roberts et al., 2008). In general the forwards are required to gain and retain possession of the ball. Conversely, the backs spend greater time in free running whilst controlling the possession of the ball obtained by the forwards (Duthie et al., 2003).

Testosterone and cortisol has been studied in relation to training and performance outcomes in sport (Beaven et al., 2008a, Crewther et al., 2009a, Passelergue and Lac, 1999). For example, testosterone concentrations are related to the neuromuscular expression of speed, power and strength (Crewther et al., 2009b, Cardinale and Stone, 2006). Mood states such as competitiveness, drive, persistence and contribution to winning are also associated with higher testosterone levels, (Henry, 1992) along with aggressive and dominant behaviours (Salvador et al., 1999). Given these associations, the measurement of testosterone before training or competition could help to predict a favourable performance outcome in those sports, which rely on these physiological and behavioural actions.

The reporting of pre-event testosterone concentrations as a predictor of successful outcomes in sport is equivocal. A study of elite fencers suggested that higher blood testosterone concentrations increased their chance for success (Obmiński et al., 1998) and a similar trend was noted in judoists, (Obmiński, 2009) although there was no clear distinction between hormone concentration and outcome. In separate studies on judoists (Suay et al., 1999) and tennis players, (Booth et al., 1989) pregame testosterone concentrations also showed no relevance to either wins or losses.

Links between the competition responses of both free and serum (total) testosterone and the sporting outcomes are also equivocal (Booth et al., 1989, Elias, 1981, Gladue et al., 1989, Gonzalez-Bono et al., 1999). In wrestlers, the pre to post competition changes in serum testosterone and cortisol concentrations were indicators for success in competition (Elias, 1981) and a similar

45 trend for only serum testosterone was seen in winners in the same sport (Fry et al., 2011). In tennis, the difference between winners and losers was also seen in the rise in mean testosterone from pre to immediately post match (Booth et al., 1989). However, in basketball (Gonzalez-Bono et al., 1999) no significant difference between the free testosterone and cortisol responses and game outcomes were observed, but the testosterone response did show a significant relationship between the ‘score/time playing’ ratio, an indicator of individual participation and perception in the outcome of that game.

The testosterone to cortisol (T/C) ratio has been examined as putative marker of the anabolic to catabolic hormonal balance within the body. This ratio was originally hypothesised as an indicator of overtraining and recovery from intensive physical activity (Banfi, 1998) although more recent investigations have queried its validity (Urhausen and Kindermann, 2002, Meeusen et al., 2013). Few studies have investigated the relationship between the T/C ratio and the performance outcomes in sport. As a predictive marker in golfers, (Doan et al., 2007) a low T/C ratio was related to lower scores (i.e. better performance). In elite rugby union players, a postmatch decline in the T/C ratio was reported after 2 matches in conjunction with a pregame increase across a 3 game period, although game outcomes were not reported (Cunniffe et al., 2011).

Previous research has often included only 1 or 2 competitive events (and often weeks apart). Thus, the purpose of this study was to improve on this by examining the game-day measures of salivary free testosterone and/or cortisol concentrations and the related outcomes of professional rugby union across 6 consecutive games. We hypothesised that salivary free-testosterone concentrations and the T/C ratio would be higher prior to a win (vs a loss) and that both measures would correlate to a game ranked performance score.

3.2 Methods

3.2.1 Subjects Twenty-two male rugby union players from a professional rugby union team were recruited for this study. The age, height and body mass details for participants are presented as a team and for the 2 major positional groupings (forwards and backs) in Table 3.1. Subjects were informed of the study protocols and signed informed consent before testing began, and ethics approval was obtained from a University Ethics committee.

46 Table 3.1 Physical Characteristics of the Rugby Union Forwards and Backs, Mean (SD) Team (N = 22) Backs (n = 11) Forwards (n = 11) Age (years) 27.8 (4.0) 27.1 (3.5) 28.4 (4.4) Height (cm) 186.9 (7.7) 184.1 (6.6)^ 189.1 (8.0) Body mass (kg) 103.4 (11.6) 93.8 (8.0)** 111.0 (7.8) Significantly different from forwards ^P <.1, **P <.01.

3.2.2 Design

This study was conducted mid-season over the course of 6 weeks. Players continued to train midweek according to their usual training schedules and the games were played during the weekend on a home-and-away basis. On each of the days immediately prior to testing, players were encouraged to sleep well (>7 h), consume a good breakfast (each player standardised across testing sessions) and maintain their fluid intake (at least 750 mL) during the 2 hours prior to each testing session. During the course of this study, the subjects were not on medications other than several for asthma control.

3.2.3 Hormone Assessment

Saliva samples were taken at times chosen to minimise any interruption to the players’ game preparation and to keep a diurnal consistency in sampling. Saliva was used due to its ease of compliance, low invasiveness and ability to track the biologically active “free” hormone (Crewther and Cook, 2010a).

Pregame samples were taken 40 minutes prior to the start of the game. Postgame samples were taken 15 minutes after the player was either substituted off the field or after the game’s completion. For each test, a 2 mL saliva sample was collected and stored at –20 °C before analysis. Saliva was assayed for testosterone and cortisol concentrations by a private commercial laboratory (HFL Sport Science Laboratories, UK) using commercially available kits (Salimetrics Enzyme Immunoassay Kits, Salimetrics, USA) according to the manufacturer’s instructions. Inter-assay coefficients of variation (based on low and high control samples) for testosterone and cortisol were both <10%.

3.2.4 Game Day Performance

Game-day performance was determined by the overall outcome, being a win, loss or draw based on the final score. A game-ranked performance score (1−6, 6 being the best score possible) was also assigned to each game (Table 3.2).

47 The win percentage was calculated from historical data (obtained via http://www.premiershiprugby.com) gathered from all of the games played against that relevant opposition (Table 3.4).

Table 3.2 Results and Classifications for Each of the Rugby Games Played Score Outcome Classification Clarification notes on appointing Classification 1 Loss Bad Due to amount of points lost by and poor skill decision-making and execution during the game. 2 Loss Unlucky A game that could have been won. A result of a close score or a game that was lost in the last few minutes. 3 Draw Draw 4 Win Average A game resulting in a close score or a game that was won in the last few minutes. A perceived lack of dominance during the game. 5 Win Good A win where the team played well against the opposition. A good performance not necessarily reflected by the score line. 6 Win Good + A dominant win. Reflected usually by the quality of the opponent (good) and the score-line. Note: Score = game ranked performance score; outcome = game result.

3.2.5 Statistical Analyses

Pregame and postgame hormone concentrations and the hormonal changes across each game (perigame) were tabulated for each player, game and result. The perigame hormone value was calculated by dividing the postgame hormone value by the pregame hormone value (postgame ÷ pregame = perigame). The hormone results for winning and losing were compared using paired t tests. Comparisons were conducted on the team data (team) and for the 2 different positional groups within the team (backs and forwards). To determine if game venue (home or away) affected our results, the hormone data was pooled according to game venue and compared using paired t tests. The bivariate relationships between the hormonal variables and the game-ranked outcomes were assessed using Pearson product-moment correlations. The level of significance was set at P ≤.05.

48 3.3 Results

The pregame, postgame and perigame measures of testosterone, cortisol and T/C ratio are presented in Table 3.3.

Table 3.3 Salivary Hormone Concentrations of Elite Rugby Union Players on Game Day, Mean (SD) Measure Game 1 Game 2 Game 3 Game 4 Game 5 Game 6 (Loss) (Loss) (Win) (Loss) (Win) (Win) Pregame testosterone (pg⋅mL-1) 78 (22) 99 (31) 121 (44) 94 (33) 111 (41) 114 (32) Pregame cortisol (µg⋅dL-1) 0.3 (0.2) 0.3 (0.1) 0.4 (0.7) 0.2 (0.1) 0.2 (0.1) 0.3 (0.1) Pregame T/C ratio 435 (177) 533 (334) 553 (338) 506 (324) 721 (377) 494 (232) Postgame testosterone (pg⋅mL-1) 133 (52) 154 (58) 143 (86) 159 (81) 143 (68) 198 (94) Postgame cortisol (µg⋅dL-1) 0.9 (0.5) 0.7 (0.4) 0.6 (0.5) 0.6 (0.2) 0.6 (0.3) 1.0 (0.5) Postgame T/C ratio 264 (105) 270 (148) 363 (256) 237 (137) 237 (64) 233 (108) Perigame testosterone 1.8 (0.7) 1.7 (0.8) 1.4 (1.2) 1.7 (0.9) 1.5 (0.9) 1.8 (0.9) Perigame cortisol 3.1 (1.7) 3.9 (2.8) 2.9 (3.2) 5.3 (7.5) 5.3 (5.6) 4.0 (3.1) Perigame T/C ratio 0.6 (0.6) 0.6 (0.3) 0.9 (0.9) 0.7 (0.3) 0.4 (0.2) 0.5 (0.2) Abbreviations: T/C = testosterone-to-cortisol. Perigame = postgame ÷ pregame hormone concentration.

Table 3.4 outlines the corresponding game information and results. Table 3.4 also shows the subjective game ratings, game-ranked score and other relevant information such as game venue and historical win percentage against the corresponding opposition. Historical win percentage had no obvious effect on pregame hormone levels. Game venues (home vs away) only had an effect on postgame testosterone (P <.05) and postgame cortisol (P <.001) concentrations, both being higher in away games (Table 3.5).

Table 3.4 Game-Day Information for Elite Rugby Union Players Game Outcome Rating Score Game Venue Historical Win % 1 Loss Bad 1 Home 70% 2 Loss Unlucky 2 Away 37% 3 Win Average 4 Home 50% 4 Loss Bad 1 Away 50% 5 Win Good+ 6 Home 48% 6 Win Good 5 Away 63% Note: Outcome = game result; rating = subjective performance rating; score = game-ranked performance score.

49 Table 3.5 Comparison of Hormone Concentrations According to Game Venue (Home and Away), Mean (SD) Measure Home Away Pregame testosterone (pg⋅mL-1) 104 (40) 103 (32) Pregame cortisol (µg⋅dL-1) 0.3 (0.4) 0.3 (0.1) Pregame T/C ratio 571 (326) 510 (292) Postgame testosterone (pg⋅mL-1) 140 (68) 171 (80)* Postgame cortisol (µg⋅dL-1) 0.6 (0.3) 0.8 (0.5)** Postgame T/C ratio 285 (164) 246 (131) Perigame testosterone 1.6 (0.9) 1.8 (0.9) Perigame cortisol 3.8 (3.9) 4.4 (4.8) Perigame T/C ratio 0.7 (0.6) 0.6 (0.4) Abbreviations: T/C = testosterone to cortisol. Note: Perigame = postgame ÷ pregame hormone concentration. Significantly different from home: *P <.05, **P <.01.

As a team, pregame testosterone concentrations were significantly higher before a win (P <.001) than a loss (Table 3.6). For the backs, pregame testosterone (P <.001), T/C (P <.05) and postgame testosterone (P <.05) concentrations were significantly greater before a win than a loss (Table 3.6). We found no significant differences between the forwards’ hormonal concentrations (pregame, postgame and perigame) across the different match outcomes.

We found significant correlations between pregame testosterone concentration and game-ranked outcome for both the team (r = .81, P <.05) and backs (r = .91, P <.05; Figure 3.1). The pregame T/C for the backs was the only other variable that correlated with game-ranked outcomes (r = .81, P = .05; Figure 3.2).

Table 3.6 Comparison of Salivary Hormones According to Outcome (Win or Loss) for All Games, Mean (SD)

Team (N=22) Backs (n=11) Forwards (n=11) Measure Win Loss Win Loss Win Loss ** ** Pregame testosterone (pg⋅mL-1) 115 (39) 91 (28) 127 (43) 86 (22) 112 (33) 97 (32) Pregame cortisol (µg⋅dL-1) 0.3 (0.3) 0.2 (0.1) 0.5 (0.5) 0.3 (0.1) 0.2 (0.1) 0.2 (0.2) Pregame T/C ratio 589 (316)^ 491 (278) 552 (367)* 390 (147) 614 (271) 575 (336) * Postgame testosterone (pg⋅mL-1) 161 (83) 149 (63) 164 (60) 134 (46) 160 (99) 160 (74) Postgame cortisol (µg⋅dL-1) 0.7 (0.4) 0.7 (0.4) 0.7 (0.3) 0.6 (0.3) 0.7 (0.5) 0.8 (0.4) Postgame T/C ratio 277 (143) 257 (130) 258 (135) 273 (153) 294 (146) 244 (107) Perigame testosterone 1.8 (1.0) 1.7 (0.8) 1.4 (0.6) 1.6 (0.6) 1.7 (1.3) 1.8 (0.9) Perigame cortisol 4.1 (3.9) 4.1 (4.0) 3.9 (3.5) 2.8 (1.6) 4.2 (3.9) 5.3 (5.0) Perigame T/C ratio 0.6 (0.5) 0.6 (0.4) 0.7 (0.6) 0.7 (0.4) 0.6 (0.3) 0.5 (0.4) Abbreviations: T/C = testosterone-to- cortisol. Note: Perigame = postgame ÷ pregame hormone concentration. Significantly different from loss: ^ P <.1,* P <.05, ** P <.01.

50

Figure 3.1 Relationship between pregame (Pre) salivary testosterone and game-ranked outcome. Game-ranked outcome: 1 = bad loss, 2 = unlucky loss, 4= average win, 5 = good win and 6 = good win +. Regression lines presented for the team, backs and forwards. Error bars represent standard error.

Figure 3.2 Relationship between the pregame (Pre) salivary testosterone-to-cortisol (T/C) ratio and game- ranked outcome. Game ranked outcome: 1 = bad loss, 2 = unlucky loss, 4= average win, 5 = good win and 6 = good win +. Regression Lines presented for the team, backs and forwards. Regression Lines presented for the Team (solid), Backs (dotted) and Forwards (dash). Error bars represent standard error.

51 3.4 Discussion

Our results suggest that differences seen in the pregame testosterone discriminates strongly between game outcomes. This held across a more subjective, detailed classification of game outcome.

Pregame testosterone as an indicator of team outcome in sport has not, to our knowledge, been presented previously. There is considerable literature from individual-based sport (Booth et al., 1989, Elias, 1981, Gladue et al., 1989, Gonzalez-Bono et al., 1999, Obmiński, 2009, Salvador et al., 1987) linking the outcomes of games and combat (wrestling and judo) bouts to testosterone concentrations generally across a single day of bouts, or a single game, but pregame testosterone is not well presented as a predictive factor. “Snapshot” studies are extremely difficult to interpret as testosterone shows very large variability across time (Crewther et al., 2012). Our study was somewhat unique in allowing us to follow the same team across 6 games removing some of the noise associated with this variability. The longitudinal approach taken in this study allowed a more comprehensive understanding of pregame testosterone relationships to game outcome, whilst also testing the stability of these outcomes. Arguably this finding may be a peculiarity of this team and group of individuals, but certainly suggests merit in repeating this across other teams.

There is reasonable rationale for using free testosterone as a pregame predictive tool. Individual variance in free-testosterone concentrations has been linked to aggressive and dominant behaviour (Elias, 1981, Gladue et al., 1989, Salvador et al., 1999) and to the expression of maximal power and short distance (time) speed (Crewther et al., 2009b). Expressed across several individuals within a team, it is not hard to speculate the advantages of these attributes in a game of rugby. The pregame T/C ratio also showed a moderated correlation to game outcome, but this of course may simply reflect changes in testosterone concentrations and the lack of statistical change in cortisol. Some studies have suggested the use of the T/C ratio as a measure of recovery; with a high T/C ratio indicative of positive anabolic state; (Banfi, 1998) hence, speculatively, it is possible that a higher T/C ratio in the present study indicates that players’ recovery was better prior to winning outcomes. The ability to recover is of particular importance across a rugby season due to the weekly exposures and extremely physical demands of the sport.

Beyond the supporting role for testosterone in performance, there are many other factors that are responsible for outcome during a game (e.g. refereeing decisions, weather, skill execution and opponents’ performance). Home-versus-away hormonal data and historical statistical win/loss data did not show any causality on pregame salivary hormones. Studies (Carré, 2009, Carré et al., 2006) have shown that playing at home can offer both psychological and hormonal advantages for some

52 athletes, when compared to an opponent’s home ground. Further to this, it could be plausible that the opposition team may also have influenced the anticipatory level of pregame testosterone, due to performance expectations based on previous matches played. However we could not conclude from our results that these specific factors had a direct effect on pregame levels of salivary hormones and in particular testosterone. The players involved in this study were well-seasoned elite athletes who at the time were a championship contending team. The driving factors behind their success may have been a collective and internal effort thereby minimising external distractions detrimental to performance (e.g. game venue).

Such a theory, however, needs more investigation across a larger number of games. This point highlights that although hormone status of an athlete is important, there are many other factors that contribute to game outcome that may not be controllable. As such pregame measurement of free- testosterone concentration has probability value but only amongst numerous other factors.

We saw no strong evidence to support an effect of game outcome on perigame testosterone. This observation is somewhat in contrast to findings in other sports (Booth et al., 1989, Elias, 1981, Gladue et al., 1989, Gonzalez-Bono et al., 1999, Fry et al., 2011). However, in those other sports the observation is complex with dependencies on gender, situation of measurement, team “culture” and individual perceptions (e.g. team outcome vs individual perception of self-performance) to name a few examples. Preliminary data (unpublished) that we have collected suggests that the relationship of testosterone changes (pregame to postgame) to the game outcome is team-culture dependent, being stronger in teams that share strongly in the outcome rather than in the individual performance.

Examining the results by playing position (forwards vs backs) gave some interesting additional observations. Pregame testosterone concentrations and T/C in the backs were the only hormonal measure to show significance to the game outcomes. Forwards showed no significant relationship between any game-day hormonal measure and the outcomes. Initially these results surprised us considering the requirements of the game, given that forwards are involved ostensibly in more contact and aggressive situations than the backs (Duthie et al., 2003). The backs clearly spend more time in free running at speed, (Duthie et al., 2003) but also in a considerable degree of aggressive encounters and contact at high speed, (Smart et al., 2008) and these findings parallel a recent study on elite rugby union players where the salivary testosterone concentrations of players, especially the backs, correlated to speed, power and strength (Crewther et al., 2009b).

53 We acknowledge that statistically only 6 games were followed; however the results seen in this study add to those seen in other sporting contexts (Booth et al., 1989, Elias, 1981, Gonzalez-Bono et al., 1999) and present the novel finding of a strong relationship between pregame testosterone concentrations and rugby outcome.

It is also important to acknowledge that the costs of performing such an intensive and comprehensive sampling number collection are high and this argues against the use of hormones as a routine monitoring tool. This study also acknowledges the limitations of using saliva over blood sampling to determine hormonal status. Although blood samples are needed to substantiate the actual status of the gonadal function, the elite athlete situation does not easily allow the opportunity to obtain blood. On the other hand using saliva to analyse endogenous hormones is noninvasive and stress-free (Lewis, 2006). Salivary testosterone and cortisol concentration measures correlate well with the blood hormones, especially the free hormone that initiates the biological response at target tissue (Crewther and Cook, 2010a) and the bioavailable hormone that is potentially available to tissue (Arregger et al., 2007). So although we are not able to understand the mechanism behind gonadal function under competitive stress, we are able to still obtain valuable descriptive data of elite athletes under real competitive situations. This type of data is lacking in the current literature and this study helps to provide that necessary first step to understand such responses.

In conclusion, this study presented a strong association between concentrations of free-testosterone prior to games and actual game outcomes in professional rugby players across 6 matches. Given that there are physical and psychological methods by which free-testosterone concentrations can be acutely elevated, it would interesting to see if an acute pre-match change in testosterone could subsequently affect game performance and the outcome.

3.5 Practical Applications

The current findings suggest a link between pregame free testosterone levels and success in an aggressive, contact-based professional team sport. At present, the cost and time to analyse pregame hormonal data are too high and too slow, respectively to suggest routine use. However, examining how physical and psychological methods can acutely alter prematch testosterone concentrations and how training-week actions influence it offers potential avenues for managing performance on game week and improving team readiness to compete on game day. In addition, exploring the link between game-day T/C and performance in rugby backs suggests that the monitoring of training and game loads (e.g. through the use of global positioning system technology) and employment of appropriate recovery techniques may also assist with optimal game-day preparation.

54 4. Relationship Between Match Statistics, Game Outcome and Pre- Match Hormonal State in Professional Rugby Union

Citation: Gaviglio, C. M., James, N., Crewther, B. T., Kilduff, L. P., & Cook, C. J. (2013). Relationship between match statistics, game outcome and pre-match hormonal state in professional rugby union. International Journal of Performance Analysis in Sport, 13(2), 522−534.

4.1 Introduction

The use of team performance indicators (PIs) as a method of match analysis is a widely used practice in sport particularly in professional team sports to help determine playing strategy and player selection on a game-by-game basis (Jones et al., 2008). The information gathered for the calculation of a performance indicator (PI) is obtained through the selection and combination of variables that define some aspect of performance (Ortega et al., 2009, Hughes, 2002). Thus a PI can provide valuable insight into the performance of individuals and their relative capacity or ability (Bracewell, 2002). Therefore, changes in PI measures from game to game may also be useful as an index of recovery and readiness to compete.

Rugby union is a sport characterised by a high degree of aggressive impacts (e.g. contacts) between players, with the addition of high skill component (e.g. passing and kicking the ball to create running space). Within a team of 15 players, there are two major groups classified as forwards (n = 8) and backs (n = 7). Generally, forwards are involved in more body-contact and relatively static high strength and power situations (e.g. rucks and scrums) than the backs. Conversely, the backs spend more time in free running, passing and kicking allowing them to cover a greater distance during a game (Duthie et al., 2003). Therefore, the use of match statistics as a measure of individual performance ideally needs to accommodate the different playing requirements of each position (James et al., 2005).

Rugby union game-related statistics have been studied to identify differences between winning and losing teams. With the exception of tries scored, the actual game-related skills that influenced outcome varied greatly. Winners tended to be more effective in territorial advances, kicking strategies, defence and effecting turnovers (rucks, scrums & lineouts) (Jones et al., 2004, Ortega et al., 2009, Vaz et al., 2010). Hunter and O’Donoghue (2001) examined data from the 1999 World

55 Cup games and noted that the frequency with which the winning team invaded the rival's 22 m zone, and the points that these teams achieved each time that they invaded the rival's 22 m zone, were higher than losing teams. Hughes (2001) also found that the forwards on winning teams are more effective in the line-out and in the scrummage. Stanhope and Hughes (1997) observed that successful teams in the 1991 World Cup had better performance in the ruck and recovered more possession. Finally, Jones et al. (2008) and James et al. (2005) demonstrated that winning teams obtain more balls in the lineout whilst in the rival’s zone of the field. In all of these studies, they identified the PI as a skill with no classification towards the type of action (e.g. aggressive-based). Furthermore, these aforementioned studies looked at correlating single game-related skills with outcome, rather than providing an overall score that would be indicative of team performance.

The creation of a suitable match statistic summary for each positional player that derives the relevant information for the coach is often time consuming. Therefore, any approaches to objectively quantify broad categories of actions that encompass numerous games plays are viewed as advantageous. Many of the studies to date suggest that some categorisation of “combat or aggressive interactions” may provide insight into the game outcome, although at present no study has appeared to investigate this.

Hormones have also been studied in relation to rugby union. Testosterone and cortisol measures have been employed with a primary focus on correlations with performance markers such as speed, strength and power, and their development outside of the game environment (Cardinale and Stone, 2006, Crewther et al., 2009b). In recent studies on elite rugby union players (Cunniffe et al., 2011, Elloumi et al., 2003), a post-match decline in the testosterone to cortisol (T/C) ratio was suggested as a marker of recovery (or lack thereof). In a subsequent study (Elloumi et al., 2008), it was demonstrated that testosterone varied with perceived importance of a match by the team, while cortisol was a more general marker of stress including training load. Advancing an understanding of hormones to behaviour, in particular behaviour that has an outcome in match statistics would seem an important step forward.

Little focus has been placed upon salivary hormones testosterone and cortisol and their relationship to game day PIs and game outcomes in rugby union. In other sports such as (Scaramella and Brown, 1978), serum testosterone correlated with aggressive behaviours (including but not exclusive to game-specific related actions) but not to actual PIs measured. In an individual-based sport of judo, significant correlations between serum testosterone and offensive behaviours (e.g. threat, fighting and attack) have been observed (Salvador et al., 1999, Obmiński, 2009). In the study by Obmiński (2009), one athlete demonstrated a pre-fight serum testosterone concentration 56 approximately 3.6 times less than his pre-study base concentration recorded 4 days prior. This athlete also demonstrated a depressed mood and passive behaviours during the fight, which he unexpectedly lost. A corresponding increase in serum cortisol concentration was also reported. In basketball (Gonzalez-Bono et al., 1999), the free testosterone response correlated positively with the ‘score/time playing’ ratio, an indicator of individual participation in the outcome. However, again little consideration was given to a clear link through PI analyses and game outcome.

Therefore, the primary aim of the current study was to analyse a season of professional rugby matches from a team involved in elite competitions and to determine if PIs from match statistics discriminated between winning and losing. Further to this, a secondary aim was to examine the relationship of PIs to game-day measures of free testosterone and/or cortisol concentrations across a subset of 6 consecutive games. It was hypothesised that an aggressive based PI would discriminate between game outcomes and that pre game testosterone concentrations would correlate to this PI score.

4.2 Methods

4.2.1 Participants

This study involved 22 elite male rugby union players (age 27.8 ± 4.0 y, height 1.87 ± 0.08 m and mass 103.4 ± 11.6 kg) from a professional rugby union team. This rugby team competed in both the English Rugby Union Premiership and European Championship, and a number of the subjects were also international level players. Subjects were informed of the study protocols and signed informed consent before testing began. Ethics approval was obtained from a University Ethics committee.

4.2.2 Design

This study was conducted over a full 35-week season of rugby union. Saliva sample collection was conducted mid-season over the course of 6 weeks. Players continued to train mid-week according to their usual training schedule and the games were played on the weekend. On each of the days immediately prior to testing, players were encouraged to sleep well (>7 h), consume a good breakfast (each player standardised across testing sessions) and maintain their fluid intake (at least 750 mL) during the two hours prior to each testing session.

4.2.3 Game Day Performance

Game-day analysis focused upon the different PI scores, as assigned by the club statistician. The PI scores were calculated by an equation based on rating specific performance based actions performed during the game. These variables were arithmetically summed for each player based 57 upon how many times they completed each action. The weighting of each variable was pre- determined by the coaching staff and formed the basis of the algorithm. Poor performance was thus reflected in a lower score due to less successful execution but was not subtracted in this simple algorithm.

The PI was divided into three indices based upon skill and aggression variables. The three scores were an aggression PI (AggPI), skill PI (SkillPI) and an equal weighted additive combination of the two original PI (AggPI + SkillPI) to give a skill and aggression PI (SkAggPI).

The equation and variable explanation used to calculate the aggression PI score: AggPI = (Tackle Wins + Ball Carries and Dominant + Clear-out: Effective) + (Contacts/2)

Variable Explanation:

1. Tackle Wins describes how aggressively the tackler competed for the ball in the tackle i.e. soak up = loss (−1 point), a covering tackle with no option to pinch the ball = neutral (0 point), a dominant hit and attempted counter ruck = win (+1 point).

2. Ball Carries and Dominant calculates if the player dominated the contact when they were tackled with the ball.

3. Clear-out: Effective calculates if the player was effective at removing the opposition player that they attempted to clear away at ruck, maul or tackle.

4. Contacts is a relationship between the following skills:

a. Ball carry – amount of times the player carried the ball into contact (i.e. to be tackled with the ball.)

b. Tackles made.

c. Offloads – attempting to pass the ball onto a teammate whilst being tackled.

d. Clear-outs – when a player attempts to “clear-out” the opposition player up and off the ball.

e. Seal – as a defender, the player “seals” the ball at the tackle attempting to make the ball unplayable.

58 The equation and variable explanation used to calculate the Skill PI score: SkillPI = Jackal + Defensive Errors + Pass % + Passes + Kicking % + Open Kicking %

Variable Explanation:

1. Jackals (turnovers) forced. A jackal is performed when the opposition player (attacking team) is tackled and the tackler will attempt to “steal” the ball.

2. Defensive errors focuses on the number of tackles made that resulted in a defensive error.

3. Pass % is a relationship between the number of successful passes and offloads and the total number of passes and offloads.

4. Passes is a value based on the total number of passes or offloads made irrespective of their success. 5. Kicking % is a relationship between the number of good kicks at goal or touch from a penalty and the total number of attempts.

6. Open Kicking % is the number of kicking errors as a percentage of total kicks in an open field of play. These types of kicks may or may not result in regaining possession. They include:

i. general kicking with the field of play,

ii. kicks that regain possession,

iii. restart kicks after the opposition scores and

iv. kicks that go out of field of play resulting in an opposition line-out.

The equation used to calculate the SkAgg PI score:

SkAggPI = SkillPI + AggPI

4.2.4 Hormone Assessment

Saliva samples were taken at times chosen to avoid interrupting the players’ game preparation and to keep a diurnal consistency in sampling. Saliva was used due to its ease of compliance, low invasiveness, and ability to track the biologically active “free” hormone (Arregger et al., 2007, Crewther and Cook, 2010a). Pregame samples were taken 40 minutes prior to the start of the game. Post-game samples were taken 15 minutes after the player was substituted off the field or after the game’s completion. For each test, a 2 mL saliva sample was collected and stored at –20 °C before assay. Saliva was assayed for testosterone and cortisol by a private commercial laboratory (HFL Sport Science Laboratories, UK) using diagnostic kits (Salimetrics Enzyme Immunoassay Kits, 59 Salimetrics, USA) and manufacturer’s instructions. Inter-assay coefficients of variation (based on low and high control samples) for testosterone and cortisol were both < 10%.

4.2.5 Statistical Analyses

The PI (AggPI, SkillPI and SkAggPI) scores for each player and game was grouped according to outcome (i.e. win and loss) and compared using dependent t-tests. Pre and post-game hormone concentrations and the hormonal changes (peri) across a sub-set of 6 games were tabulated for each player. The peri hormone value was calculated by dividing the post hormone value by the pre hormone value (post ÷ pre = peri). The testosterone to cortisol (T/C) ratio was also calculated and used during analysis. The bivariate relationships between mean pooled hormonal variables and PI scores were assessed using Pearson product moment correlation coefficients (r). The criterion level for significance was set at p ≤.05.

4.3 Results

Analysing all PI scoring variations, only the AggPI significantly (p <.01) differentiated between a win and a loss (Table 4.1). Neither SkillPI nor SkAggPI, showed any ability to discriminate between game outcomes.

Table 4.1 Comparison of match performance indicator scores of elite rugby union matches according to outcome. PI Win (n = 15) Loss (n = 16) AggPI 14.0 (4.9) 12.7 (4.6)** SkillPI 13.9 (16.6) 13.7 (17.2) SkAggPI 27.9 (17.1) 26.5 (17.5) Data presented as mean (SD). PI = performance indicator. Significant from a Win **p <.01

AggPI showed relationships with game day concentrations of pre testosterone (r = .86, p <.05; Figure 4.1), pre T/C ratio (r = .86, p <.05; Figure 4.2) and peri testosterone (r = –.87, p <.05; Figure 4.3).

Although SkillPI and SkAggPI were not able to discriminate between outcomes, both PI scores did show relationships with various game day hormone variables. We observed significant correlations between SkillPI and peri cortisol (r = –.98, p <.01) whereas SkAggPI significantly correlated with post T/C (r = .85, p <.05).

60

Figure 4.1 Relationship between pregame testosterone concentration and aggression performance indicator score. Data points represent team mean values. Regression lines presented for the team (solid).

Figure 4.2 Relationship between pregame T/C ratio and aggression performance indicator score. Data points represent team mean values. Regression lines presented for the team (solid).

61

Figure 4.3 Relationship between the change in pre- to post-game (peri) testosterone concentration and aggression performance indicator score. Data points represent team mean values. Regression lines presented for the team (solid).

4.4 Discussion

The main finding of this study was the ability for an aggressive action-based team PI score (AggPI), to differentiate between outcomes (i.e. win–loss) across 31 elite professional rugby union games. A skill-based (SkillPI) and combined skill-aggression (SkAggPI) score however, were not able to differentiate between game outcomes. A secondary observation was that in a smaller sample size of these games (n = 6), pregame concentrations of testosterone, T/C ratio and peri testosterone significantly correlated to AggPI score (and therefore also win-loss outcomes).

The ability for a single team based score to discriminate between outcomes provides an interesting point for discussion. Previous rugby union studies retrospectively looked at individual performance parameters during a game to provide a game related statistic that correlated with outcome. The performance parameters that were associated with the game outcome varied amongst studies, however, winners tended to be more effective in territorial advances, kicking strategies, defence and effecting turnovers (rucks, scrums & lineouts) (Jones et al., 2004, Ortega et al., 2009, Vaz et al., 2010). Although these findings were single skill related outcomes, all of these variables were prevalent in some form in the calculation of the AggPI used in the present study. We are unaware of any study in rugby union that has investigated using a single team based PI score to discriminate between a win and loss.

62 The ability of the aggressive based statistical score of AggPI to discriminate game outcomes could be postulated to be due to the highly aggressive based nature of rugby union. Despite a high skill factor in professional rugby, it does remain a combat sport. Elevated serum creatine kinase (CK) levels indicative of structural damage to muscle cells has been reported after competitive rugby union matches (Takarada, 2003). Gill et al. (2006) postulated that the main cause of CK elevation is the damage caused to muscle by direct impacts between opposing players. Described often as a contact sport characterised by a high degree of aggressive impacts and interactions (e.g. grappling for possession of the ball) between players, it would lend support to the theory that the measurement of aggressive based actions would correlate with successful performance.

The relationship between testosterone concentration and AggPI agrees with traditional theories of testosterone and aggressive behaviours (Bouissou, 1983, Dixson, 1980), and combativeness, territoriality and dominance in human competitive encounters (Neave and Wolfson, 2003). In other areas of human performance (e.g. share-traders) testosterone has been equated to “bullish” aggressive behaviour and reward seeking while cortisol to risk assessment and conservative behaviour (Coates et al., 2010). It is possible the results we present are very team strategy dependent and it may reflect the nature of this particular team’s winning strategy i.e. highly aggressive game play rather than a generalised rugby team attribute. Other factors may also play a role. For example, Eloumi et al. (2008) demonstrated that a rise in testosterone was in accordance to the perceived importance of a rugby match.

The T/C ratio has also been proposed as an interactive factor in human aggression (Carré and Mehta, 2011, Terburg et al., 2009, Kuoppasalmi et al., 1980, Bosco et al., 1996a, Cardinale and Stone, 2006, Kraemer et al., 2004, Häkkinen et al., 1987). In sport, the T/C ratio has traditionally been documented as a useful value in the early detection of over-training in an athlete (Adlercreutz et al., 1986, Banfi, 1998) or the inability to recover from intensive physical activity (Banfi et al., 1993, Vervoorn et al., 1991) and rugby (Cunniffe et al., 2011, Elloumi et al., 2003). As a predictive marker in golfers (Doan et al., 2007), a low T/C ratio was related to lower scores (i.e. better performance). With this exception, few studies to date have focused and identified a positive relationship between T/C ratio and performance. The relationship seen here between pregame levels of T/C and AggPI may simply be a slight exaggeration of the testosterone results; however, it might also relate to a need for the players to fully recover (as indicated by better T/C ratio) to meet higher neuromuscular demands as a result of their positional requirements and actions (aggressive based) during a game. If the pregame T/C ratio is an indication of the players’ inability to recover from the

63 high intensity physical activity of either the previous week’s game and/or training, then its recognition and monitoring may have value.

Peri-testosterone also showed a relationship with AggPI. In other sports (Booth et al., 1989, Elias, 1981, Fry et al., 2011, Gonzalez-Bono et al., 1999, Gladue et al., 1989), changes in pre to post game testosterone concentrations showed a relationship with outcome; however, in these other sports, the observation is complex with dependencies on factors such as gender, situation of measurement, team “culture” and individual perceptions (e.g. team outcome versus individual perception of self performance). The inverse relationship between peri testosterone and AggPI may primarily be due to the observed pregame testosterone concentrations. In the previous study (Chapter 3), only pregame testosterone concentrations discriminated between game outcomes unlike postgame testosterone and peri testosterone concentrations (Gaviglio et al., 2014). To contextualise this further, the sub-set of 6 games studied for hormone concentrations in this study is the same as Chapter 3. As a higher AggPI score is reflective of a win and within the sub-set of 6 games, the 3 lowest AggPI scores are losses and the 3 higher AggPI scores are wins. Although the relationship between pregame testosterone and AggPI fits well with these observations, the inverse relationship observed between peri testosterone and AggPI does not. Therefore it would be plausible that it was the pregame testosterone concentration as the major contributor to this observation. It is speculative and reductionist to think however, that the game day testosterone concentration may be a primary determinant in games outcomes.

The high variability observed in SkillPI may be due to the players’ different positions and roles within a game. For example, forwards are often considered to be “ball winners” and the backs to be “ball users” and players within these broad groups having different tasks during specific phases of the game (Quarrie et al., 1996). In particular, forwards do not kick during a game and therefore provide very minimal opportunities to contribute to the kicking calculation component of the SkillPI. Among the backs, there are 3 distinct sub-groups, (inside, mid-field and outside backs) and their primary roles differ from each other. The inside backs, for example, control the possession obtained by the forwards and have a high passing skill requirement to deliver the ball to the midfield back or outside back who are required “crash” directly into the opposition team players at speed or are required to be able to beat opposition players (Quarrie et al., 1996). Therefore an outside back may have less of an opportunity to pass the ball. Consequently each player may not be able to effectively contribute to each variable within the SkillPI variables by virtue of their positional roles within a game. Possible refinement to the factors contributing to the PI score may result in a more position specific evaluation of performance.

64 A major limitation in our study is that we have followed only one team, albeit across an entire season of data, so it must be stressed that the applied PI scoring system may only be relevant for this team. Little discussion has been given to the seemingly logical proposition that at any time teams differ in their strategies, coaching approaches and individual player strengths; such that a team that were very strong on aggressive approach to games would be far more likely to show what we have observed in win-loss analyses than perhaps a team that was less so and far more reliant on a high level skill with what ball they obtained. The test of this would be the application of the same PI scoring system to other professional rugby teams to assess its validity as a suitable performance rating system relative to their team strategy. We feel this is a novel and useful approach to viewing game statistics, namely taking into account how a team best plays with its available player and coaching resources, rather than blanket statistics averaged across multiple team results.

4.5 Conclusion

In conclusion, we have demonstrated that a combined aggressive based single team PI (AggPI) can differentiate between a win and a loss across a full of season of games. In addition, in a small subset of these games we observed a strong correlation between salivary free-testosterone concentrations and the pre-match T/C ratio and the execution of the actions producing the AggPI. The results are suggestive of the importance of these hormones in the realisation of these actions and to provide an objective marker of player readiness to perform these highlighted game roles.

4.6 Practical Applications

The present findings suggest it is possible to produce a match based team statistic to discriminate between a win and a loss. In the particular team observed, this was an aggressive based combined total statistic; in other teams it could perhaps be the skill-based statistic. Use of a single team statistic such as AggPI can highlight what strategies and elements of the players’ game needs addressing during the training week. The link between salivary free-testosterone concentrations, the T/C ratio and AggPI shows potential both experimentally and from an application perspective. Given there are physical and psychological methods by which free testosterone concentration can be elevated, it would interesting to see if an acute pre-match change in testosterone positively affects AggPI score and consequent game performance and outcome. Also, understanding how the T/C ratio is affected by a players recovery throughout the week between matches and how that can be optimised offers an interesting experimental approach that could also lead to practical avenues for managing performance come game day.

65 4.8 Acknowledgements

The author wishes to thank David Reed who was the statistician at the club and provided the data and explanations for the formulas used during this study.

66 5. Relationship Between Midweek Training Measures of Testosterone and Cortisol Concentrations and Game Outcome in Professional Rugby Union Matches

Citation: Gaviglio, C. M., & Cook, C. J. (2014). Relationship between midweek training measures of testosterone and cortisol concentrations and game outcome in professional rugby union matches. Journal of Strength and Conditioning Research, 28(12), 3447−3452.

5.1 Introduction

Testosterone and cortisol concentrations have previously been studied in relation to training and performance outcomes in sport (Beaven et al., 2008a, Crewther et al., 2009b, Gladue et al., 1989, Passelergue and Lac, 1999). In particular, increases in concentrations of endogenous testosterone have been shown to correlate positively with elements of athletic performance (Crewther et al., 2009b, Cardinale and Stone, 2006, Salvador et al., 1999). Likewise, cortisol concentrations have also been linked to competitive behavior and performance (Elias, 1981, Salvador et al., 1999, Suay et al., 1999). However, the relationship between testosterone and cortisol concentrations and successful competition outcomes remains equivocal (Booth et al., 1989, Elias, 1981, Gladue et al., 1989, Gonzalez-Bono et al., 1999).

The testosterone to cortisol (T/C) ratio has traditionally been viewed as a putative marker of the anabolic to catabolic hormonal balance within the body (Budgett, 1998, Banfi et al., 1993); however it is unclear how this may directly link to performance. The T/C ratio has further been suggested as a useful tool in the early detection of overtraining (Adlercreutz et al., 1986, Banfi, 1998) and as a measure of recovery from intensive physical activity (Banfi et al., 1993, Vervoorn et al., 1991). Few studies have investigated the relationship between the T/C ratio and performance outcomes in sport (Kraemer et al., 2004, Cunniffe et al., 2011, Doan et al., 2007).

Data on endogenous hormones and competition has historically focused on the acute response around the timing of the competition event, with little focus on the interplay of midweek hormones to subsequent competitive performance a few days later. The role and interaction of testosterone and cortisol is complex in nature, with both hormones affected via multiple mechanisms including, psychological (Archer, 2006, Archer, 1991), social (Archer, 1991, Mazur, 1985), physiological 67 (Viru et al., 1992, Urhausen et al., 1995) and competitive (Booth et al., 1989, Elias, 1981, Suay et al., 1999) events. A recent study in rugby league demonstrated that testosterone responses to mid- week resistance training workouts showed significant associations with subsequent game outcome a few days later (Crewther et al., 2013). To our knowledge this is the only study to date that has shown such a relationship.

Having data to better understand the preparation of a sporting team for an upcoming competition has important implications in advancing the management, recovery and training of the athletes leading into an event. In rugby union this preparation includes the development of decision-making skills with a goal to provide a positive transfer of player actions practiced during these training sessions to the competition environment (Passos et al., 2008). Therefore, this study of professional rugby union players assessed the relationship between absolute concentrations of salivary testosterone and cortisol, and the T/C ratio across selected midweek skill-based training sessions and the subsequent game-day performance outcome 3 days later. It is hypothesised that midweek pre- and post-testosterone concentrations and T/C ratio values would be greater when rugby union matches were won rather than lost.

5.2 Methods

5.2.1 Experimental Approach to the Problem

There is limited information on the use of salivary testosterone and cortisol as a measurement tool to identify likely outcomes in competitive team sport. Therefore, this study focused on the salivary- free testosterone and cortisol responses to midweek skill-based training sessions and subsequent outcomes of professional rugby union games. Game-day outcomes included match results (win or loss) and pregame testosterone concentrations. Pregame testosterone levels were assessed due to higher concentrations being previously associated with successful outcomes in this team (Gaviglio et al., 2014). Saliva was used due to its ease of compliance, low invasiveness and ability to track the biologically active “free” hormone (Crewther and Cook, 2010a). As this was an in-season professional rugby environment, all testing was performed under normal training conditions, which improved the internal validity of the study findings and their applications within elite sport.

5.2.2 Subjects

This study involved 22 elite male players from a professional rugby union team (age, 27.8 ± 4.0 years; height, 1.87 ± 0.08 m; and mass, 103.4 ± 11.6 kg). This team competed in both the English Rugby Union Premiership and the European Championship, and a number of the subjects were also playing at international level. Subjects were volunteers. They were informed of the study protocols

68 and they signed written informed consent documents before testing began. All subjects were above 18 years of age. Ethics approval was obtained from a University Ethics Health Sciences committee in compliance with national legislation and The Code of Ethical Principles for Medical Research involving Human Subjects of the World Medical Association (Declaration of Helsinki).

5.2.3 Assessment Procedures

This study was conducted midseason over the course of 6 weeks between the months of November to January. Players continued to train midweek according to their usual training schedules and the games were played during the weekend on a home and away basis. Weekly training sessions at the time of this study generally involved 2 gymnasium sessions, 4 rugby skill training sessions and recovery work after these skill-based sessions. As this was an in-season professional rugby environment, there were practicalities of normal training that had to be adhered to. On each of the days immediately prior to testing, players were encouraged to sleep well (>7 h), consume a nutritionally sound breakfast (each player standardised across testing sessions) and maintain their fluid intake (at least 750 mL) during the 2 hours prior to each testing session. No food was consumed within 30 minutes of sessions.

5.2.4 Midweek Skill-Based Training Session

Participants were monitored once a week across 6 midweek skill-based training sessions during the competitive season. These sessions were designed to assist with the team preparation for the subsequent game on the weekend. A skill-based training session was chosen due to the specificity of the training session to the actual game requirements. The training sessions were performed 4 days after each match-day game, to ensure that hormonal and neuromuscular recovery had been achieved (McLellan et al., 2010, McLellan et al., 2011) and 3 days before the next game. Each session was conducted at the same time each week to account for diurnal variation (Kraemer et al., 2001b) and formed the main rugby skills session for the week. Because of the importance of this session to the upcoming game, session time and content were kept as consistent as possible (Table 5.1). Session intensity and effort were monitored using a rate of perceived exertion (RPE) measurement scale where 1 = very light and 10 = maximal (Borg, 1982).

69 Table 5.1 Midweek skill training session outline and timing of events. Time (min) Event Comments 0−2 Pre saliva sample 2−10 Preparation Relevant musculoskeletal treatment. 10−30 Meeting Review of upcoming game. Vision (video) used to assist in the meeting. Outline of rugby skills session. 35-50 Warm-up General movement and jogging. Agility drills involving rugby ball. Running drills and exercises of increasing speed and intensity. 50-60 Units Backs and forwards split into their respective playing groups to practice their specific plays and routines. 60-70 Attack The backs and forwards rejoin to practice moves the team will use when attacking against the opposition. 70-80 Defence Defensive moves to be used against the opposition. 80 Conclusion of session Brief discussion about the training session. Players leave the training field. 90 Post saliva sample After post saliva sample players commence recovery modalities (i.e., ice baths, physiotherapy and massage).

5.2.5 Game Day Performance

Two factors were assessed on game day, namely, match outcome of win or loss, and pregame testosterone concentrations (Table 5.2).

Table 5.2 Subsequent competitive game performance after the midweek skill-based training session.^† Midweek training session 1 2 3 4 5 6 Outcome Loss Loss Win Loss Win Win Pre-T (pg⋅mL-1) 78 (22) 99 (31) 126 (44) 95 (32) 105 (41) 111 (31) Venue Home Away Home Away Home Away Win% 70 37 50 50 48 63 ^Outcome = match result; Pre-T = pregame testosterone concentrations; Win% = historical win percentage (http://www.premiershiprugby.com) from all of the games played previously against that relevant opposition. †Data represented as mean (SD).

5.2.6 Hormone Assessment

Saliva samples were collected at times chosen to minimise any interruption to the players’ training and game preparation. Presession samples were taken 30 minutes prior to the start of the training

70 session or match. Posttraining samples were taken 10 minutes after the conclusion of the midweek training session. For each test, approximately 2 mL of saliva was collected by passive drool into sterile containers and stored for less than three months at –20° C (Granger et al., 2004) before analysis. Saliva was assayed for testosterone (pg⋅mL-1) and cortisol (µg⋅dL-1) concentrations by a private commercial laboratory (HFL Sport Science Laboratories, UK) using commercially available kits (Salimetrics Enzyme Immunoassay Kits: Salimetrics, State College, PA, USA) and were run in duplicate according to the manufacturer’s instructions. Interassay coefficients of variation based on low and high control samples for testosterone and cortisol were both < 10%.

5.2.7 Statistical Analyses

Pre- and post-midweek hormone concentrations were pooled according to subsequent outcome (wins and losses) and compared using dependent t-tests. The testosterone to cortisol (T/C) ratio and percentage change in pre to post hormone concentration variables were also calculated and used during the analysis. The bivariate relationships between the midweek hormonal variables and pregame testosterone concentrations were assessed using Pearson’s product moment correlations. An alpha level of α = .01 was set a priori.

5.3 Results

5.3.1 Hormone Concentrations vs Wins and Losses

A significant increase in the T/C ratio from pre to post values was noted before winning games (p <.001; Figure 5.1). The preskill T/C ratio prior to a win was significantly lower (p <.01) than when compared to a loss.

Figure 5.1 Preskill and postskill training salivary testosterone to cortisol (T/C) ratio before winning and losing games (mean ± SE). Significant from postskill training (wins) ^p <.001 and preskill training (losses) *p <.01.

71 A significant increase in pre- to post-testosterone concentrations was observed in midweek training sessions when subsequent games were won (p <.01). No significant difference was observed when the match was lost (p >.01; Figure 5.2).

160 *

) 140 -1 l

m 120 · g

(p 100

e 80 on r

te

s 60

sto

e 40

T 20 0 Pre-skill Post-skill Pre-skill Post-skill

Wins (n=3) Losses (n=3) Figure 5.2 Preskill and postskill training salivary testosterone concentrations before winning and losing games (mean ± SE). Significant from preskill training *p <.01.

Cortisol concentrations did not demonstrate significant change (p >.01) across midweek training sessions before subsequent wins (pre, 0.23 ± 0.14 vs post, 0.18 ± 0.12) and subsequent losses (pre, 0.19 ± 0.15 vs. post, 0.21 ± 0.16). For both testosterone and cortisol there were no significant differences when either the preskill or postskill sample values were compared in isolation in either games won or lost.

5.3.2 Hormone Responsiveness vs Wins and Losses

The relative changes in hormone concentrations across the midweek training session did not highlight any significant differences (p >.01) between games won and lost; testosterone (win, 37 ± 60%; loss, 30 ± 64%), cortisol (win, 1.4 ± 65%; loss, 62 ± 167%) and the T/C ratio (win, 81 ± 110%; loss, 38 ± 116%).

5.3.3 Midweek Hormones vs Pregame Testosterone

The cortisol and T/C ratio responses (%) to the midweek skill session displayed strong and significant correlations to pregame testosterone concentrations (Table 5.3). We also found no significant correlations between preskill or postskill training session hormones and pregame testosterone concentrations (data not presented).

72

Table 5.3 Correlations between the relative (%) hormonal changes across each skill-based training session and pregame testosterone concentration. † Pre-T (pg⋅mL-1) Testosterone (% change) .09 Cortisol (% change) –.90* T/C ratio (% change) .90* †Pre-T = pregame testosterone concentrations. *p =.01.

5.3.4 Game Data - Rate of Perceived Exertion There was no significant difference when RPE data was grouped according to their subsequent game outcome (win, 5.9 ± 1.3; loss, 6.1 ± 1.1).

5.4 Discussion

In elite professional rugby union players, the T/C ratio and salivary testosterone concentration increased significantly in response to a midweek skill-based session prior to a win but not prior to a loss. Furthermore, the percent change in the midweek T/C ratio and cortisol concentrations demonstrated strong correlations with subsequent pregame testosterone concentrations. In this group of players, pregame testosterone concentrations have been previously associated with successful game outcomes (Gaviglio et al., 2014).

Midweek concentrations of the T/C ratio as an indicator of competition match outcomes in rugby union have not been presented previously. However, the observed increase in testosterone concentrations across the skill-based session reflected the results of a recent study of rugby league players, where an elevated midweek testosterone response across a resistance training workout was able to highlight successful outcome in games played a few days later (Crewther et al., 2013). Training and competition measures of testosterone concentrations have shown to correlate with aggressive behaviour, motivation and mood state in different athletic groups (Mazur and Lamb, 1980, Salvador et al., 1999, Suay et al., 1999, Cook et al., 2013), all of which contribute to physical performance in sport. Alternatively, transient increases in testosterone concentrations (which could be cumulative across numerous training sessions) might have a delayed effect on those adaptive systems directly influencing game-day performance, or these testosterone patterns could simply reflect the recovery state of athletes from previous games (Crewther et al., 2013).

Although salivary testosterone concentrations demonstrated an ability to highlight successful outcomes in this current study, the T/C ratio was perhaps a more comprehensive indicator of 73 competition success. A number of studies have described the interaction of testosterone and cortisol in relation to psychophysiological stressors (Zilioli and Watson, 2012). Therefore, the T/C ratio observations may be due to the stronger cognitive requirements of a skill-based training stimulus (Passos et al., 2008) as opposed to that required across a resistance training workout (Crewther et al., 2013). As such, it could be suggested that the dual mechanism of a testosterone and cortisol ratio might have more robustness across a range of midweek psychophysiological stressors.

It has been suggested that the T/C ratio is an indicator of physiological strain in training (Urhausen et al., 1995) and a marker of the anabolic-catabolic status of the athlete (Budgett, 1998, Banfi et al., 1993). Hence, it could be possible that prior to a win, the increase in the midweek T/C ratio highlighted a positive endocrine response to the training session, which was carried through to the subsequent game as indicated through the association with game-day pre-testosterone concentrations. The T/C ratio prior to a midweek skill training session was significantly lower preceding winning games than preceding games that were lost. Although this result may be interpreted as an indicator of poor recovery from the previous game, this rather may be a transient observation.

The relationship between the midweek cortisol response and pregame testosterone concentrations further highlights the interactive factor between midweek hormones and competitive performance. Typically, an increased concentration of cortisol is known to have a protein-catabolic affect on skeletal muscle (Urhausen et al., 1995). In this present study, an increase in the percentage change of midweek cortisol concentrations negatively correlated with pregame testosterone concentrations. It could therefore be suggested that a lower midweek cortisol response to a skill-based session would be preferred because an increase in cortisol concentrations has been shown to be associated with a concurrent decrement in testosterone concentration affecting aggressive and dominant behaviors (Mehta and Josephs, 2010). As indicated in previous chapters, higher pregame testosterone concentrations (Gaviglio et al., 2014) in association with an increase in aggressive based actions (AggPI) (Gaviglio et al., 2013) within this group of current athletes are associated with winning games.

The hormonal response to the midweek training session may have also been a result of other factors such as the pretraining session meeting or coaching style. For example, cautionary coaching feedback in rugby union players has previously resulted in larger increases in salivary cortisol concentrations when compared against positive coaching feedback (Crewther and Cook, 2012). Furthermore, subsequent pregame testosterone concentrations and game-ranked performance several days after the coach’s feedback reported to be significantly improved as a result of positive 74 coaching feedback. In the current study, the coaching staff did not change during the testing period and their coaching style was not monitored for any variances in feedback, language or communication. Training sessions were traditionally preceded with a team meeting discussing the upcoming game-day tactics incorporating video clips to assist with explanations. It is possible that these factors may have impacted on hormone concentrations based upon the observations from Cook and Crewther (2012b), however this warrants further investigation. Nonetheless, to our knowledge this is the first study to demonstrate midweek values of the T/C ratio across a mixed psychological-physical stressor to distinguish between competitive performance outcomes a few days later.

A limitation of this study is that only a small number of training sessions and games (n = 6) were monitored although this number is comparable to other research in rugby union (Maso et al., 2004, Takarada, 2003, Crewther et al., 2013) and rugby league (McLellan et al., 2010). We also acknowledge that training load volume and intensity can influence testosterone and cortisol responses (Beaven et al., 2008b). Therefore, training sessions were diligently planned to ensure that the starting time and duration were kept as consistent as possible. Though the level of participation could have varied slightly between player and between sessions, the RPE data indicated that there was no significant difference when grouped according to their subsequent game outcome (win, 5.9 ± 1.3; loss, 6.1 ± 1.1). Despite the proposed effects of game venue (home vs away) and historical performance statistics (i.e., win %) against the opposition (Neave and Wolfson, 2003), neither had a discernable effect on game outcome in this study. The results presented here, whilst preliminary, warrant further investigation employing a larger sample of midweek skill and game outcomes whilst taking into account coaching techniques, training session content and individuals. Valuable information that is lacking in current literature has been obtained from this study, namely the descriptive endocrinological data of elite level performing athletes under real training and competitive situations. In a team sport where skill-based training sessions form a large portion of a training week, the findings from this study may better help to better understand the overall response to training sessions and their impact on competition results a few days later.

5.5 Practical Applications

Our findings support the use of midweek measurements of salivary hormones in response to a psychophysiological stressor and as a tool for examining potential game readiness in professional rugby union. Specifically, in this study the T/C ratio and salivary testosterone concentration increased significantly in response to a midweek skill-based session prior to a win but not prior to a loss. The practical application to the strength & conditioning professional presents itself in the form 75 of how the training week is planned and how the information around the training week is presented (i.e., meetings, vision, coaching styles). Assessing neuromuscular measures (e.g., vertical jump) may also provide an alternative to sampling salivary hormones (Crewther et al., 2009b, Cardinale and Stone, 2006). However a longitudinal approach to collecting this data around subjective and objective training and competition parameters would be required to identify such relationships. This information may assist in diagnosing midweek training responses and help with match-day and training week interventions (passive and active) to increase the probability of competitive readiness and successful outcome in competition.

5.6 Acknowledgements

The authors declare that they have no conflict of interest and the results of the present study do not constitute endorsement of any products by the authors or the National Strength and Conditioning Association.

76 6. Salivary Testosterone and Cortisol Response to Four Different Rugby Training Exercise Protocols

Citation: Gaviglio, C. M., Osborne, M., Kelly, V. K., Kilduff, L. P., & Cook, C. J. (2015). Salivary Testosterone and cortisol responses to four different rugby training exercise protocols. European Journal of Sport Science, 15(6), 497−504.

6.1 Introduction

Strength, power and lean muscle mass are important requirements for rugby union athletes (Beaven et al., 2008b, Duthie et al., 2003). Consequently, resistance exercise (RE) is an integral part of a rugby union player’s training schedule. Depending on the type of RE, benefits include improvements in power, strength, muscular endurance and hypertrophy (Beaven et al., 2008b). Strongman (STRNG) and combative (COMB; boxing and wrestling) sessions are also regularly used in a rugby union player’s training schedule as an alternative training stimuli. Strongman training consists of multi-joint movements that incorporate large muscle mass groups resulting in significant neuromuscular stress (Ghigiarelli et al., 2013). Combative training sessions based on boxing and wrestling training are another popular alternative training modality. Wrestling training has been proposed to assist rugby players with elements of grappling and body positioning around an opposition player (Duthie et al., 2003). Boxing has also been traditionally used as a preferred method to condition rugby union players.

Both RE (Beaven et al., 2008b) and STRNG training (Ghigiarelli et al., 2013) have been shown to temporarily elevate acute concentrations of endogenous hormones. Although studies have investigated the hormonal response to wrestling bouts (Elias, 1981) no study to date has explored the hormonal response to a COMB training session based around a combination of wrestling and boxing exercises. It has been well documented that RE in both athletic and non-athletic populations improves the neuromuscular system through morphological (e.g. muscle size) and/or functional (e.g. strength, power) changes (Kraemer et al., 2002). The prescription of RE workouts is based on the principle that acute elevations in testosterone offer both short-term benefits from training performance and long-term adaptations (Ahtiainen et al., 2003).

77 It has traditionally been viewed that the specific loading parameters of RE influence the observed pattern of endocrine response (Kraemer and Ratamess, 2005). Hypertrophy protocols (high total work or volume, moderate loads, combined with short rest periods) have typically produced large acute increases in various anabolic (e.g. testosterone, growth hormone, growth factors) and catabolic (e.g. cortisol) hormones to promote protein metabolism and muscle growth during the recovery period (Crewther et al., 2011). Conversely maximal strength and power workouts (lower total volume with longer rest periods) have been suggested to produce little or no hormonal change (Crewther et al., 2011). However, West and Phillips (2012) suggest that the acute changes in testosterone across resistance workouts reflect a stress response rather than a hypertrophy signal. This work demonstrated that acute changes in testosterone do not correlate with hypertrophy but rather focused on the potential of testosterone to motivate effort in training as a driver to improve performance and morphological changes. This theory was reflected in the findings from a recent study of rugby union players where the acute hormonal response was specific to the workout design, and that grouping the athlete’s by their optimum individual response gave a different result to the traditional grouping as a single population approach (Beaven et al., 2008b).

Elevated hormonal concentrations have many proposed performance benefits both in training and competition (Elias, 1981, West and Phillips, 2012, Kraemer et al., 2001a). Previous studies in this group of athletes have demonstrated a strong association between mid-week and game-day salivary hormones concentrations and successful game outcome (Gaviglio and Cook, 2014, Gaviglio et al., 2014) speculating that these may be worth monitoring leading into competition. In particular the potential importance of mid-week hormones and subsequent readiness to compete highlights the importance of identifying training protocols that can elevate hormones to provide an optimal environment to maximise functional gain (Beaven et al., 2008a) and competition performance (Crewther et al., 2013).

To date, few studies have examined and compared the acute hormonal responses with commonly prescribed RE protocols in rugby union players and no studies have investigated such responses to STRNG and COMB style training sessions despite their popular inclusion in rugby union training sessions. Therefore, the purpose of the current study was to examine the effects of four exercise protocols on salivary testosterone and cortisol concentrations in elite male rugby players. It was hypothesised that the testosterone response to each protocol would differ amongst each athlete, and that standard group data may not reflect individual response.

78 6.2 Methods

6.2.1 Subjects

Twenty-seven elite male rugby union players (age 28.3 ± 4.0 y, height 1.87 ± 0.08 m and mass 107.6 ± 18.9 kg) from a professional rugby union team volunteered for this study. This rugby team competed in both the English Rugby Union Premiership and European Championship, and a number of the subjects were also international level players. Ethics approval was obtained via a University Ethics committee.

6.2.2 Design

Four exercise protocols were superimposed onto the normal in-season training regimes of the athletes over four weeks. Each athlete completed all four protocols in random order. Two protocols incorporated RE, the third protocol was based on STRNG training and the fourth was a COMB inspired protocol based on boxing and wrestling exercises.

Saliva samples were collected before (PRE), and immediately after exercise (POST) for the analysis of testosterone and cortisol concentrations. Saliva was used due to its low invasiveness, and ability to track the biologically active “free” hormone (Crewther and Cook, 2010a).

All protocols were conducted mid-week and in the morning, when the athletes would normally have a gymnasium session. Each protocol was performed at the same time of day (10 am) to avoid the effects of circadian rhythm on hormonal concentrations. As this was an in-season professional rugby environment, there were practicalities of normal training that had to be adhered to. On each of the days immediately prior to testing, players were encouraged to sleep well (>7 h), consume a consistent breakfast across testing session days and maintain their fluid intake (at least 750 mL) during the two hours prior to each testing session. No food was consumed within 30 minutes of each testing session.

6.2.3 Hormone assessment

Saliva samples (approximately 2 mL) were collected by passive drool into sterile containers and stored at –20 °C (Granger et al., 2004) before analysis. The athletes provided a saliva sample prior to the commencement of each protocol (PRE) and immediately after the session (POST).

Saliva was assayed for testosterone (pg⋅mL-1) and cortisol (µg⋅dL-1) concentrations by a private commercial laboratory (HFL Sport Science Laboratories, UK) using commercially available kits (Salimetrics Enzyme Immunoassay Kits, Salimetrics, USA) according to the manufacturer’s

79 instructions. Inter-assay coefficients of variation (based on low- and high-control samples) for testosterone were 7.2% and 9.4% for cortisol.

6.2.4 Exercise protocols

The four protocols involved exercises that activated large muscle masses. The two RE protocols were based on those used in previous studies (Smilios et al., 2003, Beaven et al., 2008b). Each involved four exercises (high pull, bench press, squat and chin-ups/prone row). All exercises were performed using free weights and both sessions typically lasted 45 minutes in duration. The 5 × 15- 55% protocol consisted of 5 sets of 15 repetitions at 55% of the athletes’ 1 repetition maximum (1RM) with 1-minute rest between sets. The 3 × 5-85% protocol consisted of three sets of five repetitions at 85% 1RM with 2-minute rest between sets.

The STRNG protocol consisted of three stations within a circuit of exercises. Station 1 included high pulls (3 repetitions) and a heavy resisted horizontal sprinting sled (2 × 5 m; AssassinTM, UKSport, UK), Station 2 was battling ropes (20 repetitions completed as fast as possible) and Station 3 was a combination of a prowler push (2 × 6 m; 120 kg), farmer’s walk (2 × 6 m; weight each hand, 60 kg – forwards, 50 kg – backs) and tyre flip (2 × 2 repetitions: tyre standing height – 1.60 m and width – 0.61 m). Where possible power output (high pull – vertical transducer: Tendo Weightlifting Analyzer, Tendo Sport Machines, Trencin, Slovak Republic) and exercise time (i.e. AssassinTM and Station 3) were used to encourage competition within the group and create a competitive environment. Each athlete completed 4 circuits with 4-minute rest between circuits for a total training time of approximately 45 minutes.

The COMB protocol consisted of wrestling and boxing exercises. In groups of two, the athletes worked through six stations. Stations 1, 3 and 5 involved boxing using various combinations whilst the other player held a pad to absorb the punches. Stations 2, 4 and 6 were wrestling based and involved one player attempting to prevent their training partner from getting out of a particular position. Each group was encouraged to work as hard as they could for 60 seconds at each station and players swapped activities half way through each work set. Thirty-seconds rest was given between stations. Two-minutes rest was given after one circuit before completing a second and final circuit for a total training time of 19 minutes.

6.2.5 Statistical analyses

Hormone variables were log transformed to reduce bias due to non-uniformity of error. PRE and POST hormone concentrations for the four protocols (5 × 15–55%, 3 × 5–85%, STRNG and COMB) were analysed using an inferential approach (Hopkins, 2007). This analysis was also

80 conducted on a subgrouping of individuals categorised according to the protocol that demonstrated the greatest absolute PRE to POST increase in testosterone concentrations (Tmax). Selection into these subgroups was based on the principle that a maximal anabolic response was one of the objectives of the protocol and provided an optimal environment for adaption. Magnitudes of standardised effects were interpreted using thresholds of 0.2, 0.6 and 1.2 for small-, moderate- and large-effect sizes, respectively (Hopkins et al., 2009). Statistical significance was set at the P <.05 level. Percentage change in PRE to POST hormone concentration was calculated for use as a comparison against previous studies.

6.3 Results

6.3.1 Testosterone concentration - group response

The mean pooled testosterone concentration measured before each protocol was significantly different (p <.05) between the following protocols: 5 × 15–55% vs. 3 × 5–85% (ES = −0.73); 5×15- 55% vs COMB (ES = −0.63) and STRNG vs COMB (ES = −0.42). There were no statistically significant differences (p >.05) in the absolute change in PRE to POST testosterone concentrations and no differences were observed amongst protocols for POST testosterone concentrations (Figure 6.1).

Figure 6.1 Salivary testosterone concentrations before (PRE) and immediately after (POST) four exercise protocols (5 × 15–55%; 3 × 5–85%; strongman – STRNG; combative – COMB). Standard error of the difference is shown as error bars. *P < .05 vs PRE 5 × 15 concentration; **P <.05 vs PRE COMB concentration.

81 6.3.2 Testosterone concentration - individual response

Individuals were then categorised into subgroups according to the protocol that demonstrated the greatest absolute increase in PRE to POST testosterone concentration (Tmax). Significant absolute PRE to POST increases in testosterone concentrations were observed as a result of the 5 × 15–55% (p <.01; ES = 0.88), COMB (p <.001; ES =1.65) and STRNG (p <.01; ES = 0.75) protocols. No change was observed in the 3 × 5–85% protocol (Figure 6.2).

Figure 6.2 Salivary testosterone concentrations grouped according to the protocol that produced the largest absolute change before (PRE) and immediately after (POST) four exercise protocols (5 × 15–55%: n = 8; 3 × 5–85%: n = 2; strongman - STRNG: n = 6; combative - COMB: n = 11). Standard error of the difference is shown as an error bar. ** P <.01, *** P <.001 vs corresponding PRE value.

6.3.3 Cortisol concentration - group response

The mean group response for cortisol concentration is displayed in Table 6.1. Significant (p <.001) absolute PRE to POST decreases were observed as a result of the 5×15-55% (ES = −0.99) and 3×5- 85% (ES = −0.80) protocols but not the COMB or STRNG protocol. No change was observed in the STRNG protocol.

82 Table 6.1 Cortisol concentration and T/C ratio group response for each protocol.

Cortisol (µg⋅dL-1) T/C Ratio

PRE POST PRE POST

5 × 15–55% 0.35 ± 0.31 0.19 ± 0.17b 474 ± 227 870 ± 422d 3 × 5–85% 0.29 ± 0.18 0.20 ± 0.15b 456 ± 224 672 ± 357d STRNG 0.24 ± 0.21a 0.24 ± 0.15 610 ± 286c 589 ± 184 COMB 0.32 ± 0.27 0.27 ± 0.23 438 ± 184 560 ± 256 a P <.01 significantly different from PRE 5 × 15–55% (ES = −0.80). b P <.001 significantly different from PRE values. c P <.05 significantly different from PRE 5 × 15–55% (ES = −0.52), 3 × 5–85% (ES = −0.46). d P <.001 significantly different from PRE values.

6.3.4 T/C ratio concentration - group response

The mean group response for testosterone to cortisol (T/C) ratio is displayed in Table 6.1. Significant (p <.001) increases in absolute PRE to POST T/C ratio were observed in the 5 × 15– 55% (ES = 1.06) and 3 × 5–85% (ES = 0.69) protocols. No difference was observed in the STRNG and COMB protocols.

The percentage change in the mean group response for all PRE to POST hormone concentrations was calculated (Table 6.2). This calculation also included Tmax.

Table 6.2 Percentage change in PRE to POST hormone concentrations for each protocol

Tmax response Group response (n = 27)

Protocol Testosterone Testosterone (%) Cortisol (%) T/C ratio (%)

5 × 15–55% 60% (n = 8) 5 −45 83 3 × 5–85% 9% (n = 2) −8 −30 48 STRNG 23% (n = 6) 9 1 −3 COMB 56% (n = 11) 17 −16 28

Tmax – mean response of the subgroup of individuals who demonstrated the greatest absolute increase in PRE to POST testosterone concentration.

6.4 Discussion

An acute elevation in endogenous hormones offers short-term benefits for training performance and long-term adaptations (Ahtiainen et al., 2003). Resistance (Beaven et al., 2008b) and strongman training (Ghigiarelli et al., 2013) in particular have been shown to temporarily elevate testosterone concentration. Participants performed different exercise protocols (5 × 15–55%, 3 × 5–85%, STRNG and COMB) aiming to elicit a maximal anabolic response and provide an optimal 83 environment for adaption. The present study identified large-individual differences in the testosterone response to four exercise protocols.

Testosterone concentration - group response The group response to each protocol was small and not significant. The lack of change appears due to the large variability in the response to specific workout design, a feature noted in previous work (Beaven et al., 2008b). The 5 × 15–55% protocol used in the present study produced a non- significant acute testosterone increase (5%; p >.05). Using a similar protocol, Beaven et al. (2008b) reported similar findings (10.5% ± 19.7%; p >.05). The current 3 × 5–85% protocol produced a non-significant acute testosterone decrease (−8%; p >.05) which is of a magnitude similar to that in previous studies that utilised a similar protocol (Beaven et al., 2008b, Smilios et al., 2003). Consequently pooling the data for all of the athletes according to each protocol may not yield a true reflection of the individual response. Reanalysing this data by grouping individuals according to their highest response however gave a different interpretation. These results are reflective of those in elite athletic populations where case studies often yield better practical data than group based analyses.

Testosterone concentration - individual response Regrouping the results according to the protocols that demonstrated the greatest absolute increase in PRE to POST testosterone concentration highlighted three subgroups of significance (5 × 15–55%, STRNG and COMB). The changes in acute testosterone concentration for the 5 × 15–55% protocol is reflective of other work (Beaven et al., 2008b). Ghigiarelli et al. (2013) reported a 74% increase in testosterone concentration as a result of STRNG training workouts whereas this study only observed a 23% increase in testosterone concentration. The protocol utilised by Ghigiarelli et al. (2013) incorporated exercises performed to failure and they recruited athletes who were high school, collegiate and recreational standard. These factors could explain the differences seen in testosterone concentration (74% vs 23%) and provide further reason to the specificity of the training protocol to the individual (Beaven et al., 2008b).

The COMB session also demonstrated a significant increase in testosterone concentration (56%). To our knowledge this training format has not been studied using salivary hormones. However simulated fights in wrestlers that contained similar movement patterns have demonstrated significant elevations (up to 30%) in serum testosterone concentration (Kraemer et al., 2001a, Elias, 1981).

84 The 3 × 5–85% protocol produced a non-significant increase (9%) in testosterone concentration which was comparable (13%) to the change reported by Beaven et al. (2008b). However, only 2 players in the 3 × 5–85% protocol demonstrated the greatest absolute change in testosterone concentration. Performance in the 3 × 5–85% protocol may have been compromised from the preceding weekend’s game. Competitive rugby union results in significant muscular trauma (Gill et al., 2006) affecting the force generating capacity of muscles and hormonal concentrations (McLellan et al., 2011). On the other hand, the 5 × 15–55% and STRNG protocol used submaximal loads and the COMB incorporated bodyweight loadings. In an in-season format, the ability to maximise acute testosterone response from a training stimulus without using heaving loadings would be deemed advantageous.

Considerations to the results observed in this study include the strength training age and training experiences of the participants. The ability to increase strength levels is reportedly limited in well- trained strength athletes (Häkkinen et al., 1987). Therefore, the increases seen in the testosterone concentration (i.e. 5 × 15–55%, STRNG, COMB) may seem more modest relative to other studies using different athletic populations. The participants in this study were well-trained athletes who had been under the instruction of the same strength coach for the previous 4 years. The coaching methods and exercise selection had therefore been consistent throughout this period leading into the study eliminating any hormonal responses being a result of unfamiliarity. Furthermore, the variations in the hormonal responses may simply be a result of the individual’s distinct response to the protocol variables (Beaven et al., 2008b).

Cortisol concentration - group response The current study demonstrated significant decreases in salivary cortisol concentration as a result of the two RE protocols (5 × 15–55%, 45% and 3 × 5–85%, 30%). These findings for both protocols are comparable to a similar study of rugby union players (Beaven et al., 2008b). Smilios et al. (2003) reported a 22% decrement in serum cortisol utilizing a 3 × 5–85% protocol whereas in the same study a 5×15-60% protocol demonstrated a 27% increase in serum cortisol (Smilios et al., 2003).

The STRNG and COMB protocol produced no change in cortisol concentration. Hormonal response to STRNG-based training is not well studied and to our knowledge this is the first study to report on the effects of this type of training on salivary cortisol concentration. Cortisol responses to competitive wrestling bouts are equivocal (Elias, 1981) with the length of such bouts being much shorter than the COMB circuit used in this study (3–5 minutes vs 19 minutes).

85 T/C ratio concentration – group response Some studies have suggested the use of the T/C ratio as an indicator of a positive anabolic state (Banfi, 1998). Hence, speculatively, it is possible that an increase in the T/C ratio during the 5 × 15–55% (83%) and 3 × 5–85% (48%) protocols indicates that players’ response was anabolic, but this of course may simply be a reflection of the change in cortisol concentration. If so, this may support the observation of no change in the T/C ratio for both the STRNG and COMB protocol.

Furthermore, the hormonal response to exercise may not only be content driven but also be derived from psychological (Archer, 2006, Archer, 1991), social (Archer, 1991), and competitive (Elias, 1981) factors. The athletes involved in this study were well trained and intrinsically motivated individuals. Consequently the training environment created by these athletes may have affected the hormonal responses observed in this study. The weight training sessions were conducted mid-week, between 72 to 96 hours post match. The natural endocrine response to competitive rugby games typically demonstrates a return to values similar to that of pregame concentrations within 38 to 96 hours post game (Cunniffe et al., 2011, Elloumi et al., 2003) speculating that full endocrine recovery had occurred prior to the training protocols. It must also be noted that there may have been a possible cumulative effect of games for particular individuals. Irrespective valuable descriptive data of elite athletes under real competitive situations have been obtained.

Similar to the speculation of Beaven et al. (2008b), weight training may potentially be viewed as a “time-out” from the rigours of rugby specific contact training. This could affect hormonal stress responses relative to other athlete populations, which may perceive an equivalent workload of weight training differently (Beaven et al., 2008b). The novelty of both the STRNG and COMB protocol in conjunction with the associated enjoyment and variance of training stimulus (e.g. from standard resistance training) may have been another factor in the acute hormonal elevation observed in the present study.

The outcomes of the present study support the findings of Beaven et al. (2008b) in identifying the “individual” RE induced hormonal responses in elite rugby union players. The observed acute testosterone concentrations across the training sessions may be reflective of the behaviour modifying properties of testosterone, in particular with respect to increased motivation (Aarts and van Honk, 2009). Increases in testosterone concentrations through endogenous observations (Cook and Beaven, 2013) and exogenous administration (Aarts and van Honk, 2009) have both demonstrated improvements in motivational measures. This is further supported by West and Phillips (2012) who proposed that acute post-exercise increases in systemic hormones are not a

86 proxy marker for anabolism as they do not underpin the capacity of the muscle to hypertrophy in any measureable way.

The findings from this study also support the use of STRNG and COMB training protocols as an alternative stimulus to resistance training when viewing exercise protocols as a method to elevate endogenous hormones. This work’s application lies within how training workouts affect acute hormone concentrations and consequential behaviours and competition outcome. This is an important step in creating a training culture that is geared towards optimising both training and competitive performance.

6.5 Practical applications

Elevated hormonal concentrations have many proposed performance benefits both in training and competition. Resistance training is well regarded as a method to elicit a desirable elevation in hormonal response. In a season long competition format, the ability to utilise multiple training formats that are behaviourally beneficial (through the elevation of hormonal response) becomes important. The results observed in this study highlight the potential usefulness of employing STRNG and COMB (boxing and wrestling) training protocols in conjunction with appropriate RE protocols as an effective training method for rugby union players. However further investigation is required to demonstrate if the hormonal changes resulting from such exercise protocols produce long term functional gains and influence athletic performance.

87 7. General Discussion

7.1 Summary and discussion

7.1.1 Assessing the measures of salivary testosterone and cortisol concentrations according to game outcome

Pregame testosterone concentrations in elite male rugby players were significantly higher before games that were won than before games that were lost. A game-ranked performance score (1 to 6) also strongly correlated (r = .81) with the team’s pregame testosterone concentrations. When the data was analysed by playing position (forwards vs backs), significant relationships were observed between the game-ranked score and both pregame testosterone (r = .91) and the T/C ratio (r = .81) in the backs only. Pregame testosterone as a predictor of team outcome in sport has, to our knowledge, not been presented previously. The rise between pre and post competition samples has previously been reported as an indicator of success; however, most of these studies were generally recorded across a single event or game. The present series of studies, however, were able to follow the same team across 6 consecutive games. High endogenous testosterone concentrations are associated with superior physical qualities of strength, speed and power (Crewther et al., 2012) in addition to dominant and aggressive behaviours (Mazur and Booth, 1998). The demands of competitive rugby union would highlight that optimising these physical qualities and behaviours would be of an advantage (Duthie et al., 2003). Therefore, ensuring elevated endogenous testosterone concentration the day of the game presents an opportunity to give the players a platform for a more consistent and favourable performance. Priming strategies through specific physical (e.g. strength training) and passive (e.g. pre-match talks, video footage) interventions implemented prior to competition could elevate testosterone concentrations and lead to performance benefits in rugby union (Kilduff et al., 2013).

The relationship between the T/C ratio and game outcome in the backs is the first study to demonstrate a predictive ability for outcome in team sport. A number of studies have reported associations between the T/C ratio and actions linked to performance outcomes; however, there has been no demonstrated relationship with performance outcome. In social studies, a higher T/C ratio is associated with heightened cognitive decision making, motivational and behavioural tendencies (Terburg et al., 2009). These are also desirable traits for success in rugby union. A higher T/C ratio could also be interpreted as a marker of physical readiness and ‘positive consequence of the total training’ (Vervoorn et al., 1991) which are both important for game performance. Therefore, a 88 higher T/C ratio may have been a result of the improved balance between training and recovery leading into the game. The physiological demands of a game between backs and forwards differ (Cunniffe et al., 2009). Hence, the lack of relationship observed in the forwards may be due to inadequate recovery from the previous week’s game and training. Alternatively, it may reflect an individual hormone response associated with a stress (Crewther et al., 2012). Coach interaction with players prior to a game through tactical and motivational pre-match talks can also affect acute testosterone concentrations (Cook and Crewther, 2012b). Different coaches are responsible for the backs and forwards and therefore the coach’s interaction (i.e. pre-match talk) with the player groups prior to the game may have been a casual factor in the observed hormonal differences.

Individualising weekly training load and the employment of position-specific recovery techniques based on regular monitoring may assist with game-day readiness. Furthermore, the findings from Cook and Crewther (2012b) highlights the potential role of the coach’s language and actions on game day with respect to how they may affect a player’s circulating hormone concentration. In rugby union, this study comprising Chapter 3 is the first to demonstrate that pregame testosterone concentration and pregame T/C ratio can predict competitive success. It also highlights the variability of acute hormonal response and the need to provide an individualised approach to recovery, training and game preparation plans for athletes according to their positional requirements.

7.1.2 Determining if performance indicator scores discriminate between game outcomes and correlate with game-day salivary hormone concentrations

Across a full of season of 31 elite professional rugby union games, an aggressive-based team performance indicator (AggPI) score was shown to differentiate between game outcomes (i.e. win- loss). The performance variables contributing to the calculation of the AggPI score required the player to perform actions of dominance and aggression whilst either carrying the ball into contact or alternatively trying to get the ball off the opposition. In a small subset of these games, strong correlations were observed between AggPI and pre-match testosterone concentrations (r = .86), peri-match testosterone concentrations (r = −.87) and pre-match T/C ratio (r = .86). A skill-based performance indicator (SkillPI) was also calculated. Although SkillPI was not predictive of game outcome, a team score demonstrated a strong relationship with peri-game cortisol (r = –.98).

Rugby union studies investigating the role of performance parameters and outcome have traditionally focused on single-skill components (e.g. effectiveness in territorial advances) as causal factors for successful outcome (Jones et al., 2004, Ortega et al., 2009, Vaz et al., 2010). The

89 advantage of a performance indicator (PI) score is that it encompasses multiple skills and actions that are fundamental to rugby union as opposed to just one predictive marker of performance. It could be suggested that a PI score would be more robust across a season of games due to being able to encompass the variations in playing different opponents weekly (e.g. opposition game plans, home vs away, refereeing, weather). To our knowledge, this is the first study in rugby union that has identified a single team-based PI score to discriminate between a win and a loss. The relationship between pregame testosterone concentrations and AggPI are consistent with documented theories regarding the association between testosterone and aggressive behaviours in human competitive encounters (Bouissou, 1983, Dixson, 1980), which are desirable behaviours in rugby union. The inverse relationship between peri testosterone and AggPI may primarily be due to the observed pregame testosterone concentrations. In Chapter 3, pregame testosterone concentrations discriminated between game outcomes unlike postgame testosterone and peri testosterone concentrations (Gaviglio et al., 2014). As a higher AggPI score is reflective of a win the positive relationship between pregame testosterone and AggPI fits well with these observations, however the inverse relationship observed between peri testosterone and AggPI does not. Therefore, it would be plausible that it was the pregame testosterone concentration as the major contributor to this observation. The T/C ratio has also been suggested as an interactive factor in human aggression, which supports the observed results with testosterone (Carré and Mehta, 2011, Terburg et al., 2009, Kuoppasalmi et al., 1980, Bosco et al., 1996a, Cardinale and Stone, 2006, Kraemer et al., 2004, Häkkinen et al., 1987). The T/C ratio may also be an indicator of recovery (Banfi et al., 1993, Vervoorn et al., 1991) or readiness to compete (Doan et al., 2007). The relationships above confirm the aggressive nature of rugby union and the complex performance variables contributing to successful outcome. The relationship between AggPI and hormone concentration further supports the concept of elevating testosterone concentrations pregame with respect to improving the chance of a successful competitive outcome.

7.1.3 Evaluating mid-week salivary hormone concentrations and subsequent match outcome

In elite professional rugby union players, the T/C ratio and salivary testosterone concentrations increased significantly in response to a midweek skill-based session prior to a win but did not increase prior to a loss. In addition, the percent change in the midweek T/C ratio and cortisol concentrations demonstrated strong correlations with subsequent pregame testosterone concentrations. Furthermore, the T/C ratio prior to a midweek skill training session was significantly lower preceding winning games than preceding games that were lost. Mid-week values of the T/C ratio as a predictor of subsequent match outcome in rugby union have not been presented previously. A mid-week physical stress test in rugby league (resistance training workout) has also 90 demonstrated significant relationships with game outcome 3 to 4 days later (Crewther et al., 2013). In rugby league, the absolute and relative elevations in mid-week salivary testosterone concentrations were significant only prior to games won. Crewther et al. (2017) observed that in rugby union an anticipatory increase in mid-week salivary cortisol concentrations between morning and a physical stress test (repeat shuttle run test) was observed prior to winning and cortisol decreased before games lost. With respect to the repeat shuttle run test, they observed no relationship between changes in pre-to-post hormone concentrations and match outcome. These observed differences highlight the individual response of hormones to a stress and may reflect the different team culture and environments. The relationship between the mid-week hormone concentrations and pregame testosterone concentrations further highlights the interaction between salivary hormones and competitive performance. In a team sport in which skill-based training sessions form a large portion of a training week, the observations from this study may help to better understand the overall response to training sessions and their impact on competition results a few days later. This information has therefore highlighted an opportunity to use salivary hormones as a marker to identify competition readiness 3 to 4 days prior to a game.

7.1.4 Evaluating the acute response of salivary testosterone and cortisol concentrations to four rugby training exercise protocols

The acute response of salivary hormones to four exercise protocols was examined. Two protocols were resistance training based (5 × 15–55% 1RM and 3 × 5–85% 1RM), while the other two sessions were based around strongman (STRNG) and combative (COMB) training (e.g. boxing and wrestling). There was no difference in the average testosterone concentration between any intervention protocol; however, cortisol concentration declined and the T/C ratio increased significantly in both of the resistance-training protocols. When data were grouped and analysed according to the specific protocol that demonstrated the greatest absolute increase in testosterone concentration for each athlete, significant increases (p <.01–.001) for the 5 × 15–55% (↑ 60%), STRNG (↑ 23%), and COMB (↑ 56%), protocols were observed. Although the hormonal response to resistance training has been well studied, there is limited research on the response to strongman training and no research on combative-based training despite their popularity in the training environment. The results observed in this study support the notion that the adaptation to any mode of training is highly individualised. This finding is important in determining a program that is geared towards optimising both training effectiveness and competitive performance.

91 7.2 Practical applications

Salivary hormones are a suitable biomarker for evaluating training response and assessing competition readiness in rugby players. Therefore, routine use of salivary hormones could help to better understand the acute effects of a training stimulus on an individual, and the interactive factors associated with training and competition. Thus, the research in this thesis led to development of the following recommendations.

1. Pregame salivary testosterone concentration predicts competition success in an aggressive, contact-based professional team sport. Incorporating specific physical and psychological interventions can elevate acute pre-match testosterone concentrations. Therefore optimising game-day routines by incorporating these interventions and validating through saliva hormone analysis could assist with improving competition readiness and performance.

2. The forwards did not demonstrate any relationship between game performance and salivary hormone concentration. The difference in positional requirements and physiological demands during a game between forwards and backs may reflect the individual hormonal response to a stress or inadequate recovery from the previous game and training performed during the week. Thus, position specific recovery protocols and monitoring of response during the training week need to be developed.

3. The determination of mid-week salivary hormone concentrations before and after a skill- based training session could be used as a tool for assessing game readiness in professional rugby union.

4. A match-based team statistic score can discriminate between game outcome. Therefore, it is important to monitor the specific skills that comprise this PI score to identify any technical or execution errors under competition stress. Thus, during the training week, coaching of these identified skills could be a focus with the goal to improve skill execution during subsequent games. Tracking these parameters in real-time during a game could assist the coach with providing more pertinent and specific feedback to the players during the game.

5. During a competition season, physical qualities are maintained through performing strength and conditioning programs. For elite male rugby players, the response to training is highly individualised. Therefore, identifying optimal strength and conditioning programs through saliva hormone analysis for each player is important. More specifically, these results demonstrate the usefulness of employing strongman and combative (boxing and wrestling)

92 training protocols in conjunction with appropriate resistance exercise protocols as an effective training method for rugby union players.

7.3 Limitations and delimitations

1. In Study 1 and Study 2, data were collected for only 6 weeks. Although this timeframe is comparable to those seen in studies in other sporting contexts, sampling across a longer timeframe would increase the statistical strength of the interpretation and likely reduce the large variability observed in testosterone measurements.

2. A limitation in Study 2 was that only one team was investigated, albeit across an entire season. Therefore, the applied scoring system may only be relevant for this team.

3. The exercise protocols in Study 4 did not account for variations in training time, volume, loadings and type of exercises. Nevertheless, these protocols are common in training sessions used by rugby players and provide valuable real-world descriptive data.

7.4 Directions for future research

This thesis has demonstrated the usefulness of measuring the acute changes in testosterone and cortisol concentrations before and in response to training and competition. The results indicate several avenues for future research.

1. Investigate various physical and psychological interventions (e.g. meeting content, use of video, different coaching styles, session structure) and the subsequent effect on acute hormone response.

2. Examine how various physical and psychological stimuli affect performance parameters, using salivary hormones as markers of the effectiveness of the stimuli. An understanding of the acute and interactive effects of the stimuli would provide potential avenues for improving the training content and structure of the training week and game day performance.

3. Further examine the same performance-indicator scoring system with other professional rugby teams to confirm its validity as a suitable performance rating system. Given that testosterone concentration can be increased by physical and psychological methods, it is also necessary to determine if an acute pre-match elevation in testosterone concentration by a particular physical or psychological intervention positively affects the score of the aggressive performance indicator, and consequent game performance and outcome.

93 4. Further research is required to investigate if the training interventions as used in Study 4 could produce differences in functional gain and subsequent improvement in competitive performance outcomes throughout the course of a season or an athlete’s career.

5. Further studies incorporating serial measures may demonstrate a stronger predictive relationship between hormone response to training at different times throughout the week and subsequent competition success. This could assist with defining a more individualised approach to training and competition performance.

6. Develop a cost-effective and rapid method for assessing salivary testosterone concentration. Such a method would provide an opportunity to regularly monitor acute testosterone responses to training and competition. This would enable coaching staff to implement strategies (both physical and psychological) to optimise performance during the training week and on competition day. Furthermore, this would circumvent the need to identify a suitable proxy for measuring testosterone concentration and the issues surrounding the validation of such methods.

7.5 Conclusion

The research presented in this thesis provides new evidence about the short-term relationship between testosterone and cortisol concentrations and competition and training in elite male rugby players. Acute changes in salivary hormone concentration in response to competitive game play of elite rugby players were related to objective and subjective measures of game outcome. Subsequent testing demonstrated that the response of mid-week hormone concentrations to a skill-based training session was also a reliable predictor of successful game outcome 3 to 4 days later. These findings support the importance of considering the relationship between hormone concentration and athletic performance throughout the entire training week, rather than focusing on the competition day alone. Further investigations reported in this thesis aimed to understand the relationship between salivary hormones and routine training practices used in a professional rugby union team environment. Players performing four rugby training exercise protocols demonstrated highly individualised responses in terms of salivary hormones.

The results of the research presented in this thesis support the use of salivary hormones as a reliable biomarker to monitor the acute response of elite male rugby union players to training and competition. Although this thesis has reported on the hormonal response to physical interventions, the research suggests that psychological methods could also improve performance outcomes through the modification of desirable behaviours (e.g. mood, cognitive function, motivation), which

94 influence game-day performance. Through practical interventions, these strategies could continue to shape and refine coaching practices for improved human athletic performance.

95 8. Reference List

Aarts, H. and J. van Honk 2009. Testosterone and unconscious positive priming increase human motivation separately. Neuroreport, 20, 1300-1303.

Adlercreutz, H., H. M., K. Kuoppasalmi, N. H., I. Huhtaniemi, H. Tikkanen, K. Remes, A. Dessypris and J. Karvonen 1986. Effect of training on plasma anabolic and catabolic steroid hormones and their response during physical exercise. International Journal of Sports Medicine, 7, 27-28.

Ahtiainen, J. P., A. Pakarinen, M. Alen, W. J. Kraemer and K. Hakkinen 2003. Muscle hypertrophy, hormonal adaptations and strength development during strength training in strength-trained and untrained men. European Journal of Applied Physiology, 89, 555-63.

Ahtiainen, J. P., A. Pakarinen, W. J. Kraemer and K. Häkkinen 2004. Acute hormonal responses to heavy resistance exercise in strength athletes versus nonathletes. Canadian Journal of Applied Physiology, 29, 527-43.

Aikey, J. L., J. G. Nyby, D. M. Anmuth and P. J. James 2002. Testosterone rapidly reduces anxiety in male house mice (Mus musculus). Hormones and Behavior, 42, 448-60.

Aleman, A., E. Bronk, R. P. Kessels, H. P. Koppeschaar and J. van Honk 2004. A single administration of testosterone improves visuospatial ability in young women. Psychoneuroendocrinology, 29, 612-7.

Anttila, K., S. Manttari and M. Jarvilehto 2008. Testosterone and Ca2+ regulation in skeletal muscle. International Journal of Sports Medicine, 29, 795-802.

Archer, J. 1991. The influence of testosterone on human aggression. British Journal of Psychology, 82 ( Pt 1), 1-28.

Archer, J. 2006. Testosterone and human aggression: an evaluation of the challenge hypothesis. Neuroscience and Biobehavioral Reviews, 30, 319-45.

Argus, C. K., N. D. Gill, J. W. Keogh, W. G. Hopkins and C. M. Beaven 2009. Changes in strength, power, and steroid hormones during a professional rugby union competition. Journal of Strength and Conditioning Research, 23, 1583-92.

96 Arnold, A. P., F. Nottebohm and D. W. Pfaff 1976. Hormone concentrating cells in vocal control and other areas of the brain of the zebra finch (Poephila guttata). Journal of Comparative Neurology, 165, 487-511.

Aronson, P. S., W. F. Boron, E. L. Boulpaep, W. F. Boron and E. L. Boulpaep 2003. Medical Physiology. A Cellular and Molecular Approach. 1049-1065.

Arregger, A. L., L. N. Contreras, O. R. Tumilasci, D. R. Aquilano and E. M. Cardoso 2007. Salivary testosterone: a reliable approach to the diagnosis of male hypogonadism. Clinical Endocrinology, 67, 656-62.

Arruda, A. F., M. S. Aoki, C. G. Freitas, L. M. Spigolon, C. Franciscon and A. Moreira 2015. Testosterone Concentration and Lower Limb Power Over an Entire Competitive Season in Elite Young Soccer Players. Journal of Strength and Conditioning Research, 29, 3380-5.

Banfi, G. 1998. Hormonal ratio (cortisol:testosterone) as a marker of exhaustion in athletes. Journal of Sports Medicine and Physical Fitness, 38, 91-92.

Banfi, G., M. Marinelli, G. S. Roi and V. Agape 1993. Usefulness of free testosterone/cortisol ratio during a season of elite speed skating athletes. International Journal of Sports Medicine, 14, 373-9.

Baxendale, P., M. Reed and V. James 1980. Testosterone in saliva of normal men and its relationship with un-bound and total testosterone levels in plasma. Journal of Endocrinology, 87, 46-47.

Beaven, C. M., C. J. Cook and N. D. Gill 2008a. Significant strength gains observed in rugby players after specific resistance exercise protocols based on individual salivary testosterone responses. Journal of Strength and Conditioning Research, 22, 419-25.

Beaven, C. M., N. D. Gill and C. J. Cook 2008b. Salivary testosterone and cortisol responses in professional rugby players after four resistance exercise protocols. Journal of Strength and Conditioning Research, 22, 426-32.

Bell, G. J., D. Syrotuik, T. P. Martin, R. Burnham and H. A. Quinney 2000. Effect of concurrent strength and endurance training on skeletal muscle properties and hormone concentrations in humans. European Journal of Applied Physiology, 81, 418-27.

Berchtold, M. W., H. Brinkmeier and M. Muntener 2000. Calcium ion in skeletal muscle: its crucial role for muscle function, plasticity, and disease. Physiological Reviews, 80, 1215-65. 97 Bhasin, S., L. Woodhouse and T. W. Storer 2001. Proof of the effect of testosterone on skeletal muscle. Journal of Endocrinology, 170, 27-38.

Bishop, D. 2008. An applied research model for the sport sciences. Sports Medicine, 38, 253-63.

Blanco, C. E., P. Popper and P. Micevych 1997. Anabolic-androgenic steroid induced alterations in choline acetyltransferase messenger RNA levels of spinal cord motoneurons in the male rat. Neuroscience, 78, 873-82.

Bleisch, W. V., A. L. Harrelson and V. N. Luine 1982. Testosterone increases acetylcholine receptor number in the “levator ani” muscle of the rat. Journal of Neurobiology, 13, 153-161.

Booth, A., G. Shelley, A. Mazur, G. Tharp and R. Kittok 1989. Testosterone, and winning and losing in human competition. Hormones and Behavior, 23, 556-71.

Borg, G. A. 1982. Psychophysical bases of perceived exertion. Medicine and Science in Sports and Exercise, 14, 377-81.

Bosco, C., J. Tihanyit and A. Viru 1996a. Relationships between field fitness test and basal serum testosterone and cortisol levels in soccer players. Clinical Physiology, 16, 317-22.

Bosco, C., J. Tihanyl, L. Rivalta, G. Parlato, C. Tranquilli, G. Pulvirenti, C. Foti, M. Viru and A. Viru 1996b. Hormonal responses in strenuous jumping effort. Japanese Journal of Physiology, 46, 93-8.

Bouissou, M. F. 1983. Androgens, aggressive behaviour and social relationships in higher mammals. Hormone Research, 18, 43-61.

Bracewell, P. 2002. Implementing Statistics in a Diagnostic Coaching Structure for Rugby. Research Letters in the Information and Mathematical Sciences, 3, 79-84.

Brownlee, K. K., A. W. Moore and A. C. Hackney 2005. Relationship between circulating cortisol and testosterone: influence of physical exercise. Journal of Sports Science & Medicine, 4, 76-83.

Budgett, R. 1998. Fatigue and underperformance in athletes: the overtraining syndrome. British Journal of Sports Medicine, 32, 107-10.

Burnstein, K. L., C. A. Maiorino, J. L. Dai and D. J. Cameron 1995. Androgen and glucocorticoid regulation of androgen receptor cDNA expression. Molecular and Cellular Endocrinology, 115, 177-86. 98 Buss, C., O. T. Wolf, J. Witt and D. H. Hellhammer 2004. Autobiographic memory impairment following acute cortisol administration. Psychoneuroendocrinology, 29, 1093-6.

Cadore, E., F. Lhullier, M. Brentano, E. Silva, M. Ambrosini, R. Spinelli, R. Silva and L. Kruel 2008. Correlations between serum and salivary hormonal concentrations in response to resistance exercise. Journal of Sports Sciences, 26, 1067-72.

Cairns, S. P., A. J. Knicker, M. W. Thompson and G. Sjøgaard 2005. Evaluation of models used to study neuromuscular fatigue. Exercise and Sport Sciences Reviews, 33, 9-16.

Cardinale, M. and M. H. Stone 2006. Is testosterone influencing explosive performance? Journal of Strength and Conditioning Research, 20, 103-7.

Carré, J. M. 2009. No place like home: testosterone responses to victory depend on game location. American Journal of Human Biology, 21, 392-394.

Carré, J. M. and P. H. Mehta 2011. Importance of considering testosterone-cortisol interactions in predicting human aggression and dominance. Aggressive Behavior, 37, 489-91.

Carré, J. M., C. Muir, J. Belanger and S. K. Putnam 2006. Pre-competition hormonal and psychological levels of elite hockey players: relationship to the ‘home advantage’. Physiology and Behavior, 89, 392-398.

Chen, S., J. Wang, G. Yu, W. Liu and D. Pearce 1997. Androgen and glucocorticoid receptor heterodimer formation. A possible mechanism for mutual inhibition of transcriptional activity. Journal of Biological Chemistry, 272, 14087-92.

Coates, J. M., M. Gurnell and Z. Sarnyai 2010. From molecule to market: steroid hormones and financial risk-taking. Philosophical Transactions of the Royal Society of London. Series B: Biological Sciences, 365, 331-43.

Coffey, D. S. 1988. Androgen action and the sex accessory tissues. The Physiology of Reproduction, 1, 1081-1120.

Cook, C. J. and C. M. Beaven 2013. Salivary testosterone is related to self-selected training load in elite female athletes. Physiology and Behavior, 116-117, 8-12.

99 Cook, C. J. and B. T. Crewther 2012a. Changes in salivary testosterone concentrations and subsequent voluntary squat performance following the presentation of short video clips. Hormones and Behavior, 61, 17-22.

Cook, C. J. and B. T. Crewther 2012b. The effects of different pre-game motivational interventions on athlete free hormonal state and subsequent performance in professional rugby union matches. Physiology and Behavior, 106, 683-8.

Cook, C. J. and B. T. Crewther 2014. The social environment during a post-match video presentation affects the hormonal responses and playing performance in professional male athletes. Physiology and Behavior, 130, 170-5.

Cook, C. J., B. T. Crewther and L. P. Kilduff 2013. Are free testosterone and cortisol concentrations associated with training motivation in elite male athletes? Psychology of Sport and Exercise, 14, 882-885.

Copeland, J. L., L. A. Consitt and M. S. Tremblay 2002. Hormonal responses to endurance and resistance exercise in females aged 19-69 years. Journals of Gerontology. Series A: Biological Sciences and Medical Sciences, 57, B158-65.

Crewther, B., J. Carruthers, L. Kilduff, C. Sanctuary and C. Cook 2016. Temporal associations between individual changes in hormones, training motivation and physical performance in elite and non-elite trained men. Biology of Sport, 33, 215-21.

Crewther, B., T. Lowe, J. Ingram and R. P. Weatherby 2010. Validating the salivary testosterone and cortisol concentration measures in response to short high-intensity exercise. Journal of Sports Medicine and Physical Fitness, 50, 85-92.

Crewther, B. T. 2009. New evidence for the short-term effects of testosterone and cortisol upon athletic performance and training adaptations in elite male rugby players. Doctor of Philosophy, Southern Cross University.

Crewther, B. T. and C. Cook 2010a. Measuring the salivary testosterone and cortisol concentrations of weightlifters using an enzyme-immunoassay kit. International Journal of Sports Medicine, 31, 486-9.

Crewther, B. T., C. Cook, M. Cardinale, R. P. Weatherby and T. Lowe 2011. Two emerging concepts for elite athletes: the short-term effects of testosterone and cortisol on the neuromuscular

100 system and the dose-response training role of these endogenous hormones. Sports Medicine, 41, 103-23.

Crewther, B. T. and C. J. Cook 2010b. Relationships between salivary testosterone and cortisol concentrations and training performance in Olympic weightlifters. Journal of Sports Medicine and Physical Fitness, 50, 371-5.

Crewther, B. T. and C. J. Cook 2012. Effects of different post-match recovery interventions on subsequent athlete hormonal state and game performance. Physiology and Behavior, 106, 471-5.

Crewther, B. T., C. J. Cook, C. M. Gaviglio, L. P. Kilduff and S. Drawer 2012. Baseline strength can influence the ability of salivary free testosterone to predict squat and sprinting performance. Journal of Strength and Conditioning Research, 26, 261-8.

Crewther, B. T., J. Cronin, J. Keogh and C. Cook 2008. The salivary testosterone and cortisol response to three loading schemes. Journal of Strength and Conditioning Research, 22, 250-5.

Crewther, B. T., J. Keogh, J. Cronin and C. Cook 2006. Possible stimuli for strength and power adaptation: acute hormonal responses. Sports Medicine, 36, 215-38.

Crewther, B. T., T. Lowe, R. P. Weatherby and N. Gill 2009a. Prior sprint cycling did not enhance training adaptation, but resting salivary hormones were related to workout power and strength. European Journal of Applied Physiology, 105, 919-927.

Crewther, B. T., T. Lowe, R. P. Weatherby, N. Gill and J. Keogh 2009b. Neuromuscular performance of elite rugby union players and relationships with salivary hormones. Journal of Strength and Conditioning Research, 23, 2046-53.

Crewther, B. T., N. Potts, L. P. Kilduff, S. Drawer and C. J. Cook 2017. Can salivary testosterone and cortisol reactivity to a mid-week stress test discriminate a match outcome during international rugby union competition? Journal of Science and Medicine in Sport.

Crewther, B. T., C. E. Sanctuary, L. P. Kilduff, J. S. Carruthers, C. M. Gaviglio and C. J. Cook 2013. The workout responses of salivary-free testosterone and cortisol concentrations and their association with the subsequent competition outcomes in professional rugby league. Journal of Strength and Conditioning Research, 27, 471-6.

Cumming, D. C., M. E. Quigley and S. S. Yen 1983. Acute suppression of circulating testosterone levels by cortisol in men. Journal of Clinical Endocrinology and Metabolism, 57, 671-3. 101 Cunniffe, B., A. J. Hore, D. M. Whitcombe, K. P. Jones, B. Davies and J. S. Baker 2011. Immunoendocrine responses over a three week international rugby union series. Journal of Sports Medicine and Physical Fitness, 51, 329-38.

Cunniffe, B., W. Proctor, J. S. Baker and B. Davies 2009. An evaluation of the physiological demands of elite rugby union using Global Positioning System tracking software. Journal of Strength and Conditioning Research, 23, 1195-203.

Dabbs, J. M. 1990. Salivary testosterone measurements: reliability across hours, days, and weeks. Physiology and Behavior, 48, 83-6.

Dabbs, J. M. 1993. Salivary testosterone measurements in behavioral studies. Annals of the New York Academy of Sciences, 694, 177-83.

Dabbs, J. M., G. J. Jurkovic and R. L. Frady 1991. Salivary testosterone and cortisol among late adolescent male offenders. Journal of Abnormal Child Psychology, 19, 469-478.

Davis, J. K. and J. M. Green 2009. Caffeine and anaerobic performance: ergogenic value and mechanisms of action. Sports Medicine, 39, 813-32.

Deutsch, M. U., G. J. Maw, D. Jenkins and P. Reaburn 1998. Heart rate, blood lactate and kinematic data of elite colts (under-19) rugby union players during competition. Journal of Sports Sciences, 16, 561-70.

Dixson, A. F. 1980. Androgens and aggressive behavior in primates: A review. Aggressive Behavior, 6, 37-67.

Doan, B. K., R. U. Newton, W. J. Kraemer, Y. H. Kwon and T. P. Scheet 2007. Salivary cortisol, testosterone, and T/C ratio responses during a 36-hole competition. International Journal of Sports Medicine, 28, 470-9.

Dubois, R., T. Paillard, M. Lyons, D. McGrath, O. Maurelli and J. Prioux 2017. Running and Metabolic Demands of Elite Rugby Union Assessed Using Traditional, Metabolic Power, and Heart Rate Monitoring Methods. Journal of Sports Science & Medicine, 16, 84-92.

Dunn, J. F., B. C. Nisula and D. Rodbard 1981. Transport of steroid hormones: binding of 21 endogenous steroids to both testosterone-binding globulin and corticosteroid-binding globulin in human plasma. Journal of Clinical Endocrinology and Metabolism, 53, 58-68.

102 Duthie, G., D. Pyne and S. Hooper 2003. Applied physiology and game analysis of rugby union. Sports Medicine, 33, 973-91.

Eik-Nes, K. B. 1964. Effects of Gonadotrophins on Secretion of Steroids by the Testis and Ovary. Physiological Reviews, 44, 609-30.

Ekblom, B. 1986. Applied physiology of soccer. Sports Medicine, 3, 50-60.

Eliakim, A., S. Portal, Z. Zadik, J. Rabinowitz, D. Adler-Portal, D. M. Cooper, F. Zaldivar and D. Nemet 2009. The effect of a practice on anabolic hormones and inflammatory markers in elite male and female adolescent players. Journal of Strength and Conditioning Research, 23, 1553- 9.

Elias, M. 1981. Serum cortisol, testosterone, and testosterone-binding globulin responses to competitive fighting in human males. Aggressive Behavior, 7, 215-224.

Elloumi, M., O. Ben Ounis, Z. Tabka, E. Van Praagh, O. Michaux and G. Lac 2008. Psychoendocrine and physical performance responses in male Tunisian rugby players during an international competitive season. Aggressive Behavior, 34, 623-32.

Elloumi, M., F. Maso, O. Michaux, A. Robert and G. Lac 2003. Behaviour of saliva cortisol [C], testosterone [T] and the T/C ratio during a rugby match and during the post-competition recovery days. European Journal of Applied Physiology, 90, 23-8.

Enoka, R. 2002. The motor system: multi-joint systems: chapter 7. Neuromechanics of Human Movement. 3rd ed. Champaign (IL): Human Kinetics, 313-58.

Estrada, M., A. Espinosa, M. Muller and E. Jaimovich 2003. Testosterone stimulates intracellular calcium release and mitogen-activated protein kinases via a G protein-coupled receptor in skeletal muscle cells. Endocrinology, 144, 3586-97.

Farrell, P. A., T. L. Garthwaite and A. B. Gustafson 1983. Plasma adrenocorticotropin and cortisol responses to submaximal and exhaustive exercise. Journal of Applied Physiology, 55, 1441-4.

Fiers, T., J. Delanghe, G. T'Sjoen, E. Van Caenegem, K. Wierckx and J. M. Kaufman 2014. A critical evaluation of salivary testosterone as a method for the assessment of serum testosterone. Steroids, 86, 5-9.

103 Filaire, E., D. Alix, C. Ferrand and M. Verger 2009. Psychophysiological stress in tennis players during the first single match of a tournament. Psychoneuroendocrinology, 34, 150-7.

Florini, J. R. 1987. Hormonal control of muscle growth. Muscle and Nerve, 10, 577-98.

Fry, A. C., W. J. Kraemer, M. H. Stone, L. P. Koziris, J. T. Thrush and S. J. Fleck 2000. Relationships between serum testosterone, cortisol, and weightlifting performance. Journal of Strength and Conditioning Research, 14, 338-43.

Fry, A. C., B. K. Schilling, S. J. Fleck and W. J. Kraemer 2011. Relationships between competitive wrestling success and neuroendocrine responses. The Journal of Strength and Conditioning Research, 25, 40-45.

Galbo, H. 1986. The hormonal response to exercise. Diabetes/Metabolism Reviews, 1, 385-408.

Gallagher, P., M. M. Leitch, A. E. Massey, R. H. McAllister-Williams and A. H. Young 2006. Assessing cortisol and dehydroepiandrosterone (DHEA) in saliva: effects of collection method. Journal of Psychopharmacology, 20, 643-649.

Gatti, R. and E. F. De Palo 2011. An update: salivary hormones and physical exercise. Scandinavian Journal of Medicine and Science in Sports, 21, 157-69.

Gaviglio, C. M. and C. J. Cook 2014. Relationship between midweek training measures of testosterone and cortisol concentrations and game outcome in professional rugby union matches. Journal of Strength and Conditioning Research, 28, 3447-52.

Gaviglio, C. M., B. T. Crewther, L. P. Kilduff, K. A. Stokes and C. J. Cook 2014. Relationship between pregame concentrations of free testosterone and outcome in rugby union. International Journal of Sports Physiology and Performance, 9, 324-31.

Gaviglio, C. M., N. James, B. T. Crewther, L. P. Kilduff and C. J. Cook 2013. Relationship between match statistics, game outcome and pre-match hormonal state in professional rugby union. International Journal of Performance Analysis in Sport, 13, 522-534.

Ghigiarelli, J. J., K. M. Sell, J. M. Raddock and K. Taveras 2013. Effects of strongman training on salivary testosterone levels in a sample of trained men. Journal of Strength and Conditioning Research, 27, 738-47.

104 Gibson, E. L., S. Checkley, A. Papadopoulos, L. Poon, S. Daley and J. Wardle 1999. Increased salivary cortisol reliably induced by a protein-rich midday meal. Psychosomatic Medicine, 61, 214- 24.

Gill, N. D., C. M. Beaven and C. Cook 2006. Effectiveness of post-match recovery strategies in rugby players. British Journal of Sports Medicine, 40, 260-3.

Gladue, B. A., M. Boechler and K. D. McCaul 1989. Hormonal response to competition in human males. Aggressive Behavior, 15, 409-422.

Goncharov, N., G. Katsya, A. Dobracheva, A. Nizhnik, G. Kolesnikova, V. Herbst and J. Westermann 2006. Diagnostic significance of free salivary testosterone measurement using a direct luminescence immunoassay in healthy men and in patients with disorders of androgenic status. The Aging Male, 9, 111-122.

Gonzalez-Bono, E., A. Salvador, M. A. Serrano and J. Ricarte 1999. Testosterone, cortisol, and mood in a sports team competition. Hormones and Behavior, 35, 55-62.

Gozansky, W. S., J. S. Lynn, M. L. Laudenslager and W. M. Kohrt 2005. Salivary cortisol determined by enzyme immunoassay is preferable to serum total cortisol for assessment of dynamic hypothalamic--pituitary--adrenal axis activity. Clinical Endocrinology, 63, 336-41.

Granger, D. A., E. A. Shirtcliff, A. Booth, K. T. Kivlighan and E. B. Schwartz 2004. The "trouble" with salivary testosterone. Psychoneuroendocrinology, 29, 1229-40.

Griffin, J. E. and J. D. Wilson 1998. Disorders of the testes and the male reproductive tract. Williams Textbook of Endocrinology, 8, 709-769.

Groschl, M. 2008. Current status of salivary hormone analysis. Clinical Chemistry, 54, 1759-69.

Hackney, A. C. 1996. The male reproductive system and endurance exercise. Medicine and Science in Sports and Exercise, 28, 180-9.

Häkkinen, K., M. Alén, W. J. Kraemer, E. Gorostiaga, M. Izquierdo, H. Rusko, J. Mikkola, A. Häkkinen, H. Valkeinen, E. Kaarakainen, S. Romu, V. Erola, J. Ahtiainen and L. Paavolainen 2003. Neuromuscular adaptations during concurrent strength and endurance training versus strength training. European Journal of Applied Physiology, 89, 42-52.

105 Häkkinen, K. and A. Pakarinen 1993. Acute hormonal responses to two different fatiguing heavy- resistance protocols in male athletes. Journal of Applied Physiology, 74, 882-7.

Häkkinen, K., A. Pakarinen, M. Alén, H. Kauhanen and P. V. Komi 1987. Relationships between training volume, physical performance capacity, and serum hormone concentrations during prolonged training in elite weight lifters. International Journal of Sports Medicine, 8 Suppl 1, 61-5.

Hansen, S., T. Kvorning, M. Kjaer and G. Sjogaard 2001. The effect of short-term strength training on human skeletal muscle: the importance of physiologically elevated hormone levels. Scandinavian Journal of Medicine and Science in Sports, 11, 347-54.

Henry, J. P. 1992. Biological basis of the stress response. Integrative Physiological and Behavioral Science, 27, 66-83.

Herman, J. P., H. Figueiredo, N. K. Mueller, Y. Ulrich-Lai, M. M. Ostrander, D. C. Choi and W. E. Cullinan 2003. Central mechanisms of stress integration: hierarchical circuitry controlling hypothalamo-pituitary-adrenocortical responsiveness. Frontiers in Neuroendocrinology, 24, 151-80.

Holloszy, J. O. and E. F. Coyle 1984. Adaptations of skeletal muscle to endurance exercise and their metabolic consequences. Journal of Applied Physiology, 56, 831-838.

Hopkins, W. G. 2007. How to analyze a straightforward crossover trial [Excel spreadsheet]. Available at: newstats. org/xcrossover. xls., 18.

Hopkins, W. G., S. W. Marshall, A. M. Batterham and J. Hanin 2009. Progressive statistics for studies in sports medicine and exercise science. Medicine and Science in Sports and Exercise, 41, 3- 13.

Hughes, M. D., Bartlett, R. M. 2001. An analysis of forward play in the 1999 rugby union World Cup for men. In: Books of abstracts Fifth World Congress of Performance Analysis in Sports, 85- 90.

Hughes, M. D., Bartlett, R. M. 2002. The use of performance indicators in performance analysis. Journal of Sports Sciences, 20, 739-754.

Hunter, P. and P. O’Donoghue. A match analysis of the 1999 rugby union world cup. Books of abstracts Fifth World Congress of performance analysis in sports, 2001. 85-90.

106 James, N., S. D. Mellalieu and N. M. Jones 2005. The development of position-specific performance indicators in professional rugby union. Journal of Sports Sciences, 23, 63-72.

Jeffreys, I. and J. Mood 2016. Strength and Conditioning for Sports Performance, New York, Routledge.

Johnson, S. G., G. F. Joplin and J. M. Burrin 1987. Direct assay for testosterone in saliva: relationship with a direct serum free testosterone assay. Clinica Chimica Acta, 163, 309-318.

Jones, N. M., N. James and S. D. Mellalieu 2008. An objective method for depicting team performance in elite professional rugby union. Journal of Sports Sciences, 26, 691-700.

Jones, N. M., S. D. Mellalieu and N. James 2004. Team performance indicators as a function of winning and losing in rugby union. International Journal of Performance Analysis in Sport, 4, 61- 71.

Kilduff, L. P., C. V. Finn, J. S. Baker, C. J. Cook and D. J. West 2013. Preconditioning strategies to enhance physical performance on the day of competition. International Journal of Sports Physiology and Performance, 8, 677-81.

Kindermann, W., A. Schnabel, W. M. Schmitt, G. Biro, J. Cassens and F. Weber 1982. Catecholamines, growth hormone, cortisol, insulin, and sex hormones in anaerobic and aerobic exercise. European Journal of Applied Physiology and Occupational Physiology, 49, 389-99.

Kraemer, W. J., K. Adams, E. Cafarelli, G. A. Dudley, C. Dooly, M. S. Feigenbaum, S. J. Fleck, B. Franklin, A. C. Fry, J. R. Hoffman, R. U. Newton, J. Potteiger, M. H. Stone, N. A. Ratamess and T. Triplett-McBride 2002. American College of Sports Medicine position stand. Progression models in resistance training for healthy adults. Medicine and Science in Sports and Exercise, 34, 364-80.

Kraemer, W. J., N. D. Duncan and J. S. Volek 1998. Resistance training and elite athletes: adaptations and program considerations. Journal of Orthopaedic and Sports Physical Therapy, 28, 110-9.

Kraemer, W. J., S. J. Fleck, R. Callister, M. Shealy, G. A. Dudley, C. M. Maresh, L. Marchitelli, C. Cruthirds, T. Murray and J. E. Falkel 1989. Training responses of plasma beta-endorphin, adrenocorticotropin, and cortisol. Medicine and Science in Sports and Exercise, 21, 146-53.

Kraemer, W. J., D. N. French, N. J. Paxton, K. Häkkinen, J. S. Volek, W. J. Sebastianelli, M. Putukian, R. U. Newton, M. R. Rubin, A. L. Gomez, J. D. Vescovi, N. A. Ratamess, S. J. Fleck, J. 107 M. Lynch and H. G. Knuttgen 2004. Changes in exercise performance and hormonal concentrations over a big ten soccer season in starters and nonstarters. Journal of Strength and Conditioning Research, 18, 121-8.

Kraemer, W. J., A. C. Fry, M. R. Rubin, T. Triplett-McBride, S. E. Gordon, L. P. Koziris, J. M. Lynch, J. S. Volek, D. E. Meuffels, R. U. Newton and S. J. Fleck 2001a. Physiological and performance responses to tournament wrestling. Medicine and Science in Sports and Exercise, 33, 1367-78.

Kraemer, W. J., S. E. Gordon, S. J. Fleck, L. J. Marchitelli, R. Mello, J. E. Dziados, K. Friedl, E. Harman, C. Maresh and A. C. Fry 1991. Endogenous anabolic hormonal and growth factor responses to heavy resistance exercise in males and females. International Journal of Sports Medicine, 12, 228-35.

Kraemer, W. J., C. C. Loebel, J. S. Volek, N. A. Ratamess, R. U. Newton, R. B. Wickham, L. A. Gotshalk, N. D. Duncan, S. A. Mazzetti, A. L. Gomez, M. R. Rubin, B. C. Nindl and K. Häkkinen 2001b. The effect of heavy resistance exercise on the circadian rhythm of salivary testosterone in men. European Journal of Applied Physiology, 84, 13-8.

Kraemer, W. J. and N. A. Ratamess 2005. Hormonal responses and adaptations to resistance exercise and training. Sports Medicine, 35, 339-61.

Kraemer, W. J. and A. D. Rogol 2008. The Encyclopaedia of Sports Medicine An IOC Medical Commission Publication, The Endocrine System in Sports and Exercise, Wiley-Blackwell.

Kuoppasalmi, K., H. Naveri, M. Härkönen and H. Adlercreutz 1980. Plasma cortisol, androstenedione, testosterone and luteinizing hormone in running exercise of different intensities. Scandinavian Journal of Clinical and Laboratory Investigation, 40, 403-9.

Kvorning, T., M. Andersen, K. Brixen and K. Madsen 2006. Suppression of endogenous testosterone production attenuates the response to strength training: a randomized, placebo- controlled, and blinded intervention study. American Journal of Physiology: Endocrinology and Metabolism, 291, E1325-32.

Lane, A. R. and A. C. Hackney 2015. Relationship between salivary and serum testosterone levels in response to different exercise intensities. Hormones (Athens), 14, 258-64.

108 Leslie, M., N. G. Forger and S. M. Breedlove 1991. Sexual dimorphism and androgen effects on spinal motoneurons innervating the rat flexor digitorum brevis. Brain Research, 561, 269-73.

Lewis, J. G. 2006. Steroid analysis in saliva: an overview. Clinical Biochemist, 27, 139-46.

Linnamo, V., A. Pakarinen, P. V. Komi, W. J. Kraemer and K. Häkkinen 2005. Acute hormonal responses to submaximal and maximal heavy resistance and explosive exercises in men and women. Journal of Strength and Conditioning Research, 19, 566-71.

Lippi, G., F. De Vita, G. L. Salvagno, M. Gelati, M. Montagnana and G. C. Guidi 2009. Measurement of morning saliva cortisol in athletes. Clinical Biochemistry, 42, 904-6.

Lippi, G., M. Dipalo, R. Buonocore, C. Gnocchi, R. Aloe and R. Delsignore 2016. Analytical Evaluation of Free Testosterone and Cortisol Immunoassays in Saliva as a Reliable Alternative to Serum in Sports Medicine. Journal of Clinical Laboratory Analysis, 30, 732-5.

Mangine, G. T., J. R. Hoffman, A. M. Gonzalez, J. R. Townsend, A. J. Wells, A. R. Jajtner, K. S. Beyer, C. H. Boone, R. Wang, A. A. Miramonti, M. B. LaMonica, D. H. Fukuda, E. L. Witta, N. A. Ratamess and J. R. Stout 2017. Exercise-Induced Hormone Elevations Are Related to Muscle Growth. Journal of Strength and Conditioning Research, 31, 45-53.

Maso, F., G. Lac, E. Filaire, O. Michaux and A. Robert 2004. Salivary testosterone and cortisol in rugby players: correlation with psychological overtraining items. British Journal of Sports Medicine, 38, 260-3.

Mazur, A. 1985. A biosocial model of status in face-to-face primate groups. Social Forces, 64, 377- 402.

Mazur, A. and A. Booth 1998. Testosterone and dominance in men. Behavioral and Brain Sciences, 21, 353-63.

Mazur, A. and T. A. Lamb 1980. Testosterone, status, and mood in human males. Hormones and Behavior, 14, 236-46.

McArdle, W. D., F. I. Katch and V. L. Katch 2010. Exercise physiology: nutrition, energy, and human performance, Lippincott Williams & Wilkins.

McKenzie, D. C. 1999. Markers of excessive exercise. Canadian Journal of Applied Physiology, 24, 66-73.

109 McLean, D. 1992. Analysis of the physical demands of international rugby union. Journal of Sports Sciences, 10, 285-296.

McLellan, C. P., D. I. Lovell and G. C. Gass 2010. Creatine kinase and endocrine responses of elite players pre, during, and post rugby league match play. Journal of Strength and Conditioning Research, 24, 2908-19.

McLellan, C. P., D. I. Lovell and G. C. Gass 2011. Markers of postmatch fatigue in professional Rugby League players. Journal of Strength and Conditioning Research, 25, 1030-9.

Meeusen, R., M. Duclos, C. Foster, A. Fry, M. Gleeson, D. Nieman, J. Raglin, G. Rietjens, J. Steinacker and A. Urhausen 2013. Prevention, diagnosis, and treatment of the overtraining syndrome: joint consensus statement of the European College of Sport Science and the American College of Sports Medicine. Medicine and Science in Sports and Exercise, 45, 186-205.

Mehta, P. H. and R. A. Josephs 2010. Testosterone and cortisol jointly regulate dominance: evidence for a dual-hormone hypothesis. Hormones and Behavior, 58, 898-906.

Miller, W. L. 1988. Molecular biology of steroid hormone synthesis. Endocrine Reviews, 9, 295- 318.

Morton, R. W., S. Y. Oikawa, C. G. Wavell, N. Mazara, C. McGlory, J. Quadrilatero, B. L. Baechler, S. K. Baker and S. M. Phillips 2016. Neither load nor systemic hormones determine resistance training-mediated hypertrophy or strength gains in resistance-trained young men. Journal of Applied Physiology, 121, 129-38.

Neave, N. and S. Wolfson 2003. Testosterone, territoriality, and the 'home advantage'. Physiology and Behavior, 78, 269-75.

O’Connor, P. J. and D. L. Corrigan 1987. Influence of short-term cycling on salivary cortisol levels. Medicine and Science in Sports and Exercise, 19, 224-228.

Obmiński, Z. 2009. Pre- and post-start hormone levels in blood as an indicator of psycho- physiological load with junior judo competitors. Polish Journal of Sport & Tourism, 16, 158-161.

Obmiński, Z., L. Borkowski, M. Ladyga and E. Hübner-Wozniak 1998. Concentrations of cortisol, testosterone and lactate, and power output in repeated, supra-maximal exercise in elite fencers. Biology of Sport, 15, 19-24.

110 Obmiński, Z. and R. Stupnicki 1997. Comparison of the testosterone-to-cortisol ratio values obtained from hormonal assays in saliva and serum. Journal of Sports Medicine and Physical Fitness, 37, 50-5.

Ortega, E., D. Villarejo and J. M. Palao 2009. Differences in game statistics between winning and losing rugby teams in the Six Nations Tournament. Journal of Sports Science and Medicine, 8, 523- 527.

Paccotti, P., M. Minetto, M. Terzolo, M. Ventura, G. P. Ganzit, P. Borrione, A. Termine and A. Angeli 2005. Effects of high-intensity isokinetic exercise on salivary cortisol in athletes with different training schedules: relationships to serum cortisol and lactate. International Journal of Sports Medicine, 26, 747-55.

Pardridge, W. M. 1986. Serum bioavailability of sex steroid hormones. Clinics in Endocrinology and Metabolism, 15, 259-78.

Passaquin, A. C., P. Lhote and U. T. Ruegg 1998. Calcium influx inhibition by steroids and analogs in C2C12 skeletal muscle cells. British Journal of Pharmacology, 124, 1751-9.

Passelergue, P. and G. Lac 1999. Saliva cortisol, testosterone and T/C ratio variations during a wrestling competition and during the post-competitive recovery period. International Journal of Sports Medicine, 20, 109-13.

Passelergue, P. A. and G. Lac 2012. Salivary hormonal responses and performance changes during 15 weeks of mixed aerobic and weight training in elite junior wrestlers. Journal of Strength and Conditioning Research, 26, 3049-58.

Passos, P., D. Araújo, K. Davids and R. Shuttleworth 2008. Manipulating constraints to train decision making in Rugby Union. International Journal of Sports Science and Coaching, 3, 125- 140.

Peters, J., R. Walker, D. Riad‐Fahmy and R. Hall 1982. Salivary cortisol assays for assessing pituitary‐adrenal reserve. Clinical Endocrinology, 17, 583-592.

Pfaffe, T., J. Cooper-White, P. Beyerlein, K. Kostner and C. Punyadeera 2011. Diagnostic potential of saliva: current state and future applications. Clinical Chemistry, 57, 675-87.

Ponzi, D., S. Zilioli, P. H. Mehta, A. Maslov and N. V. Watson 2016. Social network centrality and hormones: The interaction of testosterone and cortisol. Psychoneuroendocrinology, 68, 6-13. 111 Pope, J. P. and P. M. Wilson 2012. Understanding Motivational Processes in University Rugby Players: A Preliminary Test of the Hierarchical Model of Intrinsic and Extrinsic Motivation at the Contextual Leve. International Journal of Sports Science & Coaching, 7, 89-107.

Popma, A., R. Vermeiren, C. A. M. L. Geluk, T. Rinne, W. van den Brink, D. L. Knol, L. Jansen, H. van Engeland and T. A. H. Doreleijers 2007. Cortisol moderates the relationship between testosterone and aggression in delinquent male adolescents. Biological Psychiatry, 61, 405-411.

Port, K. 1991. Serum and saliva cortisol responses and blood lactate accumulation during incremental exercise testing. International Journal of Sports Medicine, 12, 490-4.

Quarrie, K. L., P. Handcock, M. J. Toomey and A. E. Waller 1996. The New Zealand rugby injury and performance project. IV. Anthropometric and physical performance comparisons between positional categories of senior A rugby players. British Journal of Sports Medicine, 30, 53-6.

Ramamani, A., M. M. Aruldhas and P. Govindarajulu 1999. Impact of testosterone and oestradiol on region specificity of skeletal muscle-ATP, creatine phosphokinase and myokinase in male and female Wistar rats. Acta Physiologica Scandinavica, 166, 91-7.

Roberts, S. P., G. Trewartha, R. J. Higgitt, J. El-Abd and K. A. Stokes 2008. The physical demands of elite English rugby union. Journal of Sports Sciences, 26, 825-33.

Rønnestad, B. R., H. Nygaard and T. Raastad 2011. Physiological elevation of endogenous hormones results in superior strength training adaptation. European Journal of Applied Physiology, 111, 2249-59.

Rosa, C., J. Vilaca-Alves, H. M. Fernandes, F. J. Saavedra, R. S. Pinto and V. M. dos Reis 2015. Order effects of combined strength and endurance training on testosterone, cortisol, growth hormone, and IGF-1 binding protein 3 in concurrently trained men. Journal of Strength and Conditioning Research, 29, 74-9.

Rosner, W. 1991. Plasma steroid-binding proteins. Endocrinology and Metabolism Clinics of North America, 20, 697-720.

Salvador, A., V. Simón, F. Suay and L. Llorens 1987. Testosterone and cortisol responses to competitive fighting in human males: A pilot study. Aggressive Behavior, 13, 9-13.

Salvador, A., F. Suay, S. Martinez-Sanchis, V. M. Simon and P. F. Brain 1999. Correlating testosterone and fighting in male participants in judo contests. Physiology and Behavior, 68, 205-9. 112 Sannikka, E., P. Terho, J. Suominen and R. Santti 1983. Testosterone concentrations in human seminal plasma and saliva and its correlation with non-protein-bound and total testosterone levels in serum. International Journal of Andrology, 6, 319-30.

Sapolsky, R. M., L. M. Romero and A. U. Munck 2000. How do glucocorticoids influence stress responses? Integrating permissive, suppressive, stimulatory, and preparative actions. Endocrine Reviews, 21, 55-89.

Scaramella, T. J. and W. A. Brown 1978. Serum testosterone and aggressiveness in hockey players. Psychosomatic Medicine, 40, 262-5.

Schiml, P. A. and E. F. Rissman 1999. Cortisol facilitates induction of sexual behavior in the female musk shrew (Suncus murinus). Behavioral Neuroscience, 113, 166-75.

Schulkin, J., B. S. McEwen and P. W. Gold 1994. Allostasis, amygdala, and anticipatory angst. Neuroscience and Biobehavioral Reviews, 18, 385-96.

Schumann, M., D. Eklund, R. S. Taipale, K. Nyman, W. J. Kraemer, A. Häkkinen, M. Izquierdo and K. Häkkinen 2013. Acute neuromuscular and endocrine responses and recovery to single- session combined endurance and strength loadings: "order effect" in untrained young men. Journal of Strength and Conditioning Research, 27, 421-33.

Sedeaud, A., A. Marc, J. Schipman, M. Tafflet, J. P. Hager and J. F. Toussaint 2012. How they won Rugby World Cup through height, mass and collective experience. British Journal of Sports Medicine, 46, 580-4.

Smart, D., N. D. Gill, C. M. Beaven, C. Cook and A. Blazevich 2008. The relationship between changes in interstitial creatine kinase and game-related impacts in rugby union. British Journal of Sports Medicine, 42, 198-201.

Smilios, I., T. Pilianidis, M. Karamouzis and S. P. Tokmakidis 2003. Hormonal responses after various resistance exercise protocols. Medicine and Science in Sports and Exercise, 35, 644-54.

Stanhope, J. and M. Hughes 1997. An analysis of scoring in the 1991 Rugby Union World Cup for men. Notational Analysis of Sport I & II. University of Wales Institute Cardiff. Cardiff, Wales, 167-176.

113 Staron, R. S., D. L. Karapondo, W. J. Kraemer, A. C. Fry, S. E. Gordon, J. E. Falkel, F. C. Hagerman and R. S. Hikida 1994. Skeletal muscle adaptations during early phase of heavy- resistance training in men and women. Journal of Applied Physiology, 76, 1247-55.

Stumpf, W. E. 1970. Estrogen-neurons and estrogen-neuron systems in the periventricular brain. American Journal of Anatomy, 129, 207-17.

Stupnicki, R. and Z. Obmiński 1992. Glucocorticoid response to exercise as measured by serum and salivary cortisol. European Journal of Applied Physiology and Occupational Physiology, 65, 546-9.

Suay, F., A. Salvador, E. Gonzalez-Bono, C. Sanchis, M. Martinez, S. Martinez-Sanchis, V. M. Simon and J. B. Montoro 1999. Effects of competition and its outcome on serum testosterone, cortisol and prolactin. Psychoneuroendocrinology, 24, 551-66.

Taipale, R. S. and K. Häkkinen 2013. Acute hormonal and force responses to combined strength and endurance loadings in men and women: the "order effect". PloS One, 8, e55051.

Takarada, Y. 2003. Evaluation of muscle damage after a rugby match with special reference to tackle plays. British Journal of Sports Medicine, 37, 416-9.

Tanner, A. V., B. V. Nielsen and J. Allgrove 2014. Salivary and plasma cortisol and testosterone responses to interval and tempo runs and a bodyweight-only circuit session in endurance-trained men. Journal of Sports Sciences, 32, 680-9.

Tarullo, A. R. and M. R. Gunnar 2006. Child maltreatment and the developing HPA axis. Hormones and Behavior, 50, 632-9.

Taylor, R. 1990. Interpretation of the Correlation Coefficient: A Basic Review. Journal of Diagnostic Medical Sonography, 6, 35-39.

Terburg, D., B. Morgan and J. van Honk 2009. The testosterone-cortisol ratio: A hormonal marker for proneness to social aggression. International Journal of Law and Psychiatry, 32, 216-23.

Tremblay, M. S., J. L. Copeland and W. Van Helder 2004. Effect of training status and exercise mode on endogenous steroid hormones in men. Journal of Applied Physiology, 96, 531-9.

Turan, B. 2015. Predictors of anticipatory cortisol reactivity to subsequent stressors. Physiology and Behavior, 149, 239-46.

114 Umeda, T., R. Hiramatsu, T. Iwaoka, T. Shimada, F. Miura and T. Sato 1981. Use of saliva for monitoring unbound free cortisol levels in serum. Clinica Chimica Acta, 110, 245-53.

Urhausen, A., H. Gabriel and W. Kindermann 1995. Blood hormones as markers of training stress and overtraining. Sports Medicine, 20, 251-76.

Urhausen, A. and W. Kindermann 2002. Diagnosis of overtraining: what tools do we have? Sports Medicine, 32, 95-102.

Van Honk, J., R. P. Kessels, P. Putman, G. Jager, H. P. Koppeschaar and A. Postma 2003. Attentionally modulated effects of cortisol and mood on memory for emotional faces in healthy young males. Psychoneuroendocrinology, 28, 941-8.

VanBruggen, M. D., A. C. Hackney, R. G. McMurray and K. S. Ondrak 2011. The relationship between serum and salivary cortisol levels in response to different intensities of exercise. International Journal of Sports Physiology and Performance, 6, 396-407.

Vaz, L., M. Van Rooyen and J. Sampaio 2010. Rugby game-related statistics that discriminate between winning and losing teams in IRB and Super twelve close games. Journal of Sports Science and Medicine, 9, 51-55.

Vermeulen, A., T. Stoica and L. Verdonck 1971. The apparent free testosterone concentration, an index of androgenicity. Journal of Clinical Endocrinology and Metabolism, 33, 759-67.

Vervoorn, C., A. M. Quist, L. J. Vermulst, W. B. Erich, W. R. de Vries and J. H. Thijssen 1991. The behaviour of the plasma free testosterone/cortisol ratio during a season of elite rowing training. International Journal of Sports Medicine, 12, 257-63.

Vining, R. F. and R. A. McGinley 1986. Hormones in saliva. Critical Reviews in Clinical Laboratory Sciences, 23, 95-146.

Vining, R. F. and R. A. McGinley 1987. The measurement of hormones in saliva: possibilities and pitfalls. Journal of Steroid Biochemistry, 27, 81-94.

Vining, R. F., R. A. McGinley, J. J. Maksvytis and K. Y. Ho 1983a. Salivary cortisol: a better measure of adrenal cortical function than serum cortisol. Annals of Clinical Biochemistry, 20 (Pt 6), 329-35.

115 Vining, R. F., R. A. McGinley and R. G. Symons 1983b. Hormones in saliva: mode of entry and consequent implications for clinical interpretation. Clinical Chemistry, 29, 1752-6.

Virkkunen, M. 1985. Urinary free cortisol secretion in habitually violent offenders. Acta Psychiatrica Scandinavica, 72, 40-4.

Viru, A., K. Karelson and T. Smirnova 1992. Stability and variability in hormonal responses to prolonged exercise. International Journal of Sports Medicine, 13, 230-5.

Viru, A. and M. Viru 2004. Cortisol--essential adaptation hormone in exercise. International Journal of Sports Medicine, 25, 461-4.

Walker, R., D. Wilson, G. Read and D. Riad‐Fahmy 1980. Assessment of testicular function by the radioimmunoassay of testosterone in saliva. International Journal of Andrology, 3, 105-120.

Wang, C., S. Plymate, E. Nieschlag and C. A. Paulsen 1981. Salivary testosterone in men: further evidence of a direct correlation with free serum testosterone. Journal of Clinical Endocrinology and Metabolism, 53, 1021-4.

Weitzman, E. D., C. A. Czeisler, J. C. Zimmerman and M. C. Moore-Ede 1981. Biological rhythms in man: relationship of sleep-wake, cortisol, growth hormone, and temperature during temporal isolation. Advances in Biochemical Psychopharmacology, 28, 475-99.

West, D. J., C. V. Finn, D. J. Cunningham, D. A. Shearer, M. R. Jones, B. J. Harrington, B. T. Crewther, C. J. Cook and L. P. Kilduff 2014. Neuromuscular function, hormonal, and mood responses to a professional rugby union match. Journal of Strength and Conditioning Research, 28, 194-200.

West, D. W. and S. M. Phillips 2012. Associations of exercise-induced hormone profiles and gains in strength and hypertrophy in a large cohort after weight training. European Journal of Applied Physiology, 112, 2693-702.

Wood, P. 2009. Salivary steroid assays - research or routine? Annals of Clinical Biochemistry, 46, 183-96.

Yehuda, R. and J. S. Meyer 1984. A role for serotonin in the hypothalamic-pituitary-adrenal response to insulin stress. Neuroendocrinology, 38, 25-32.

116 Zilioli, S. and N. V. Watson 2012. The hidden dimensions of the competition effect: basal cortisol and basal testosterone jointly predict changes in salivary testosterone after social victory in men. Psychoneuroendocrinology, 37, 1855-65.

117 APPENDIX 1: PARTICIPANT INFORMATION DOCUMENT

Bath Research Ethics Committee Room 11, John Apley Building Research Ethics Office Royal United Hospital Combe Park Bath BA1 3NG Tel/Fax: 01225 825725 [email protected]

24 August 2009

Mr Oliver Peacock Research Officer School for Health University of Bath Bath, BA2 7AY

Dear Mr Peacock

Study Title: Monitoring physiological strain during exercise training and recovery in athletes REC reference number: 09/H0101/66 Protocol number: 1

The Research Ethics Committee reviewed the above application at the meeting held on 20 August 2009. The Committee would like to thank Dr Keith Stokes for attending to discuss the study.

Ethical opinion

Record of ethical issues discussed

1. Suitability of the applicant and supporting staff: It was noted that you are a Research Officer at the School for Health, University of Bath and are very suitable to conduct this research. 2. Quality of the facilities: The facilities, in the Sport and Exercise Department at the University of Bath, are very good. 3. Relevance of the research and research design: The following points were raised: 3.1 It was noted that a wide range of sporting activities would be included in the research and that, directly before or after a session, a participant would be requested to meet with a researcher immediately in order for tests to be administered. It was asked how, logistically,

118 this would happen with so many diverse locations, and whether the research team have sufficient researchers to cover this. Dr Stokes replied that the research team was confident they would be able to cover this. Dr Stokes noted that you had been employed full time for three years so this is a major research project for you. This was accepted by the Committee. 4. Anticipated benefits and risks: There are minimal risks and benefits to participants. 5. Care and protection of the research subject: The following points were raised: 5.1 It was noted that the training to be undertaken by participants was substantial, but that this was normal for elite athletes. 5.2 It was noted that there might be some compunction on the part of athletes to participate but agreed that this was not the case as athletes expected to participate in this kind of research and were known to the research team. 6. Hazards, discomfort and distress of subjects: This is adequate. 7. Consent: This is suitable. 8. Adequacy of written information: Participant Information Sheet The following point was raised: 8.1 QA13 of the application form notes that the research team would be unsure when the tests will take place when they sign up participants for the study. Therefore it was noted that the Information Sheet should inform participants that a time will be fixed with them and their coach by mutual agreement. This was agreed by Dr Stokes. 9. Recruitment arrangements: The following point was raised: 9.1 It was noted that the project would be advertised locally and by word of mouth. It was asked what the research team would do if word of mouth proved insufficient to recruit participants. Dr Stokes replied that some of the research team have external contacts in the wider UK sport environment whom they could ask. Dr Stokes also noted that people are currently approaching the School of Health team about research so the interest is there. This was accepted by the Committee. 10. Confidentiality: This is satisfactory. 11. Provision of indemnity: The following point was noted: 11.1 It was noted that the date on the indemnity document, which is provided by the University of Bath, has expired. An updated version was requested. This was agreed by Dr Stokes. 12. General issues: The following point was raised: 12.1 QB4 of the application form implies that blood samples will be stored for future use, although it is stated elsewhere in the application that samples won’t be stored. Clarification was requested. Dr Stokes confirmed that this was an administrative error and that all blood samples would be destroyed once the relevant measurements have been taken. This was accepted by the Committee.

The members of the Committee present gave a favourable ethical opinion of the above research on the basis described in the application form, protocol and supporting documentation, subject to the conditions specified below.

119 Ethical review of research sites The favourable opinion applies to all NHS sites taking part in the study, subject to management permission being obtained from the NHS/HSC R&D office prior to the start of the study (see “Conditions of the favourable opinion” below).

Conditions of the favourable opinion

The favourable opinion is subject to the following conditions being met prior to the start of the study.

Management permission or approval must be obtained from each host organisation prior to the start of the study at the site concerned.

For NHS research sites only, management permission for research (“R&D approval”) should be obtained from the relevant care organisation(s) in accordance with NHS research governance arrangements. Guidance on applying for NHS permission for research is available in the Integrated Research Application System or at http://www.rdforum.nhs.uk. Where the only involvement of the NHS organisation is as a Participant Identification Centre, management permission for research is not required but the R&D office should be notified of the study. Guidance should be sought from the R&D office where necessary.

Sponsors are not required to notify the Committee of approvals from host organisations.

Requested Conditions

1. A revised Participant Information Sheet should be submitted, to inform participants that a time for them to participate in tests will be fixed with them and their coach by mutual agreement (point 8.1). 2. An updated version of the University of Bath indemnity document should be submitted (point 11.1).

If further clarification is requested following receipt of the decision letter, it is requested that you contact Coordinator Miss Vanessa Bishop for advice.

It is responsibility of the sponsor to ensure that all the conditions are complied with before the start of the study or its initiation at a particular site (as applicable).

Approved documents

The documents reviewed and approved at the meeting were: Document Version Date Participant Consent Form 1 04 August 2009 Participant Information Sheet 1 04 August 2009 Questionnaire: Medical Questionnaire Compensation Arrangements 28 August 2008 Peer Review 04 August 2009 Covering Letter 04 August 2009 Protocol 1 04 August 2009 Investigator CV 04 August 2009 REC application 2.3 04 August 2009 120 Membership of the Committee

The members of the Ethics Committee who were present at the meeting are listed on the attached sheet.

Statement of compliance

The Committee is constituted in accordance with the Governance Arrangements for Research Ethics Committees (July 2001) and complies fully with the Standard Operating Procedures for Research Ethics Committees in the UK.

After ethical review

Now that you have completed the application process please visit the National Research Ethics Service website > After Review You are invited to give your view of the service that you have received from the National Research Ethics Service and the application procedure. If you wish to make your views known please use the feedback form available on the website. The attached document “After ethical review – guidance for researchers” gives detailed guidance on reporting requirements for studies with a favourable opinion, including: • Notifying substantial amendments • Adding new sites and investigators • Progress and safety reports • Notifying the end of the study

The NRES website also provides guidance on these topics, which is updated in the light of changes in reporting requirements or procedures. We would also like to inform you that we consult regularly with stakeholders to improve our service. If you would like to join our Reference Group please email [email protected].

09/H0101/66 Please quote this number on all correspondence

With the Committee’s best wishes for the success of this project

Yours sincerely

Dr Jeff Handel Alternate Vice-Chair

Enclosures: List of names and professions of members who were present at the meeting and those who submitted written comments “After ethical review – guidance for researchers” [SL-AR1 for CTIMPs, SL- AR2 for other studies]

Copy to: Professor Ken Judge, Head of School for Health, School for Health, Norwood House, University of Bath, Claverton Down, Bath, BA2 7AY

121 Bath Research Ethics Committee

Attendance at Committee meeting on 20 August 2009

Committee Members:

Name Profession Present Notes Dr Jenny Bell General Practitioner Yes Dr Paul F Bennett General Practitioner Yes Mr Michael Bowman Engineer Yes Professor David Brown Professor of Biochemistry Yes Mr John Brownrigg Partner Yes Reverend Alastair Davies Hospital Chaplain No Dr Fiona Finlay Consultant Paediatrician No Mr Grey Giddins Consultant Hand Surgeon No Dr Jeff Handel Consultant Yes Anaesthetist/Alternate Vice-Chair Mr Richard James Database Manager Yes Dr Eleanor Korendowych Consultant Rheumatologist No Dr Martin Marlowe Consultant Psychiatrist Yes Mr Philip Morgan Local Government Officer Yes Dr Karen Rodham Lecturer/Director for Yes Studies MSc Health Psychology Dr Jenny Scott Senior Lecturer in No Pharmacy Practice/Chair Mr Paul Shipley Civil Servant/Vice-Chair Yes Dr Andrew Taylor Consultant Biochemist Yes Mr Trevor Wild Staff Nurse - CCU Yes

Also in attendance:

Name Position (or reason for attending) Miss Vanessa Bishop REC Co-ordinator Miss Naomi Fisher Observer

122 Appendix 2: Participant Information Document

REF: 09/H0101/66 DATE: 04/08/09 VERSION: 2

Participant Information Sheet

You are being invited to take part in a research study. Before you decide it is important for you to understand why the research is being done and what it will involve. Please take time to read the following information carefully and discuss it with others if you wish. Ask us if there is anything that is not clear or if you would like more information. Take time to decide whether or not you wish to take part.

Study Title Monitoring physiological strain during exercise training and recovery in athletes

Background Information Monitoring physiological responses to training is commonplace within elite sport. Such information can be used to refine training programmes and ensure that athletes maximise their potential for adaptation, while minimising their risk of overtraining or incurring injury. The purpose of this research is to monitor selected physiological responses during typical training sessions in elite athletes and to use this information to help guide exercise prescription and enhance athletic performance.

Invitation to participate We invite you to participate in a research project examining selected physiological responses of elite athletes performing their normal training practices.

Do you have to take part? Participation in this study is completely voluntary and it is up to you to decide whether or not to take part. If you decide to take part you will be given this information sheet to keep and asked to sign a consent form. If you do decide to take part you are still free to withdraw at any time and without giving a reason. If you decide not to take part or to withdraw from the study, this will not affect your future training or athlete status.

Exclusion Criteria Please check these criteria carefully and ask if you feel that you might fall into one of the following groups: - Non-elite athlete - Aged less than 18 years, aged more than 39 years - Smoker - Previous medical history of heart arrhythmia

123 - Women who are known to be pregnant or who are intending to become pregnant over the course of the study.

Procedures Should you agree to participate in this study then you will be asked to continue with your normal training programme. A time for you to participate in tests will be arranged with yourself and your coach by mutual agreement. During selected training sessions, saliva samples and finger prick blood samples will be taken pre and post-exercise and post- recovery. This process will generally not surpass more than 3 sessions per week, although there may be occasions whereby measurements will be recorded on up to 10 consecutive days depending on your training and competition schedule. This will not interfere with the content of your training session, and will take place in an area away from other athletes. The finger-prick blood sample and saliva sample will be collected over a period of less than 10 minutes in duration. On the first of these sessions, you will be instructed on how to provide a saliva sample and you will be talked through the finger-prick blood sampling procedure. This monitoring procedure may continue for up to 12 months, but may be shorter depending on the specific demands of your training.

Requirements - You will be asked to continue with your normal training programme - On testing days, you will be asked to report to a researcher immediately before and after training and after recovery for sample collection - You will be required to continue recording your normal training activities in a training log, and will allow investigator access to this training history on study completion.

What are the benefits of taking part? The research staff at the School for Health are scientists and are not medically trained (i.e. not medical doctors). They cannot therefore give medical advice or interpretation of your results. The test kits and equipment they use is for scientific research - it is not certified for medical diagnosis. You will be provided with the results of your measures taken for this research project. Where a ‘normal’ range for this measure is known we will also tell you what this is. If your results fall outside of this ‘normal’ range this is not necessarily an indication of any medical problem. ‘Normal’ values have not been established for every measure, so we may not always be able to give you these. Factors like diet, minor illness e.g. colds and exercise can influence some results. If data is found to be irregular, the anonymised results will be sent to an independent medical officer who will contact you if a follow-up is required.

- The information we will get from this study may help inform future research into refining training programmes for elite athletes - At the end of the study you will be able to receive results from all physiological measures taken - You will receive individual feedback on how this information relates to your training and performance - There are no other benefits of taking part

What are the risks of taking part? - The risks involved are those associated with your normal training programme

124 What will happen to the results of this study? All results will remain confidential, and names will be removed prior to publication. You can only receive your own results and not those of others.

Who has reviewed this study? This study has been reviewed by the Bath Local Research Ethics Committee.

Who to contact in case of a claim? This proposal is covered by the general compensation arrangements of the University of Bath. For more information on these arrangements please contact Simon Holt on: 01225 385129

Many thanks for considering participation in the study If you have any further queries, please do not hesitate to contact us. If you agree to take part in this study then you will be given a copy of the Information Sheet and a signed Consent Form to keep.

Investigator contact details Chris Gaviglio E-mail: [email protected] Principal Investigator

Dr Christian Cook E-mail: [email protected] Supervisor

125 Appendix 3: Consent Form

REF: 09/H0101/66 DATE: 04/08/09 VERSION: 1

CONSENT FORM

Title of Project: Monitoring physiological strain during exercise training and recovery in athletes

Researchers: Chris Gaviglio E-mail: [email protected] Dr Christian Cook E-mail: [email protected]

Please initial box

1. I confirm that I have read the information sheet dated …./.…/2009 (version……) for the above study and have had the opportunity to ask questions.

2. I understand that my participation is voluntary and that I am free to withdraw at any time, without giving any reason and without my legal rights being affected.

3. I agree to take part in this study.

Name of participant Date Signature

Name of chief investigator Date Signature

126