GASTROINTESTINAL DAMAGE, INFLAMMATION AND CENTRAL FATIGUE DURING EXERCISE IN THE HEAT

John Owen Osborne Bachelor of Exercise and Nutrition Sciences (Hons)

Submitted in fulfilment of the requirements for the degree of Doctor of Philosophy

School of Exercise and Nutrition Sciences Faculty of Health Queensland University of Technology 2019 Keywords

endotoxemia exercise heat stress thermoregulation central fatigue cycling cytokines gastrointestinal permeability glutamine supplementation heat acclimation hyperthermia

LPS neuromuscular performance inflammation

ii Gastrointestinal damage, inflammation and central fatigue during exercise in the heat Abstract

Exercise in the heat results in reduced performance due to an earlier onset of fatigue.

This issue is relevant to both professional and recreational athletes, as well as certain military and civilian occupations. Current literature suggests that some of this impairment in exercise performance may potentially arise from a centrally-mediated decrease in muscle activation. However, the causal link between exercise, heat and central fatigue remains equivocal.

During exercise in the heat, high core temperatures and reductions in gut blood flow increase the permeability of the gastrointestinal lining to endotoxins, which translocate through the damaged barrier and into systemic circulation. An elevation in circulating endotoxins triggers the release of pro-inflammatory cytokines, which have been linked to the occurrence of fatigue, and may modulate neuronal activity.

This PhD investigated the relationship between exercise performance in the heat, gut damage, endotoxin release, inflammatory cytokines, and neuromuscular function. The initial proof-of-concept study found that 60 min of moderate-to-vigorous cycling in the heat resulted in a significant decrease in voluntary activation of knee extensor muscles and an elevation in circulating markers of gut damage, compared to a temperate condition. However, no condition differences were observed for endotoxin concentration, level of inflammatory cytokines or intestinal permeability. The findings from this study suggest that exercise in the heat results in diminished neuromuscular activation and hyperthermia-induced gut damage.

The second study investigated the efficacy of glutamine supplementation on the performance of twelve well-trained cyclists during a 20 km time trial in the heat.

Although glutamine did not improve cycling performance over a placebo, it was

Gastrointestinal damage, inflammation and central fatigue during exercise in the heat iii observed to maintain voluntary strength and attenuate an increase in particular inflammatory cytokines and markers of gut damage. The findings from this study provided evidence that glutamine supplementation does not benefit shorter duration

(~33 min) cycling tasks in the heat. However, glutamine-mediated protection of gut barrier integrity and reduced inflammation, resulting in improved preservation of knee extensor strength, could potentially enhance athletic performance over longer-duration exercise protocols, common in many competitive events (e.g., marathons, triathlons, time trials).

Finally, the third investigation examined the effect of 5 days of heat acclimation training on 20 km time trial cycling performance in the heat, inflammation and neuromuscular performance. Heat acclimation was found to improve both 20 km performance and knee extensor strength, without inducing additional inflammatory stress, central fatigue or gut damage. In contrast, 5 days of training in thermoneutral conditions did not enhance self-paced 20 km performance in the heat and voluntary strength was reduced from initial baseline values, possibly due to cumulative peripheral fatigue. In conclusion, the data from this study supported an ergogenic effect from short-term heat acclimation training for athletes who undertook strenuous exercise in the heat. Moreover, the similar level of central fatigue and inflammation following heat acclimation, despite a larger workload, indicated that this intervention might preserve neuromuscular function and protect against exertional-endotoxemia.

Collectively, this body of work demonstrates that strenuous exercise in the heat damages the gastrointestinal tract, impairs voluntary activation of skeletal muscle, and may result in elevated pro-inflammatory cytokines. In contrast to the thesis hypothesis and previous literature, hyperthermic exercise did not induce detectable endotoxin translocation, despite inflammation and central fatigue. This outcome potentially

iv Gastrointestinal damage, inflammation and central fatigue during exercise in the heat suggests that transient endotoxemia is unlikely to be a mechanistic link between elevated core temperatures and CNS-mediated downregulation of voluntary drive during exercise in the heat. However, the findings of this PhD do tentatively support a possible association between heat, gut damage, inflammation and diminished neuromuscular function. Acute (i.e., glutamine) or chronic (i.e., heat acclimation) interventions that disrupt this cascade may result in beneficial performance improvements.

Gastrointestinal damage, inflammation and central fatigue during exercise in the heat v Table of Contents

Keywords ...... ii Abstract ...... iii Table of Contents ...... vi List of Figures ...... viii List of Tables ...... x List of Abbreviations ...... xii Statement of Original Authorship ...... xiv Publications...... xvii Chapter 1: Introduction ...... 1 1.1 Summary ...... 5 Chapter 2: Literature Review ...... 8 2.1 Thermoregulation during exercise ...... 8 2.2 Exercise performance in the heat ...... 9 2.3 Models of fatigue in the heat ...... 11 2.3.1 Critical core temperature theory ...... 12 2.3.2 Cardiovascular factors ...... 14 2.3.3 Central nervous system factors ...... 16 2.4 Neuroinflammatory fatigue ...... 19 2.4.1 Exertional endotoxemia ...... 20 2.5 Protective Interventions...... 32 2.5.1 Glutamine Supplementation ...... 33 2.5.2 Heat Acclimation ...... 41 2.6 Summary and Implications ...... 53 Chapter 3: Study 1 – The effects of cycling in the heat on gastrointestinal inflammation and neuromuscular fatigue...... 54 3.1 Abstract ...... 54 3.2 Introduction ...... 55 3.3 Methods ...... 57 3.4 Results ...... 66 3.5 Discussion ...... 72 3.6 Conclusion ...... 77 3.7 Acknowledgments ...... 78 3.8 Linking Section ...... 84 Chapter 4: Study 2 – Glutamine supplementation does not improve 20 km cycling time trial performance in the heat...... 87 4.1 Abstract ...... 87

vi Gastrointestinal damage, inflammation and central fatigue during exercise in the heat 4.2 Introduction ...... 88 4.3 Methods ...... 90 4.4 Results ...... 98 4.5 Discussion ...... 105 4.6 Conclusion ...... 108 4.7 Acknowledgments ...... 109 4.8 Tables ...... 111 4.9 Linking Section ...... 116 Chapter 5: Study 3 – Short-duration heat acclimation training improves 20 km cycling performance in the heat and enhances knee extensor strength...... 118 5.1 Abstract ...... 118 5.2 Introduction ...... 119 5.3 Methods ...... 121 5.4 Results ...... 130 5.5 Discussion ...... 144 5.6 Conclusion ...... 149 5.7 Acknowledgments ...... 150 5.8 Tables ...... 151 Chapter 6: General Discussion ...... 166 6.1 Thesis aims ...... 166 6.2 Principal findings ...... 167 6.3 Integrative Neuroinflammatory Model of Fatigue – A Concept Revisited ...... 170 6.4 Conclusion ...... 177 6.5 Limitations and delimitations ...... 178 6.6 Practical applications ...... 182 6.7 Recommendations for future research ...... 183 Chapter 7: Bibliography ...... 185 Chapter 8: Appendices...... 210 8.1 Appendix A - Timeline ...... 210 8.2 Appendix B – Neuromuscular Variables and Definitions ...... 213 8.3 Appendix C – Publications and Conference Abstracts ...... 214 8.4 Appendix D – Scales & Datasheets (Study 1) ...... 216 8.5 Appendix E – Scales & Datasheets (Study 2) ...... 223 8.6 Appendix F – Scales & Datasheets (Study 3) ...... 234 8.7 Appendix G – Consent Forms...... 244

Gastrointestinal damage, inflammation and central fatigue during exercise in the heat vii List of Figures

Figure 2.1. The effect of hypoxia and hyperthermia on intestinal barrier permeability during exercise in the heat. Reprinted from “Role of Gastrointestinal Permeability in Exertional Heatstroke,” by G. P. Lambert, 2004, Exercise and Sport Sciences Reviews, 32(4), pg. 186...... 21 Figure 3.1. Posterior predicted mean [95% CI] for MVC torque (N∙m), VA (%) and Pt torque (N∙m) at pre, post and 1 h post-exercise in HOT and CON. * indicates a statistical time difference compared to pre- exercise values within a condition; † indicates a statistical condition difference at a respective time point...... 69 Figure 3.2. Posterior predicted mean [95% CI] serum concentrations of inflammatory cytokines (IL-1β and TNF-α), markers of intestinal damage (I-FABP) and permeability (CLDN-3) at pre, post and 1 h post-exercise in HOT and CON. * indicates a statistical time difference compared to pre-exercise values within a condition; † indicates a statistical condition difference at a respective time point ...... 71 Figure 4.1. Physiological and perceptual measures during 20 km TT in the heat following ingestion of glutamine (GLUT) or placebo (CON). (A) HR; (B) Tre; (C) Tsk; (D) RPE; (E) Thermal sensation; (F) Thermal Comfort. Data displayed as posterior predicted...... 100 Figure 4.2. The neuromuscular function of the knee extensors pre and post-20TT following ingestion of glutamine (GLUT) or placebo (CON). Data displayed as posterior predicted mean with ± 95% CI and overlayed with raw individual responses. † indicates a significant difference from pre-exercise values in the same condition...... 102 Figure 4.3. Plasma concentration of TNF-α, IL-6 and I-FABP at pre- and post-20TT timepoints. Data displayed as posterior predicted mean with ± 95% CI and overlayed with raw individual responses. † indicates a significant difference from the pre-exercise value of the same condition...... 104 Figure 5.1. Experimental schematic of a training block. NM: neuromuscular assessment...... 122

Figure 5.2. Individual 20TTFINAL completion time improvement for CON and HA...... 133

Figure 5.3. Mean power output of 20TTFINAL (relative to 20TTINITIAL values) for CON and HA...... 133 Figure 5.4. Mean and 95% CI for (A) HR, (B) core and (C) skin temperature during CON and HA 20TT tests...... 135 Figure 5.5. Mean and 95%CI for MVC torque throughout each training block. * indicates HA is statistically different to CON at same time- point. † indicates statistically different pre- to post-exercise torque in the same condition; # indicates torque is statistically different from same time-point and condition in 20TTINITIAL...... 138

viii Gastrointestinal damage, inflammation and central fatigue during exercise in the heat Figure 5.6. Mean and 95%CI for I-FABP for each 20TT and training days. † indicates statistically different pre- to post-exercise torque in the same condition; # indicates torque is statistically different from same time-point and condition in 20TTINITIAL...... 143

Gastrointestinal damage, inflammation and central fatigue during exercise in the heat ix List of Tables

Table 2.1. Summarised exercise and heat-stress research with measured outcomes of LPS, inflammatory cytokines, GI permeability and damage...... 25 Table 2.2. Cytokine reference values from previous exercise-heat stress and endotoxemia research...... 29 Table 2.3. Summarised glutamine supplementation research with measured outcomes of LPS or GI permeability and/or damage...... 40 Table 2.4. Adaptations associated with heat acclimation...... 42 Table 3.1. Posterior predicted mean [95% credible interval] for pre, post and 1 h post-exercise for neuromuscular variables in HOT or CON...... 79 Table 3.2. Posterior predicted mean [95% credible interval] for pre-, post and 1-h post-exercise for blood markers in HOT or CON...... 81 Table 3.3. Posterior predicted values [95% credible interval] for physiological and perceptual measures during exercise...... 82 Table 4.1. Baseline, post and pre-post posterior predicted data (mean [95% credible interval])...... 111 Table 4.2. Gastrointestinal distress post-exercise questionnaire (N = response rate/12; mean of responders (range)) ...... 112 Table 4.3. Posterior predicted mean data for 20TT exercise parameters (mean [95% credible interval])...... 113 Table 4.4. Posterior predicted data for pre- and post-20TT neuromuscular function parameters (mean [95% credible interval])...... 114 Table 4.5. Posterior predicted data for blood markers (mean [95% credible interval])...... 115 Table 5.1. Pre-trial data for initial and final 20TT (mean [95% credible interval])...... 151 Table 5.2. Performance data for initial and final 20TT (mean [95% credible interval])...... 153 Table 5.3. GI symptom questionnaire data for initial and final 20TT...... 155 Table 5.4. Neuromuscular properties for initial and final 20TT (mean [95% credible interval])...... 156 Table 5.5. Plasma concentrations of inflammatory markers for initial and final 20TT (mean [95% credible interval])...... 158 Table 5.6. Baseline data for Day 1 and Day 5 of training (mean [95% credible interval])...... 159 Table 5.7. Exercise data for Day 1 and Day 5 of training (mean [95% credible interval])...... 161

x Gastrointestinal damage, inflammation and central fatigue during exercise in the heat Table 5.8. Neuromuscular properties for Day 1 and Day 5 of training (mean [95% credible interval])...... 163 Table 5.9. Plasma concentrations of inflammatory markers for Day 1 and Day 5 of training (mean [95% credible interval])...... 165

Gastrointestinal damage, inflammation and central fatigue during exercise in the heat xi List of Abbreviations

~ approximately EDTA ethylenediaminetetraacetic acid

° degree EMG electromyography

°C degrees Celsius ELISA enzyme-linked immunosorbent assay

∆ delta (change) EU endotoxin units

½ RT half relaxation time F force

% percentage FM fat mass

x̄ mean FFM fat-free mass

μs microsecond g gram

μL microlitre GI gastrointestinal

ACSM American College of Sports Medicine GLUT glutamine

ANOVA analysis of variance h hour

BBB blood-brain barrier HA heat acclimation

BM body mass Hb haemoglobin

BMI body mass index Hct haematocrit

BP blood pressure Hz Hertz bpm beats per minute HR heart rate

CAR central activation ratio HSF Heat Shock Factor

CD contraction duration HSP Heat Shock Protein

CI credible interval ICC interclass correlation

CIT critical internal temperature IL interleukin cm centimetre kg kilogram

CON control kHz kilohertz

CNS central nervous system km kilometre

CV coefficient of variation L litre d Cohen’s d La-1 blood lactate

DIC deviance information criterion LAL Limulus amebocyte lysate

xii Gastrointestinal damage, inflammation and central fatigue during exercise in the heat LBM lean body mass s second

LPS Lipopolysaccharide SD standard deviation

m metre SEM standard error of measurement

MD mean difference sRPE sessional rating of perceived exertion

min minute SWC smallest worthwhile change

mL millilitre Tc core temperature

mm millimetre Tcomfort thermal comfort

ms millisecond TEM typical error of measurement

mV millivolt THN thermoneutral

MVC maximum voluntary contraction TMS transcranial magnetic stimulation

N Newtons TNF-α tumour necrosis factor alpha

NF-κB nuclear factor kappa B TPt time to peak twitch torque

-1 ng∙mL nanograms per millilitre Tre rectal temperature

N∙m torque Tsensation thermal sensation

nm nanometres Tsk skin temperature

np2 partial eta squared TT time trial

pg∙mL-1 picograms per millilitre USG urine specific gravity

Pt peak twitch torque V volt

PV plasma volume VA voluntary activation

r Pearson’s product moment correlation VL vastus lateralis

RH relative humidity VM vastus medialis

ρ Spearman’s rho correlation VO2max maximal oxygen uptake

RPE rating of perceived exertion VO2peak peak oxygen uptake

RPM revolutions per minute W Watt

RR rate of relaxation W∙kg-1 Watts per kilogram

RTD rate of torque development WBGT wet-bulb globe temperature

Gastrointestinal damage, inflammation and central fatigue during exercise in the heat xiii Statement of Original Authorship

The work contained in this thesis is composed of original work and to the best of my knowledge and belief, contains no material previously published or written by another person except where due reference is made. No part of this thesis has been previously submitted to meet requirements for an award of any other degree at any other higher education institution.

Signature: QUT Verified Signature

Date: 31.07.2019

xiv Gastrointestinal damage, inflammation and central fatigue during exercise in the heat Acknowledgments

First and foremost, I would like to acknowledge the tremendous guidance I have received from my principal supervisor, Dr. Geoffrey Minett. Thank you for your endless patience, attention to detail and belief in my abilities. The feedback, direction and support you provided in every aspect of this PhD has been invaluable and has set me in good stead for my future academic career.

Thank you also to my associate supervisor, Professor Ian Stewart. Your experience, as both a researcher and as a supervisor, has been invaluable to me over the course of this PhD. I really appreciate your thought-provoking questions, calm demeanour, good humour and ability to make me step back and see the bigger picture.

To my associate supervisor, Professor Ken Beagley, thank you for your biochemical expertise, for providing me access to the facilities at QIMR, and for welcoming me into the Beagley lab group. A special thanks to Logan Trim, for volunteering hours of your own time assisting me with the various biochemical analyses that were completed for this PhD. I won’t forget your generosity and the many long nights we spent running those ELISA kits.

I am grateful for each and every one of the ENS postgrads that have made the past few years so enjoyable: Aaron, Benny, Rob, Pat, Nicki, Jess, Kate and Jesse.

Thank you for the chats, laughs, beers, coffees, and for participating in my studies. To my main PhD colleague, David Borg: thank you for your continued encouragement, work ethic, advice and friendship; especially after we spent multiple 14 hours days together in the lab. I would not have been able to

Gastrointestinal damage, inflammation and central fatigue during exercise in the heat xv complete this PhD without you, and I look forward to our continued collaborations in the future. I would like to acknowledge QUT for providing an

Australian Government Research Training Program Scholarship and also my external examiners for their valuable feedback that improved this final thesis.

To mum and dad, thank you for the endless love and support over the years. You instilled in me an inquisitive nature, a love of reading, and the importance of a job well done. I would never have considered beginning this journey without such wonderful parents and am grateful for the opportunities that you have provided me.

Finally, to Emma, thank you for the continuous encouragement, motivation and belief in my ability to complete this PhD. Your unwavering support got me through even the most difficult times and coming home to you and Izzy always made things better. Your love and encouragment has made this thesis possible, and so I dedicate it to you.

xvi Gastrointestinal damage, inflammation and central fatigue during exercise in the heat Publications

Osborne, J.O., Stewart, I.B., Beagley, K.W., & Minett, G. M. (2019). The effect of cycling in the heat on gastrointestinal-induced damage and neuromuscular fatigue. Eur. J. Appl. Physiol, 119(8): 1829-1840. DOI: 10.1007/s00421-019-

04172-z.

Gastrointestinal damage, inflammation and central fatigue during exercise in the heat xvii

Chapter 1: Introduction

Athletes often compete in hot environments, with many sporting events (e.g., 2020

Summer Olympics in Tokyo, Qatar 2022 FIFA World Cup and the annual UCI Tour

Down Under) scheduled during summer months. The combination of exercise and environmental heat is widely acknowledged to be detrimental to athletic performance, and this impairment has been reported since the early 1900s during both laboratory tests and field-based competitions (Dill, Edwards, Bauer, & Levenson, 1931;

Hargreaves, 2008; Lee & Scott, 1916; Nybo, Rasmussen, & Sawka, 2014; Sawka,

Leon, Montain, & Sonna, 2011; Schlader, Stannard, & Mündel, 2011). For example, athletes exercising at a fixed intensity in the heat will reach volitional exhaustion earlier (Galloway & Maughan, 1997), while self-paced tasks result in lower work outputs and therefore, impaired performance (Périard, Cramer, Chapman, Caillaud, &

Thompson, 2011a; Tucker, Marle, Lambert, & Noakes, 2006). However, despite considerable research, the exact causal mechanism(s) behind the development of hyperthermic fatigue is yet to be clearly identified (Cheung & Sleivert, 2004b; Nybo et al., 2014).

One mechanism that has been proposed to explain hyperthermic fatigue emphasises a centrally-mediated downregulation in neuromuscular activation of skeletal muscle

(Nybo et al., 2014). Previous research has demonstrated an inverse association between core temperature and voluntary activation of skeletal muscle, with exercise in the heat observed to diminish central drive to the motor neuron pool (Nybo & Nielsen,

2001a; Thomas, Cheung, Elder, & Sleivert, 2006; Tucker, Rauch, Harley, & Noakes,

2004a). Further, a gradual elevation in core temperature via passive heating has also

Chapter 1: Introduction 1

been associated with a concurrent attenuation in voluntary activation, and reversed upon cooling (Morrison, Sleivert, & Cheung, 2004). Cumulatively, this evidence supports the possibility of a centrally-mediated reduction in neuromuscular activation to explain the occurrence of hyperthermia-induced fatigue, although the causal factor(s) that initiate this development is still undetermined (Cheung & Sleivert,

2004b; Nybo et al., 2014).

The simultaneous stressors of exercise and a hot environment pose a substantial challenge to the cardiovascular system, due to increased demand for blood flow to the exercising muscles, and to the skin for thermoregulation (González-Alonso, Crandall,

& Johnson, 2008). To achieve this required cardiac output, blood flow is diverted from areas of lesser importance, such as the gastrointestinal tract (Lambert, 2008; Wendt, van Loon, & Lichtenbelt, 2007). Reductions in splanchnic blood flow lead to hypoxia and the subsequent death of mucosal epithelial cells, damaging the gastrointestinal tract (van Wijck et al., 2012). Heat also directly affects this intestinal mucosal layer, as a raised core temperature damages cell membranes and stimulates the opening of tight junctions, resulting in increased gastrointestinal permeability (Lambert, 2009).

In thermoneutral conditions, the gastrointestinal tract presents an impermeable barrier to large toxic molecules, such as endotoxins (lipopolysaccharides; LPS), which reside harmlessly within the intestinal lumen (Lim & Mackinnon, 2006). However, as a consequence of hyperthermia-induced gut permeability, endotoxins can translocate through the weakened intestinal barrier and into the portal circulation (Dokladny, Zuhl,

& Moseley, 2016; Marshall, 1998). While the immune system initially scavenges and clears these endotoxins, higher rates of leakage overwhelm this capability and result in systemic circulation of endotoxins, known as endotoxemia (Lambert, 2004; Lim &

Mackinnon, 2006). Circulating endotoxins provoke a strong immune response from

2 Chapter 1: Introduction

the body, eliciting the release of pro-inflammatory cytokines, such as tumour necrosis factor alpha (Heled, Fleischmann, & Epstein, 2013; Lambert, 2009; Lim &

Mackinnon, 2006).

Numerous field- and laboratory-based studies have evidenced a link between strenuous exercise, heat, endotoxemia and inflammatory cytokines (Barberio et al., 2015; Camus et al., 1998; Lambert, 2008; Lim & Mackinnon, 2006; Pires et al., 2016; Selkirk,

McLellan, Wright, & Rhind, 2008). Nevertheless, most of this research has considered endotoxemia and cytokinemia as a trigger for, and contributing to, systemic inflammation and the development of exertional heat stroke (Epstein & Roberts, 2011;

Heled et al., 2013; Lambert, 2004; Lim & Mackinnon, 2006). While the relationship between inflammation and exertional heat stroke cannot be understated, even mild endotoxemia could impair exercise performance via a transient release of inflammatory cytokines (Camus et al., 1997; Jeukendrup et al., 2000). Increases in inflammatory cytokines have been consistently linked with sensations of fatigue and the symptomatic reporting of nausea and malaise (Dantzer, 2004; Robson-Ansley, de

Milander, Collins, & Noakes, 2004; Rohleder, Aringer, & Boentert, 2012). Therefore, cytokines produced during hyperthermic-exercise may result in a manifestation of transient fatigue, similar to the ‘sickness behaviour’ symptoms of a diseased state, such as listlessness, weakness and an inability to concentrate (Dantzer, 2004; Phillips, 2015;

Vargas & Marino, 2014). These pro-inflammatory cytokines have also been reported to alter effort-related motivation and behavioural states (Dantzer, 2001; Dantzer,

Heijnen, Kavelaars, Laye, & Capuron, 2014), which could contribute to an increase in perceived exertion during exercise-heat stress (Nybo & Nielsen, 2001b).

The exact mechanism(s) by which endotoxins and inflammatory cytokines could modulate the CNS remains equivocal, although various pathways have been proposed

Chapter 1: Introduction 3

(Banks & Erickson, 2010; Dantzer et al., 2014). These include: a humoral pathway through the circumventricular organs (Dantzer et al., 2014; Vargas & Marino, 2014); activating a neuro-immune link between afferent peripheral (e.g., vagus) nerves and the CNS (Konsman, Parnet, & Dantzer, 2002); or directly crossing the blood-brain barrier (Banks & Erickson, 2010). Intriguingly, hyperthermia appears to affect the blood-brain barrier similarly to that of the gastrointestinal wall, namely opening tight junctions and therefore increasing permeability (Sharma & Hoopes, 2003).

Furthermore, increased circulating LPS and inflammatory cytokines can modulate the bioavailability of amino acid precursors of certain brain neurotransmitters, namely the impairment of serotonin, dopamine and norepinephrine (Dantzer et al., 2014). This potential neurochemical basis links neatly with previously published hypotheses that proposed the occurrence of central fatigue may result from altered neurotransmitter activity (Cheung & Sleivert, 2004b; Nybo et al., 2014; Roelands & Meeusen, 2010).

Therefore, this thesis proposes an integrative model that mechanistically merges the aforementioned cardiovascular, inflammatory and neural factors. Specifically, it is proposed that prolonged aerobic exercise in the heat could potentially facilitate a centrally-mediated downregulation in muscular activation as a consequence of neuroinflammatory stress, namely exertional-endotoxemia and cytokinemia. This theory suggests a causal link between two independent areas of hyperthermia research; a reduction in central drive during exercise-heat stress and the occurrence of inflammation arising from gastrointestinal barrier dysfunction. While a link between these factors could be considered a novel concept, considerable work has predicated the independent occurrence of CNS fatigue (Morrison et al., 2004; Nybo & Nielsen,

2001a; Thomas et al., 2006; Tucker et al., 2004a) and exertional-endotoxemia

(Lambert, 2008; Lim & Mackinnon, 2006; Pyne, Guy, & Edwards, 2014) during

4 Chapter 1: Introduction

prolonged exercise in a hot environment. Indeed, several authors have briefly speculated that exertional-endotoxemia and the resultant pro-inflammatory cascade of cytokines could influence exercise capacity (Cheung & Sleivert, 2004b; Nybo et al.,

2014; Pyne et al., 2014).

1.1 SUMMARY

A hot environment can impair both gastrointestinal integrity and exercise performance, resulting in inflammation and the development of hyperthermia-induced fatigue, respectively (Lambert, 2008; Nybo et al., 2014; Pires et al., 2016). This PhD thesis proposes a neuroinflammatory model of fatigue, which links elevated core temperatures to reduced central motor drive via an endotoxin-mediated production of pro-inflammatory cytokines. Accordingly, the present research aims to test the relationship between these factors, as well as investigate the interaction of inflammatory markers and exercise-heat stress on the CNS. The efficacy of practical interventions (i.e., glutamine supplementation and heat acclimation training) that might maintain optimal neuromuscular activation and attenuate declines in exercise performance is also investigated. Therefore, the following body of work is comprised of three original studies: 1) a proof of concept investigation into the relationship between hyperthermia, central fatigue and exertional-endotoxemia; 2) efficacy of glutamine supplementation as an intervention to protect the intestinal barrier and attenuate declines in physical performance during heat stress; and 3) effect of a short- duration heat acclimation protocol on intestinal damage, inflammation, and cycling performance.

Chapter 1: Introduction 5

Glutamine supplementation has been reported to provide rapid-acting protection to the gastrointestinal system against endotoxemia and systemic inflammation (Guy &

Vincent, 2018; Wischmeyer, 2008). Previous research has focused on the application of glutamine for clinical patients who have experienced considerable internal trauma, such as severe burns or sepsis (Wischmeyer, 2006). Considerable evidence has demonstrated that glutamine supplementation enhances intestinal integrity and reduces the occurrence of endotoxic shock in these clinical populations (Castell, 2003; Hall,

Heel, & McCauley, 1996; Preiser & Wernerman, 2003). Thus, it stands to reason that glutamine supplementation may also provide similar benefits to athletes competing in a hot environment, namely a reduction in gastrointestinal permeability and attenuation in exertional-endotoxemia and cytokinemia (Guy & Vincent, 2018; Lambert, 2009).

As hyperthermic inflammation could potentially induce central fatigue, a glutamine- mediated reduction in pro-inflammatory cytokines may preserve voluntary activation, and result in improved athletic performance.

Repeated and deliberate exposure to heat stress, a process known as heat acclimation, produces beneficial and protective physiological changes, such as increased sweat rate and a lower heart rate (Sawka et al., 2011). Heat acclimation has been widely researched in the field of exercise physiology and has a strong base of evidence to support integration into the training programs of athletes before undertaking competition in a hot environment (Périard, Racinais, & Sawka, 2015). However, traditional heat acclimation requires a significant period (>7 days) before optimal physiological benefits and adaptations are realised (Racinais, Alonso, et al., 2015).

Given the time constraints of athletes and sports teams, shorter duration (≤ 7 days) heat acclimation training has become popular as a time-efficient method of gaining partial adaptations before competition in the heat (Garrett, Creasy, Rehrer, Patterson, &

6 Chapter 1: Introduction

Cotter, 2012). Notably, only a handful of studies have investigated the effect that short- duration heat acclimation training may have on exertional-endotoxemia, gut damage and exercise performance (Barberio et al., 2015; Guy, Pyne, Deakin, Miller, &

Edwards, 2016; Kuennen et al., 2011). However, all previous studies have methodological issues that limit the applicability of the results to the field. For example, both Kuennen et al. (2011) and Barberio et al. (2015) only utilised fixed- intensity exercise tasks, resulting in limited external validity. While Guy et al. (2016) reported that short duration heat acclimation training had no significant effect on the level of circulating endotoxins, this may have been due to low participant aerobic capacity, as similar performance increases were observed in a workload-matched control group. Further, none of these studies included any measures of neuromuscular function. Accordingly, additional research is required to consider how short duration heat acclimation training could influence exertional-endotoxemia, gut damage, central fatigue and exercise performance in the heat.

The three studies contained in this thesis provide original data regarding neuromuscular fatigue and gastrointestinal damage during exercise in the heat. Further, these studies also present the results of two separate prophylactic interventions on hyperthermic exercise performance, namely glutamine supplementation and heat acclimation training.

Chapter 1: Introduction 7

Chapter 2: Literature Review

2.1 THERMOREGULATION DURING EXERCISE

As endothermic beings, humans regulate and maintain core temperature within a narrow range, regardless of the external environment (Tansey & Johnson, 2015). Due to the inefficiency of metabolic oxidation, active skeletal muscles generate a large amount of heat when completing work, which the body must remove to maintain thermoregulatory homeostasis (Casa, 1999; Sawka, Wenger, & Pandolf, 2010). Blood flow transfers heat from the core to the skin, and four main mechanisms transmit heat to the external environment: convection, conduction, radiation and evaporation

(Tansey & Johnson, 2015). In thermoneutral environments, these mechanisms release heat to, or gain heat from, the surroundings and maintain body core temperature in homeostasis (Casa, 1999). However, exercise in a hot and humid environment poses a greater thermoregulatory challenge, as the effectiveness of evaporative heat loss mechanisms are reduced (Maughan, Otani, & Watson, 2012; Sawka et al., 2010).

Indeed, evaporation of sweat is the sole mechanism for heat dissipation once ambient temperature rises above body temperature (Wendt, van Loon, & Marken Lichtenbelt,

2007b). Accordingly, sweat production and skin blood flow are dramatically elevated in an attempt to attenuate the increased thermal load (Sawka et al., 2010; Tansey &

Johnson, 2015). This increased pooling of blood in the compliant venous beds of the skin improves heat transfer to the external environment. However, this has a deleterious effect on central blood volume, arterial pressure and therefore cardiac filling, necessitating an increased heart rate to maintain cardiac output (Nybo et al.,

2014).

8 Chapter 2: Literature Review

Heightened sweat rate and associated fluid loss, particularly from plasma volume, also negatively decreases cardiac output (Nybo, 2008; Sawka et al., 2010). This is an additional burden on an already overloaded cardiovascular system, which must also meet the competing demands for blood flow to active skeletal muscles and peripheral vascular beds to maintain thermoregulatory control (Casa, 1999; González-Alonso et al., 2008). To provide additional cardiac output, heart rate is elevated, and blood flow is redirected from splanchnic tissues to active muscles (Nybo et al., 2014; Sawka et al., 2011). This reduction in splanchnic blood flow has been observed to reach more than 50% after even short duration (10-15 min) sub-maximal and maximal exercise

(80-100% HRmax) (Otte, Oostveen, Geelkerken, Groeneveld, & Kolkman, 2001; Qamar

& Read, 1987), resulting in severe hypoxia and subsequent death of intestinal epithelial cells in extreme cases (Selkirk et al., 2008).

2.2 EXERCISE PERFORMANCE IN THE HEAT

Exercise in a hot environment has been repeatedly shown to prematurely reduce aerobic performance compared to cool conditions (Chan, Wong, & Chen, 2008;

Galloway & Maughan, 1997; Peiffer & Abbiss, 2011; Périard, Cramer, et al., 2011a;

Sawka et al., 2011; Tatterson, Hahn, Martini, & Febbraio, 2000; Tucker et al., 2004a).

Many models have been proposed to explain the occurrence of hyperthermia-induced fatigue (Cheung, 2007; Cheung & Sleivert, 2004b; Nybo et al., 2014; Schlader,

Stannard, et al., 2011), though all highlight the greater physiological demands incurred as being detrimental to sustained performance. This diminished exercise performance has been observed in both real-world athletic competitions, as well as in controlled laboratory environments (Marino et al., 2000; Mountjoy et al., 2012; Périard, Cramer,

Chapman, Caillaud, & Thompson, 2011b; Tatterson et al., 2000; Tucker et al., 2004a).

Chapter 2: Literature Review 9

An example of the detrimental effect that heat stress has on exercise performance was reported by Tatterson et al. (2000), who found that elite cyclists had a 6.5% lower power output in the heat (32 °C) than in a temperate environment (23 °C). A similar response has also been observed in studies involving cyclists undertaking 20, 40 and

100 km time trials in the heat, with reductions in power output, and therefore performance, when compared to identical trials in thermoneutral environments

(Abbiss et al., 2010; Périard, Cramer, et al., 2011a; Tucker et al., 2004a). Likewise,

Tucker et al. (2006) found participants exercising at a ‘fixed’ rating of perceived exertion in a hot environment significantly reduced their power output compared to cooler temperatures. This attenuation in exercise intensity due to hot environmental conditions has been proposed to occur in response to an elevated body temperature

(Flouris & Schlader, 2015), which is the primary driving factor behind perception of effort, thermal sensation and comfort; although additional factors such as skin wetness and skin temperature may also potentially modulate these perceptual variables

(Schlader, Simmons, Stannard, & Mündel, 2011; Schlader & Vargas, 2019).

Behavioural thermoeffectors play a critical role in the regulation of self-paced exercise intensity in the heat, as increased perceptions of thermal sensation and effort appear to downregulate work-rate, even before a significant elevation in core temperature

(Flouris & Schlader, 2015). This reduction in exercise intensity has been proposed to occur as a protective anticipatory mechanism, to avoid hyperthermia and the possibility of a catastrophic physiological collapse due to extreme heat stress (Marino,

2004a; Tucker et al., 2006). During self-paced exercise in the heat, an increased perception of effort results in behavioural alterations and the implementation of a

‘positive’ pacing strategy, whereby the athlete’s work distribution gradually declines throughout the event (Abbiss & Laursen, 2008). However, the prevalence of exertional

10 Chapter 2: Literature Review

heatstroke in athletic competitions suggests this protective self-regulatory mechanism can be potentially overridden (Nichols, 2014). Interestingly, multiple studies have shown that athletes often increase their speed in the latter section of an event, resulting in a curved pacing strategy, and indeed this has also been observed in hot environmental conditions (Borg et al., 2018; Peiffer & Abbiss, 2011; Roelands, de

Koning, Foster, Hettinga, & Meeusen, 2013; Tucker et al., 2004a).

In summary, the evidence from this previous research unequivocally supports the notion that hot environments detrimentally impact the performance of self-paced exercise (Roelands et al., 2013). This impairment of exercise performance due to hyperthermia is recognised as an example of fatigue, a concept that is widely acknowledged throughout exercise physiology literature (Cheung & Sleivert, 2004b;

Marino, 2004b; Nybo et al., 2014; Schlader, Stannard, et al., 2011). The various factors that could potentially contribute to the development of acute fatigue will be presented in more detail in the following chapter.

2.3 MODELS OF FATIGUE IN THE HEAT

Fatigue is a commonly used term in many fields of research; however, the definition of the concept varies and is specific to the particular scientific discipline. Researchers of the exercise sciences consider fatigue to be characterised by an acute impairment in performance, which leads to a reduction in required force production, resulting in volitional exhaustion (Ament & Verkerke, 2009; Schlader, Stannard, et al., 2011). The development of exercise-induced fatigue has been suggested to be a protective mechanism which attempts to avoid homeostatic failure (Ament & Verkerke, 2009;

Walters, Ryan, Tate, & Mason, 2000). As exercise in a hot environment places additional stress on maintaining thermoregulatory homeostasis, fatigue occurs more quickly and exercise performance is reduced (Cheung & Sleivert, 2004b; Nybo et al.,

Chapter 2: Literature Review 11

2014; Schlader, Stannard, et al., 2011). However, the potential underlying mechanisms which trigger this onset of fatigue during exercise in the heat are currently inconclusive

(Nybo et al., 2014). A broad array of theories have been developed, which are often viewed as mutually exclusive, although some authors have attempted to reconcile these hypotheses into a combined, multifaceted approach (Cheung, 2007; Cheung &

Sleivert, 2004b). The following sections will discuss the critical core temperature hypothesis and mechanisms that may influence the development of hyperthermic fatigue: 1) cardiovascular factors and 2) central nervous system factors.

2.3.1 Critical core temperature theory

An interesting outcome from an early study into exercise tolerance and heat acclimation by Nielsen et al. (1993), found that voluntary exhaustion occurred at a similar core temperature, both before and after acclimation intervention. Additional research supported the possibility of a consistent endpoint, with several studies reporting a cessation of exercise at a similar core temperature, regardless of initial or final core temperature, or heat storage rate (Fuller, Carter, & Mitchell, 1998; González-

Alonso, Teller, et al., 1999; Walters et al., 2000). Evolving from this research, a critical temperature hypothesis was developed, which suggested that volitional exhaustion at a specific internal temperature safeguards the body against hyperthermic damage.

The ‘critical internal temperature’ set point theory was initially championed as the origin for central fatigue during exercise in the heat (González-Alonso, Teller, et al.,

1999). However, conflicting evidence has revealed that certain athletes can continue to exercise with core temperatures beyond the hypothesised ‘critical’ set-point of ~40

°C (Byrne, Lee, Chew, Lim, & Tan, 2006; Ely et al., 2009; Racinais, Périard, Karlsen,

& Nybo, 2015). Furthermore, an elegant study by Morrison et al. (2004) demonstrated an inverse relationship between muscular force and core temperature, with a

12 Chapter 2: Literature Review

progressive reduction in force occurring as core temperature increased. This gradual impairment, as opposed to a sudden termination at a critical core temperature, suggests that alternative factors may also influence the development of hyperthermic fatigue.

In a recent editorial, Nybo and González-Alonso (2015) labelled the critical core temperature theory as ‘over-simplistic’, and highlighted that variables such as training status (Cheung & McLellan, 1998) or dehydration (González-Alonso, Calbet, &

Nielsen, 1999) could alter aforementioned ‘critical’ termination temperature. The authors concluded by urging researchers to consider alternative mechanisms which may incur central fatigue (Nybo & González-Alonso, 2015).

In a similar fashion to constant work protocols, self-paced exercise in the heat has also been found to result in decreased power output and motor recruitment (Marino et al.,

2000; Tatterson et al., 2000; Tucker et al., 2006; Tucker et al., 2004a). Self-paced exercise protocols, as opposed to fixed-intensity, arguably better represent real-world competitive events, as it allows the athlete to self-regulate their own pace (Marino,

2004b). Therefore, a self-paced protocol in a hot environment will result in an athlete making behavioural power output adjustments to ensure the exercise task can be completed without the occurrence of volitional exhaustion (Cheung, 2007). Several authors have proposed that pacing alterations during exercise in the heat may occur due to anticipatory down-regulation in work output (Abbiss & Laursen, 2008;

Schlader, Stannard, et al., 2011). Specifically, a complex system of feedforward and feedback loops may integrate afferent input, from both environmental and thermal stressors, and regulate pacing to ensure task completion without hyperthermic failure

(Abbiss & Laursen, 2008; Marino, 2004a; Marino, 2004b).

The simultaneous development of these two thermoregulatory paradigms, critical core temperature and anticipatory regulation, were initially viewed as mutually exclusive

Chapter 2: Literature Review 13

(Cheung, 2007; Marino, 2004a). However, Cheung (2007) proposed an integrative approach, suggesting both models may exist on a fatigue continuum. Although self- mediated reductions in work output would normally maintain thermoregulatory homeostasis during exercise, volitional exhaustion would occur as a physiological safety switch upon reaching a critical core temperature (Schlader, Stannard, et al.,

2011). While this provides a novel integrative approach, research has repeatedly reported core temperatures of greater than 40 °C (usually involving competitive, highly motivated athletes), demonstrating that this ‘terminal’ thermoregulatory endpoint can be potentially overridden (Maron, Wagner, & Horvath, 1977; Racinais,

Périard, et al., 2015). This evidence directly conflicts with the critical core temperature hypothesis, specifically the ability to exercise without volitional exhaustion, despite high core temperatures.

Ultimately, there is a significant basis of research to support a centrally-mediated downregulation of skeletal muscle activation during exercise in the heat; however, the processes which trigger this mechanism are currently equivocal (Nybo et al., 2014).

2.3.2 Cardiovascular factors

As previously stated, researchers have long attempted to identify the exact mechanisms that limit physical performance in the heat. The classic cardiovascular model proposes that failure of the heart to maintain required cardiovascular output may be the key fatigue mechanism to explain the reduction in performance observed during heat-stress

(Nybo, 2008). This theory is supported by the increased cardiovascular strain experienced during exercise in hot environments (González-Alonso et al., 2008;

Sawka et al., 2011), due to the competing demand for blood supply to either the active skeletal muscle for metabolism or to the skin for thermoregulation (Casa, 1999; Nybo et al., 2014).

14 Chapter 2: Literature Review

To regulate body temperature during exercise-heat stress, skin blood flow and venous compliance dramatically increase, leading to a larger peripheral blood volume and cutaneous transit time, maximising potential heat exchange with the environment

(Nybo et al., 2014; Rowell, 2011). However, increased cutaneous blood volume leads to reductions in central blood volume, venous pressure and cardiac filling, lowering stroke volume (Rowell, Marx, Bruce, Conn, & Kusumi, 1966). In an attempt to maintain adequate cardiac output, heart rate increases, further shortening the cardiac cycle and additionally compromising stroke volume (Fritzsche, Switzer, Hodgkinson,

& Coyle, 1999; González-Alonso et al., 2008). An increased heart rate can initially maintain cardiac output by offsetting the reduced stroke volume (Sawka et al., 2011).

Nevertheless, prolonged submaximal exercise with significant heat stress results in a continue rise in heart rate, termed ‘cardiovascular drift’, reaching a maximal limit and subsequent voluntary cessation of exercise (Nybo et al., 2014; Sawka et al., 2011).

Dehydration further exacerbates cardiovascular drift by reducing the stroke volume of each cardiac cycle, and subsequently results in a diminished cardiac output (González-

Alonso, Mora-Rodríguez, Below, & Coyle, 1997). The reduction in central blood volume during exercise in the heat is also thought to result in unloading of baroreceptors (Wendt, van Loon, & Marken Lichtenbelt, 2007a) and may lead to a reduction in sweating and skin blood flow in an effort to defend blood pressure. While considerable research has investigated the complex relationship between heat and the baroreflex regulation of blood pressure and possible modulation of heart rate, the findings are currently unclear and require further investigation (Crandall, 2008;

Crandall & González-Alonso, 2010).

The physiological cascade in response to exercise in the heat, a ‘cardiovascular limit’, was initially proposed to explain the decrease in aerobic performance in hot

Chapter 2: Literature Review 15

environments (Cheung, 2007). During high-to-maximal intensity exercise this mechanism appears to hold true; demand for oxygenated blood by the active skeletal muscle exceeds the capacity of the cardiovascular system, leading to impaired oxygen delivery to musculature and fatigue (Galloway & Maughan, 1997; González-Alonso

& Calbet, 2003; Nybo et al., 2014; Tatterson et al., 2000). Conversely, submaximal exercise necessitates a lower cardiac output, which is achieved via heart rate elevation and redistribution of splanchnic, renal and gastrointestinal blood flow (Rowell,

Blackmon, Martin, Mazzarella, & Bruce, 1965; van Wijck et al., 2012). Although cardiac output is adequately maintained during submaximal exercise in the heat, aerobic performance has still been observed to decrease (Nielsen et al., 1993; Périard,

2012). As inadequate cardiovascular output may not be the primary limiting factor during shorter-duration submaximal exercise in the heat, alternative mechanisms that may drive the development of hyperthermic fatigue should also be considered (Nybo,

2008).

In conclusion, the occurrence of fatigue during maximal exercise in the heat appears to be due to a limitation of the cardiovascular system which is unable to maintain required blood flow, whereas a different mechanism(s) is potentially responsible for the premature fatigue seen during submaximal exercise in the heat.

2.3.3 Central nervous system factors

Alterations in the central nervous system (CNS) function have also been proposed as a potential causal mechanism behind the development of hyperthermic fatigue (Nybo et al., 2014). Central fatigue describes a failure in force production which occurs

‘centrally’ in either the brain or spinal column, minimising voluntary activation, and therefore force production of skeletal muscle (Davis, 1995; Kent-Braun, 1999; Nybo

& Nielsen, 2001a). To test if the loss of force production is occurring due to central

16 Chapter 2: Literature Review

fatigue, the peripheral mechanisms can be isolated and independently stimulated via an electrical pulse during a maximum voluntary contraction (MVC) (Shield & Zhou,

2004). The difference between the suprastimulated force and the voluntary force can then quantify if the fatigue is occurring centrally or distal to the neuromuscular junction (Bigland-Ritchie, Jones, Hosking, & Edwards, 1978).

In a novel study, Nybo and Nielsen (2001a) suggested that hyperthermic fatigue may originate from the CNS. Participants cycled in the heat (40 °C) until exhaustion (~50 min) or for 60 min in a thermoneutral (18 °C) environment, before a sustained MVC protocol. The researchers found an inverse relationship between core temperature and neuromuscular performance (i.e., maximal voluntary force production and voluntary activation) in the hot environment. However, supramaximal electrical stimulation of the muscle elicited force similar to that of the thermoneutral condition, indicating a comparable level of peripheral, but not central, fatigue. A similar outcome was observed for a handgrip task, despite this muscle group being comparatively inactive during the preceding cycling activity, thus ruling out muscle temperature as a potential influencing factor. A limitation of this study is the use of exercise, as opposed to passive heating, to induce hyperthermia, due to the different cardiovascular and thermoregulatory responses. Périard, Caillaud, and Thompson (2011) addressed this limitation by comparing neuromuscular function during passive and exercise-induced hyperthermia. Voluntary activation was similarly reduced in both conditions, although the exercise task resulted in a greater impairment in force, which was suggested to stem from increased peripheral fatigue associated with exercise-induced alteration of muscle contractile properties. This study demonstrated that central fatigue appears to develop regardless of the source of the hyperthermia.

Chapter 2: Literature Review 17

The outcomes of these research projects evidenced a CNS-mediated decline in voluntary activation as a potential mechanism behind the occurrence of fatigue during submaximal exercise in the heat (Nybo & Nielsen, 2001a). Similarly, participants who were passively heated, but had one leg held at a thermoneutral temperature, still reported a decrease in voluntary contraction and activation of both legs (Thomas et al.,

2006). This evidence supports the possibility of a centrally-mediated impairment of neuromuscular activation, independent of muscle temperature. Interestingly, Périard,

Cramer, et al. (2011b) found that although exercise performance in the heat was decreased, both hot and thermoneutral environments resulted in a similar reduction in force production and voluntary activation following a 40 km time trial, during both brief and sustained MVCs. The authors suggest that impaired contractile properties of hot skeletal muscle could potentially explain the reduction in post-exercise force.

Conversely, Saboisky, Marino, Kay, and Cannon (2003) found no reduction in central activation ratio in a non-exercised muscle, suggesting that afferent feedback from a hot muscle may decrease voluntary activation as a safety mechanism.

Building upon the earlier research by Nybo and Nielsen (2001a), a number of studies have confirmed that hyperthermia appears to contribute to the occurrence of central fatigue, via a CNS-mediated reduction in the activation of skeletal muscle (Morrison et al., 2004; Thomas et al., 2006; Tucker et al., 2004a). Subsequent research has revealed hyperthermia-induced alterations in muscle contractile properties, such as faster relaxation rates (Todd, Butler, Taylor, & Gandevia, 2005), which theoretically would require higher motor unit firing rates to produce tetanic contraction. Therefore, it has been proposed that a failure of the CNS to sufficiently maintain activation of hot skeletal muscle may explain the reduction in voluntary force during hyperthermia

(Périard, Caillaud, et al., 2011; Todd et al., 2005). These results were supported by

18 Chapter 2: Literature Review

recent research using both passive-heating and exercise-induced hyperthermia, where high temperatures increased muscle relaxation rate (Périard, Christian, Knez, &

Racinais, 2014). Arguably, an upregulation of motor unit activity could meet increased demand during a brief contraction, however, sustained contraction results in decrements in force output (Périard et al., 2014).

A limitation for the majority of these studies is the use of MVC and twitch interpolation techniques to assess available descending drive, as this has arguably limited real-world validity and requires maximal participant effort (Shield & Zhou, 2004). The combination of heat stress and exhaustive exercise before sustained MVCs may reduce participant motivation, and therefore decrease CNS drive and motor neurone firing

(Enoka & Stuart, 1992). Conceivably, this lack of motivation could partially explain the reduced levels of central drive seen in MVC protocols following exercise in a hot environment (Nybo & Nielsen, 2001a; Saboisky et al., 2003). However, the importance of providing encouragement and feedback to participants’ during MVCs is standard practice, even if not explicitly detailed in the literature (Gandevia, 2001;

Shield & Zhou, 2004).

In conclusion, hyperthermic conditions appear to elicit the development of central fatigue, observable via a reduction in voluntary force. However, it is unclear if this centrally-mediated downregulation of muscle activation entirely accounts for the decrease in exercise performance, and the driving factor/s which initiate this procedure is still, as yet, unclear.

2.4 NEUROINFLAMMATORY FATIGUE

Hyperthermic exercise, resulting in increased gastrointestinal (GI) permeability and subsequent inflammation, has been proposed to potentially induce central fatigue

Chapter 2: Literature Review 19

(Pyne et al., 2014; Vargas & Marino, 2014). While speculative, a ‘neuroinflammatory fatigue model’ that may occur in hyperthermic environmental conditions has considerable implications for influencing exercise performance in the heat (Lambert,

2008; Pyne et al., 2014; Vargas & Marino, 2016). How the rise in exertional endotoxemia, and the resultant production of cytokines, may influence neuromuscular activation is currently equivocal. However, the identification and evaluation of potential intervention strategies to address this issue may prove critically important.

2.4.1 Exertional endotoxemia

A key difference between exercise in a hot, as opposed to a thermoneutral, environment, is the increased demand for cardiac output (Casa, 1999; González-

Alonso et al., 2008). To achieve this heightened circulatory requirement, blood flow is redirected from splanchnic regions, resulting in inadequate blood supply to the gastrointestinal system and death of mucosal epithelial cells (Lambert, 2008; van

Wijck et al., 2012; van Wijck et al., 2011). Heat also directly affects this mucosal layer, as elevated internal temperatures damage cell membranes and open tight junctions, resulting in increased GI permeability (Lambert, 2009; Pires et al., 2016). Indeed a systematic review of the literature concluded that a core temperature of at least 38.5

°C is strongly associated with a rise in intestinal permeability; and a core temperature that exceeds 39.0 °C is always associated with augmented permeability (Pires et al.,

2016). A consequence of this increased GI damage and permeability is the translocation of highly toxic molecules, called endotoxins, through the weakened intestinal barrier and into circulation, as seen in Figure 2.1 (Lim & Mackinnon, 2006;

Marshall, 1998).

While low levels of endotoxins can be removed via specialised macrophages in the liver, higher rates of leakage overwhelm these defences and spill into central

20 Chapter 2: Literature Review

circulation, causing a medical condition known as endotoxemia (Andreasen et al.,

2008; Lim & Mackinnon, 2006). Unlike clinical endotoxemia experienced by patients with septic shock, exercise-induced endotoxemia expresses milder and transient symptoms. That said, the two conditions share similar symptoms, such as shivering, nausea and diarrhoea (Lambert, 2008; Lim & Mackinnon, 2006).

Figure 2.1. The effect of hypoxia and hyperthermia on intestinal barrier permeability during exercise in the heat. Reprinted from “Role of Gastrointestinal Permeability in Exertional Heatstroke,” by G. P. Lambert, 2004, Exercise and Sport Sciences Reviews, 32(4), pg. 186.

The entry of endotoxins, also known as lipopolysaccharides (LPS), into the bloodstream triggers a strong host immune response and the resultant production of

Chapter 2: Literature Review 21

inflammatory signalling molecules, known as cytokines (Andreasen et al., 2008; Lim

& Mackinnon, 2006; van Deventer et al., 1990). Cytokines can be synthesised by a range of differing cell types and are utilised as intercellular messengers at both a local as well as a systemic level (Peake, Della Gatta, Suzuki, & Nieman, 2015).

Considerable evidence has identified that exercise induces the systemic release of various cytokines, such as tumour necrosis factor alpha (TNF-α), interleukin 1 beta

(IL-1β) and interleukin 6 (IL-6), although the duration, intensity and modality of the exercise task appears to modulate cytokine kinetics (Suzuki, 2018). Normative values of common cytokines are provided in Table 2.2.

Endotoxins, as well as several different inflammatory cytokines, have been observed to interfere with muscular contractile properties and force-generation in both respiratory and limb muscles (Supinski & Callahan, 2007) (Callahan, Nethery, Stofan,

DiMarco, & Supinski, 2001; Shindoh, Dimarco, Nethery, & Supinski, 1992).

Specifically, administration of exotoxins to skinned muscle fibres has been reported to affect the sarcolemma membrane and sarcoplasmic reticulum (Supinski et al., 2000), which could possibly explain the previously observed impairment in force-generating capacity (Supinski, Nethery, Stofan, & DiMarco, 1996). An endotoxin-mediated increase in oxygen free-radicals also appears to be a possible causal factor behind the observed reduction in force generation, and can be reversed upon the administration of free-radical scavengers (Supinski & Callahan, 2007). The presence of endotoxins may also induce the release of inflammatory cytokines, such as TNF-α, which have been observed to impair muscle contractile function by attenuating the myofilament response to calcium (Reid, Lännergren, & Westerblad, 2002). Although these previous studies have exclusively utilised in-vitro or in-vivo animal models, this research

22 Chapter 2: Literature Review

collectively supports the concept that endotoxins and/or inflammatory cytokines may directly impair muscle force-generation during exertional-endotoxemia.

Inflammatory cytokines, such as TNF-α, IL-1β and IL-6, have also been consistently linked with fatigue and implicated in signalling the brain to produce symptoms of sickness, such as nausea, elevated body temperature and malaise (Dantzer, 2004;

Rohleder et al., 2012). Vargas and Marino (2014) postulated that the production of cytokines during exercise, in particular, IL-6, may result in a manifestation of transient fatigue, similar to the response during a diseased state. This could be due to the indiscriminate nature of IL-6 receptors, which exert a similar response regardless of cytokine origin (Taga & Kishimoto, 1997; Vargas & Marino, 2014). Although speculative, research has suggested that inflammatory cytokines could potentially modulate the CNS through afferent feedback via peripheral nerves (e.g., vagus nerve) or circumventricular organs (Dantzer, 2004; Szelényi, 2001; Vargas & Marino, 2014).

How this may relate to the cytokine release following exertional endotoxemia, and implications for an altering of exercise performance via central mechanisms remains unclear.

The link between strenuous exercise in the heat, intestinal permeability and endotoxemia was first observed by Brock-Utne et al. (1988). The authors reported that

81% of randomly-selected athletes in the medical tent following an ultra-distance marathon could be classified as endotoxemic (≥5 pg·mL-1 of LPS in the blood). Further research by this same group seemed to confirm these findings, and other investigators reported similar increases in endotoxemia for marathon runners as well as triathletes competing in strenuous, long distance competitions in the heat (Bosenberg, Brock-

Utne, Gaffin, Wells, & Blake, 1988; Camus et al., 1997; Jeukendrup et al., 2000).

Exertional-endotoxemia was proposed to arise from an exercise-mediated reduction in

Chapter 2: Literature Review 23

splanchnic blood flow, resulting in intestinal barrier dysfunction and permitting the translocation of endotoxins into systemic circulation (Lim & Mackinnon, 2006; van

Wijck et al., 2011). Relevant support for this hypothesis was provided by Øktedalen,

Lunde, Opstad, Aabakken, and Kvernebo (1992), who reported that runners had increased intestinal permeability following participation in a marathon, and similar results were also reported for triathletes in ultra-distance competitions (Lambert et al.,

1999).

The majority of these early, exertional-endotoxemia studies were field-based, and data was collected from trained athletes competing in a sporting event (Bosenberg et al.,

1988; Brock-Utne et al., 1988; Camus et al., 1998; Camus et al., 1997; Jeukendrup et al., 2000; Lambert et al., 1999; Øktedalen et al., 1992). This made it impossible to control for the environmental conditions of humidity and temperature, both of which potentially play a pivotal role in the development of exertional endotoxemia. However, more recent studies in this area have primarily focused on laboratory-based testing of trained participants in climatic chambers, as can be seen in Table 2.1 (Gill, Teixeira, et al., 2015; Lim et al., 2009; Marchbank et al., 2011; Ng, Lee, Byrne, Ho, & Lim,

2008; Selkirk et al., 2008; Shing et al., 2014; van Wijck et al., 2011; Yeh, Law, & Lim,

2013).

24 Chapter 2: Literature Review

Table 2.1. Summarised exercise and heat-stress research with measured outcomes of LPS, inflammatory cytokines, GI permeability and damage.

Exercise Exercise Environment LPS GI Perm. or Inflammatory Study Sample Fitness level Modality Duration Intensity (°C; % RH) levels Damage Cytokines

(Pugh, Sage, et Recreationally-trained ↑ 241% (perm.); 10 Running (treadmill) 60 min 70% VO 30 °C, 40-45% - - al., 2017) (52 ± 6 mL·kg-1·min-1) 2max ↑70% (damage)

(Pugh, Impey, Trained 18x 400m ↑ 59% (perm.); 11 Running 120% VO 30 °C, 40-45% - - et al., 2017) (60 ± 3 mL·kg-1·min-1) sprints 2max ↑72% (damage)

(Davison, Marchbank, March, Recreationally-trained 8 Running (treadmill) 20 min 80% VO - - ↑ 208% (perm.) - Thatcher, & (60 ± 2 mL·kg-1·min-1) 2max Playford, 2016) 3x 10 min (Guy et al., Trained 50, 60, 70% No 24 Cycling (ergometer) cycle; 5 km 35 °C; 70% - ↑ ~300% (IL-6) 2016) (45 ± 5 mL·kg-1·min-1) VO ; Race change cycle 2max

(Zuhl et al., Trained 30 °C; 12- 7 Running (treadmill) 60 min 70% VO ↑ 3% ↑ 177% (perm.) No change (TNF-α) 2015) (52 ± 7 mL·kg-1·min-1) 2max 20%

Trained ~24 - 26 min (Barberio et 41 °C; 40- ↑ 50 - ↑ 48% (IL-6); 8 (55 ± 74 mL·kg-1·min- Running (treadmill) (until T ≥ +2 78% VO ↑ 200% (damage) al., 2015) c 2max 41% 67% No change (TNF-α) 1) °C rest) ↑ 3436% (IL-6); (Gill, Hankey, 0-20 °C, 54- 17 Trained 24 h ultra-marathon 122-208 km Race ↑ 37% - ↑ 35% (TNF-α); et al., 2015) 82% RH ↑ 332% (IL-1β) ↑ 13 - 240% (IL-6); (Gill, Teixeira, Multi-day ultra- 30 – 40 °C; 19 Trained 230 km total Race ↑ 3 – 21% - ↑ 25 - 103% (TNF-α); et al., 2015) marathon 31 - 40% ↑ 17 - 67% (IL-1β)

(Zuhl et al., Trained 65 – 70% 30 °C; 12- 8 -1 -1 Running (treadmill) 60 min - ↑ 177% (perm.) - 2014) (51 ± 7 mL·kg ·min ) VO2max 20%

Chapter 2: Literature Review 25

Voluntary 80% ↑ 77% (IL-6); (Shing et al., Trained 10 Running (treadmill) Exhaustion ventilatory 35 °C; 40% ↑ 24% - No change (TNF-α); 2014) (63 ± 2 mL·kg-1·min-1) (~33 min) threshold 7 Trained (Morrison, (64 ± 4 mL·kg-1·min- 90 min (15 ↑563% (trained); 15 50 – 80% HR Cheung, & 1); Running (treadmill) cycle; 2x 30 30 °C; 50% - ↑177% ↑ ~400% (IL-6); (7 + 8) Reserve Cotter, 2014) 8 Untrained run; 15 min) (untrained) (46 ± 4 mL·kg-1·min-1)

(Yeh et al., Recreationally-trained 33 °C; 50% & ↑ 54% ↑ 24% (hot); 15 Running (treadmill) 60 min 70% VO - 2013) (49 ± 3 mL·kg-1·min-1) 2max ~25 °C; 60% (hot) ↑ 13% (cool)

50 min, 10 (Kuennen et Recreationally-trained 46 - 47 °C; 20 No No change ↑256 - 616% (IL-6); 7 Running (treadmill) min rest, 50 50% VO al., 2011) (56 ± 2 mL·kg-1·min-1 2max - 22% change (perm.) No change (TNF-α) min

(Marchbank Recreationally-trained 12 Running (treadmill) 20 min 80% VO - - ↑ 142% (perm.) - et al., 2011) (53 ± 7 mL·kg-1·min-1 2max

(van Wijck et 70% No ↑ ~139% (perm.); 20 Recreationally-trained Cycling (ergometer) 60 min - - al., 2011) Workloadmax change ↑ 99% (damage)

Voluntary (Lim et al., Trained 35 ± 1 °C; ↑ 719% (IL-6); 18 Running (treadmill) Exhaustion or 70% VO ↑71-92% - 2009) (64 mL·kg-1·min-1) 2max 40 ± 2% ↑ 33% (TNF-α) Tc ≥ 39.5 °C ↑ 65% (IL-6); (Ng et al., Trained 21.1 km ~26.5 °C; 30 Half marathon Race ↑32% - No change (TNF-α); 2008) (59 ± 4 mL·kg-1·min-1) (107 ± 9 min) ~83.5% No change (IL-1β)

(Lambert et Trained 20 Running (treadmill) 60 min 70% VO 24 °C; 33% - ↑ 79% (perm.) - al., 2008) (56 ± 2 mL·kg-1·min-1) 2max

↑ 187% 23 Voluntary (Trained); (Selkirk et al., 12 Trained; 11 Walking in NBC 4.5 km/h, 2% ↑ 1530% (IL-6); (12 + Exhaustion 40 °C; 30% ↑ 247% - 2008) Untrained suit (treadmill) slope ↑ 25% (TNF-α) 11) (~134 min) (Untraine d) (Lambert, Trained 23.2 ± 0.1 °C; No change Boylan, 8 Running (treadmill) 60 min 70% VO - - (61 ± 6 mL·kg-1·min-1) 2max 36 ± 1% (perm.) Laventure, 26 Chapter 2: Literature Review

Bull, & Lanspa, 2007)

(Nieman et al., 11 - 25 °C; No ↑ 4761% (IL-6); 25 Trained Ultra- marathon 160 km Race - 2006) 56% change No change (TNF-α)

(Ashton et al., Recreationally-trained Voluntary 10 Cycling (ergometer) Incremental - ↑71% - - 2003) Exhaustion (Lambert, Broussard, Mason, Trained 17 Running (treadmill) 60 min 70% VO 22.4 °C; 48% - ↑40% (perm.) - Mauermann, (62 ± 2 mL·kg-1·min-1) 2max & Gisolfi, 2001) 3.8 km swim, (Jeukendrup Ultra-distance ↑ 3129% (IL-6); 29 Trained 180 km ride, Race 9.4 - 32.1 °C ↑ et al., 2000) triathlon No change (TNF-α) 42 km run 3.9km swim, (Lambert et Ultra-distance 41 Trained 185 km cycle, Race - - ↑ 427% (perm.) - al., 1999) triathlon 42.2 km run

(Smetanka et No change 34 Trained Marathon 42.2 km Race - - - al., 1999) (perm.)

1.5km swim, ↓ anti- (Camus et al., 14 Trained Triathlon 40km cycle, Race - LPS (↑ - ↑ 55% (TNF-α) 1998) 10km run LPS) (Pals, Chang, Recreationally-trained 40, 60 & ↑ 123% (perm.) Ryan, & 6 Running (treadmill) 60 min 22 °C; 50% - - (57 ± 2 mL·kg-1·min-1) 80% VO @80% VO Gisolfi, 1997) 2peak 2peak

(Camus et al., ↑ 7172% (IL-6); 18 Trained Marathon 42.2 km Race - ↑ (n = 9) - 1997) ↑ 72% (TNF-α)

Half marathon (n = (Øktedalen et 21.1 & 42.2 9 + 8 Trained 10); marathon (n = Race - - ↑166% (perm.) al., 1992) km 6)

Chapter 2: Literature Review 27

3.2 km swim, (Bosenberg et 18 Trained Triathlon 140 km cycle, Race 23.1 – 22.3 °C ↑ 263% - - al., 1988) 42.2 km run

(Brock-Utne et 20.3-22.3 °C 89 Trained Ultra- marathon 89.4 km Race ↑ - - al., 1988) WBGT

RH = relative humidity; perm. = GI permeability; IL-6 = interleukin-6; TNF-α = tumour necrosis factor α; IL-1β = interleukin-1β; NBC = nuclear, biological and chemical protective suit.

28 Chapter 2: Literature Review

Table 2.2. Cytokine reference values from previous exercise-heat stress and endotoxemia research.

Marker Reference Normative Response Origin Study Pre-exercise values Post-exercise values

Various: IL-1β (Dinarello, 1997) Pro-inflammatory Skeletal Muscle

(Gill, Hankey, et al., 0.1 (0.0 to 0.3) pg∙mL-1 0.6 (0.1 to 1.1) pg∙mL-1 2015)

(Gill, Teixeira, et al., 0.6 ± 0.3 pg∙mL-1 1.0 ± 0.3 pg∙mL-1 2015)

Marker Reference Normative Response Origin Study Pre-exercise values Post-exercise values

Various: (Tanaka, Narazaki, & Pro-inflammatory and/or IL-6 Skeletal muscle; Kishimoto) Ani-inflammatory Brain.

(Guy et al., 2016) ~1.5 ± 0.7 pg∙mL-1 ~6 ± 2.5 pg∙mL-1

(Barberio et al., 2015) 10.6 ± 4.5 pg∙mL-1 14.8 ± 5.6 pg∙mL-1

(Gill, Hankey, et al., 0.4 (0.3 to 0.5) pg∙mL-1 14.5 (9.3 to 19.7) pg∙mL-1 2015)

(Gill, Teixeira, et al., 8.2 ± 4.5 pg∙mL-1 27.9 ± 23.4 pg∙mL-1 2015)

(Shing et al., 2014) ~2 ± 0.8 pg∙mL-1 ~3 ± 0.7 pg∙mL-1

Chapter 2: Literature Review 29

(Morrison et al., 2014) ~0.25 ± 0.05 pg∙mL-1 ~1.5 ± 0.5 pg∙mL-1

(Kuennen et al., 2011) 0.64 ± 0.09 pg∙mL-1 4.58 ± 0.58 pg∙mL-1

(Lim et al., 2009) ~0.5 ± 0.2 pg∙mL-1 ~4 ± ~2.5 pg∙mL-1

(Ng et al., 2008) 9.2 ± 4.1 pg∙mL-1 15.2 ± 5.3 pg∙mL-1

(Selkirk et al., 2008) ~2 pg∙mL-1 ~13 pg∙mL-1

(Nieman et al., 2006) 0.94 ± 0.14 pg∙mL-1 45.7 ± 6.1 pg∙mL-1

(Jeukendrup et al., 1.5 ± 0.3 pg∙mL-1 ~40 ± 5 pg∙mL-1 2000)

(Camus et al., 1997; 1.1 ± 0.5 pg∙mL-1 88 ± 13 pg∙mL-1 Pals et al., 1997)

Marker Reference Normative Response Origin Study Pre-exercise values Post-exercise values

(Aggarwal, Gupta, & TNF-α Kim, 2012; Gill, Pro-inflammatory Macrophages Teixeira, et al., 2015)

(Zuhl et al., 2015) ~1.9 ± 0.2 pg∙mL-1 ~2.0 ± 0.2 pg∙mL-1

30 Chapter 2: Literature Review

(Barberio et al., 2015) ~4 ± 2.5 pg∙mL-1 ~4 ± 2 pg∙mL-1

(Gill, Hankey, et al., 2.8 (2.5 to 3.2) pg∙mL-1 3.8 (3.5 to 4.2) pg∙mL-1 2015)

(Gill, Teixeira, et al., 3.1 ± 2.9 pg∙mL-1 6.5 ± 5.0 pg∙mL-1 2015)

(Shing et al., 2014) ~5 ± 0.5 pg∙mL-1 ~6.25 ± 1.25 pg∙mL-1

(Kuennen et al., 2011) ~9 ± 1 pg∙mL-1 ~10 ± 1.5 pg∙mL-1

(Lim et al., 2009) ~0.7 ± 0.2 pg∙mL-1 ~0.85 ± 0.3 pg∙mL-1

(Ng et al., 2008) ~17 ± 8 pg∙mL-1 ~18 ± 9 pg∙mL-1

(Selkirk et al., 2008) ~1.25 ± 0.25 pg∙mL-1 ~1.5 ± 0.25 pg∙mL-1

(Nieman et al., 2006) 1.49 ± 0.37 pg∙mL-1 1.63 ± 0.32 pg∙mL-1

(Jeukendrup et al., 0.84 ± 0.20 pg∙mL-1 ~0.75 ± 0.25 pg∙mL-1 2000)

(Camus et al., 1998) 18 (SEM 1.5) pg∙mL-1 28 (SEM 2.2) pg∙mL-1

IL-1β = interleukin-1β; IL-6 = interleukin-6; TNF-α = tumour necrosis factor alpha.

Chapter 2: Literature Review 31

2.5 PROTECTIVE INTERVENTIONS

As has been previously evidenced, hot environmental temperatures are known to impair aerobic exercise performance (Sawka et al., 2011). While the underlying mechanisms continue to be debated (Cheung, 2007; Nybo et al., 2014), it could be hypothesised that exertional endotoxemia may contribute to central fatigue and inhibit neuromuscular activation of skeletal muscle. Accordingly, practical interventions capable of reducing endotoxemia inflammatory cytokines could maintain optimal neuromuscular activation and attenuate any decline in exercise performance (Lambert,

2004; Vargas & Marino, 2014). The following chapter will explore applied intervention strategies, specifically glutamine supplementation and heat acclimation training, and the potential benefits that may improve athletic performance in hot conditions.

32 Chapter 2: Literature Review

2.5.1 Glutamine Supplementation Due to the similarities between exercise-induced endotoxemia and medical inflammatory sepsis, promising results in clinical patients may also translate into potentially beneficial interventions for athletes in maintaining GI function. While a variety of different interventions have been proposed, such as ascorbic acid or bovine colostrum (Guy & Vincent, 2018; Lambert, 2004), the most promising may be glutamine supplementation, which has a strong body of literature supporting its use in animal and medical subjects (Ashton et al., 2003; Kucuktulu, Guner, Kahraman,

Topbas, & Kucuktulu, 2013; Lambert, 2009; Pugh, Sage, et al., 2017; Singleton &

Wischmeyer, 2006; Wischmeyer et al., 2001; Wischmeyer, Musch, Madonna, Thisted,

& Chang, 1997; Zuhl et al., 2015; Zuhl et al., 2014).

Glutamine, a non-essential amino acid, is one of the primary oxidative fuels for epithelial cells in the gastrointestinal system, and in particular, the small intestine mucosa (Fleming, Zambell, & Fitch, 1997; Windmueller & Spaeth, 1990). Glutamine supplementation research in animal and clinical populations has been conducted for more than 20 years (Wischmeyer, 2006, 2008). Collectively, these studies have found that administration of this amino acid enhances gut barrier integrity, improves immune function and protects intestinal epithelial cells against oxidative stress (Calder &

Yaqoob, 1999; Camilleri, Madsen, Spiller, Van Meerveld, & Verne, 2012; Dugan &

McBurney, 1995; Oriordain, deBeaux, & Fearon, 1996; Singleton & Wischmeyer,

2006; Soares et al., 2014; Wischmeyer et al., 2001; Zhou et al., 2003). This evidence has led to the widely accepted classification of glutamine as a conditionally essential amino acid (Preiser & Wernerman, 2003), as high metabolic stress exceeds the amount of glutamine that can be synthesized (Hall et al., 1996) or consumed in the diet (~3-6

Chapter 2: Literature Review 33

g; based on a daily protein consumption of 0.8-1.6 g∙kg-1 body mass for a 70 kg individual (Gleeson, 2008)).

Mechanistically, the protective effects that glutamine may provide the gut during exercise in the heat appear twofold: 1) a greater stimulation of the heat shock response through activation of HSP; and 2) the strengthening of the intestinal barrier, specifically the tight junctions, due to an overexpression of occludin (Hall et al., 1996;

Rao & Samak, 2012; Wischmeyer, 2008). Following exposure to heat stress, denatured proteins are segregated and refolded by molecular chaperones, HSP (De Maio, 1999;

Moseley, 1997). In addition to maintaining damaged cells, HSP is also considerably elevated following stressful conditions, to protect the organism against subsequent hyperthermic shock (Salgado, White, Schneider, & Mermier, 2014; Skidmore,

Gutierrez, Guerriero, & Kregel, 1995). Therefore, HSP is a crucial overcompensation response to external stress, protecting against further harm (Kuennen et al., 2011;

Moseley, 1997; Sawka et al., 2011).

Glutamine supplementation has been repeatedly proven to protect cells against a variety of stresses, including both heat and oxidation (Akagi, Ohno, Matsubara, Nakai,

& Inouye, 2013; Calder & Yaqoob, 1999; Kucuktulu et al., 2013; Oriordain et al.,

1996; Singleton & Wischmeyer, 2006; Wischmeyer et al., 2001; Wischmeyer et al.,

1997). Despite this, the exact mechanism by which this occurs is still under investigation (Preiser & Wernerman, 2003). A recent review of the current evidence highlighted several key mechanisms by which glutamine may act, including increased expression of HSPs; a similar result to the outcome of heat acclimation (Gibson et al.,

2015; Wischmeyer, 2008). Specifically, glutamine supplementation appears to activate heat shock factor 1 (HSF-1), which results in an enhanced expression of HSP-70, and

34 Chapter 2: Literature Review

has been demonstrated both in vitro and in vivo (animals and humans) (Singleton &

Wischmeyer, 2006; Wischmeyer et al., 1997; Zuhl et al., 2015; Zuhl et al., 2014).

While there is now a significant body of literature linking glutamine to enhanced HSP expression, this has not always been the case. Indeed glutamine’s effect on HSP production was first discovered by accident, where an investigation into various amino acids demonstrated that glutamine greatly increased HSP expression (Sanders & Kon,

1991). Further research in vitro glutamine research was led by Wischmeyer’s group, who demonstrated that glutamine induces HSP-70 activation in intestinal epithelial cells and provides dose-dependent protection against heat and oxidative stress

(Wischmeyer et al., 1997). Likewise, the application of quercetin, a HSP inhibitor, attenuates this improvement, although the non-specific inhibitory qualities of this chemical limit the findings of this research (Wischmeyer et al., 1997). Several other studies have also presented similar evidence which supports a dose-dependent response between glutamine supplementation, enhanced HSP-70 expression and cell survival following heat shock (Chow & Zhang, 1998; Musch, Hayden, Sugi, Straus,

& Chang, 1997).

Building upon this initial research, glutamine was also observed to be a potent enhancer of HSP-70 expression in both unstressed and endotoxemic rats (Wischmeyer et al., 2001). Importantly, glutamine resulted in a marked attenuation of damage and mortality in the endotoxemic animals, suggesting a potent protective effect against sepsis and inflammation (Singleton et al., 2005; Wischmeyer et al., 2001). Glutamine appears to enhance HSP-70 induction through activation of HSF-1, with the associated gene deletion at either of these steps resulting in a loss of any protective effect (Akagi et al., 2013; Morrison et al., 2006; Singleton & Wischmeyer, 2007). Furthermore, both acute and chronic glutamine supplementation has been shown to increase HSP-70

Chapter 2: Literature Review 35

expression and reduce exercise-induced intestinal barrier permeability, possibly through activation of HSF-1, in human participants (Zuhl et al., 2015; Zuhl et al.,

2014).

The process by which glutamine increases HSF-1 appears complex and multifactorial, although one possible mechanism may be through the heosamine biosynthetic pathway

(HBP). This pathway has been found to activate O-linked β-N-acetylglucosamines (O-

GlcNAc), monosaccharides that regulate the stress response (Gong & Jing, 2011;

Hamiel, Pinto, Hau, & Wischmeyer, 2009). Recent research has also suggested another two potential mechanisms by which glutamine may activate HSF-1: an increase in

HSF-1 trimerisation; and a decrease in CCAAT enhancer-binding protein, which normally suppresses HSF-1 (Ropeleski, Riehm, Baer, Musch, & Chang, 2005; Xue,

Slavov, & Wischmeyer, 2012). Oral supplementation with glutamine in exercising rats has been reported to increase HSF-1 activation considerably (Petry, Cruzat, Heck,

Homem de Bittencourt, & Tirapegui, 2015), arguably modulating inflammation, although the link with HSP expression was ambiguous. While the exact mechanistic relationship between these factors remains elusive, cumulative evidence suggests that glutamine supplementations appear to elicit an increase in HSP-70 expression, potentially due to increased transcription of HSF-1 (Wischmeyer, 2008).

Medical studies have demonstrated that glutamine supplementation may provide a protective effect for clinical patients with septic shock or severe burns (Kucuktulu et al., 2013; Wischmeyer, 2008; Wischmeyer et al., 1997). Due to the similar origin of clinical- and exertional-endotoxemia, glutamine supplementation has been proposed as a possible athletic intervention that could minimise endotoxin translocation and the subsequent inflammatory cascade (Gleeson, 2008; Guy & Vincent, 2018; Lambert,

2009; Zhou et al., 2003). Currently, only a few studies have examined the effects of

36 Chapter 2: Literature Review

glutamine supplementation on athletic populations (see Table 2.3). In the earliest of these studies, acute glutamine ingestion was found to reduce intestinal permeability following exercise (Lambert et al., 2001). However, several methodological issues with the study confounded the outcome measures and reduced the relevance of these findings to current glutamine literature. For example, exercise was combined with aspirin, resulting in increased stress on the gastrointestinal system, and glutamine was not tested independently, only combined with carbohydrates.

More recently, Zuhl et al. (2014) demonstrated that 7 days of glutamine supplementation before high-intensity running in the heat resulted in lower intestinal permeability. The authors proposed that glutamine’s protective effects may stem from increased heat shock factor 1 (HSF-1) and HSP-70 activation, leading to a rise in occludin expression at intestinal tight junctions. Although glutamine was reported to reduce GI permeability, plasma endotoxin levels were not measured, making it difficult to confirm the effect of supplementation on endotoxins translocation.

Subsequent research found that an acute 0.9 g.kg-1 fat-free mass dose of glutamine ameliorates GI permeability and reduces post-exercise plasma endotoxin levels compared to a placebo (Zuhl et al., 2015). Similarly, Pugh, Sage, et al. (2017) reported an inverse dose-response relationship between glutamine supplementation and intestinal permeability and damage, although endotoxin levels were not measured.

However, the use of discredited statistical methods (magnitude-based inferences;

(Sainani, 2018)) diminishes the credibility of these results and makes interpretation of the study outcomes somewhat challenging (Pugh, Sage, et al., 2017). Further, as all three studies utilised fixed-pace exercise protocols, the effect of glutamine supplementation on an externally valid exercise task, such as a self-paced time trial, remains unknown (Pugh, Sage, et al., 2017; Zuhl et al., 2015; Zuhl et al., 2014).

Chapter 2: Literature Review 37

Exercise-heat stress has been repeatedly found to increase GI damage and elevate intestinal barrier permeability to endotoxins (Lim & Mackinnon, 2006). The subsequent endotoxic leakage and production of inflammatory cytokines cause further damage to the intestinal epithelium in a vicious loop (Lambert, 2009). A key protein complex in this inflammatory pathway is nuclear factor κB (NF-κB), which is activated by the presence of endotoxins, and results in the promotion of pro- inflammatory cytokines (Liu & Malik, 2006). Interestingly, glutamine supplementation before exercise in the heat has been shown to decrease endotoxin- mediated cytokine production, and this has been linked to an overexpression of HSP-

70 (Zuhl et al., 2015; Zuhl et al., 2014). Strong evidence from multiple studies has indicated that HSP-70 inhibits endotoxin-induced activation of NF-κB, resulting in a suppressed production of inflammatory cytokines (Chen et al., 2006; Dokladny, Lobb,

Wharton, Ma, & Moseley, 2010; Shi et al., 2006). Thus, glutamine may attenuate the release of cytokines following endotoxin exposure due to increased HSP-70 expression.

Glutamine supplementation has also been suggested to exert a protective effect on the intestinal barrier, specifically the tight junctions, through increased expression of occludin (Rao & Samak, 2012). Occludin, a transmembrane protein, plays a critical role in the maintenance of intestinal tight junctions that compose the paracellular barrier, and prevents the translocation of endotoxins (Furuse et al., 1993). Previous research by Dokladny, Ye, Kennedy, Moseley, and Ma (2008) has linked HSF-1 activation to an up-regulation of occludin, as inhibition of this transcription factor resulted in a subsequent decrease in occludin expression. Since glutamine has been shown to be a potent activator of HSF-1, it has been suggested that HSF-1 overexpression of occludin, may explain the reduced internal permeability observed

38 Chapter 2: Literature Review

following supplementation (Rao & Samak, 2012). Recent in vitro research supports this theory, as glutamine supplementation before heat stress increased HSF-1 activation and occludin expression; conversely, glutamine deprivation reduced tight junction protein expression (Li, Lewis, Samuelson, Liboni, & Neu, 2004; Zuhl et al.,

2014).

In conclusion, glutamine supplementation has been found to attenuate intestinal permeability and decrease the translocation of endotoxins into circulation (Dugan &

McBurney, 1995; Wischmeyer, 2008; Wischmeyer et al., 2001; Zuhl et al., 2015).

Mechanistically, glutamine may act through increased activation of HSF-1, leading to elevated HSP-70 and amplified occludin expression (Rao & Samak, 2012;

Wischmeyer, 2008). This results in a host of beneficial and protective effects; improved refolding or removal of damaged cells, suppressed inflammatory cytokine production, and improved tight junction stability (Dokladny et al., 2008; Li et al., 2004;

Morrison et al., 2006; Singleton et al., 2005; Singleton & Wischmeyer, 2007;

Wischmeyer et al., 1997). Although glutamine supplementation has been found to be effective in animal studies and clinical populations, there has only been limited research focused on athletic supplementation (Pugh, Sage, et al., 2017; Wischmeyer,

2008; Zuhl et al., 2015; Zuhl et al., 2014). As the occurrence of exercise-induced endotoxemia closely models that of clinical endotoxemia, glutamine supplementation may prove beneficial in combating increased GI damage, permeability and inflammation arising from exercise in a hot environment.

Chapter 2: Literature Review 39

Table 2.3. Summarised glutamine supplementation research with measured outcomes of LPS or GI permeability and/or damage. Glutamine Glutamine Exercise LPS GI Perm. or Study Sample Fitness level Modality Environment Findings Dose Timing Protocol levels Damage Recreationally 0.25, 0.5 & 60 min @ (Pugh, Sage, trained Running 30 °C; Dose-response effect with 10 0.9 g∙kg-1 2 h prior 70% - ↓ et al., 2017) (52 ± 6 (treadmill) 40-45% RH I-FABP FFM VO mL·kg-1·min-1) 2max Trained 60 min @ (Zuhl et al., 0.9 g∙kg-1 Running 30 °C; GLUT ↓ LPS and TNF-α at 7 (52 ± 7 mL·kg- 2 h prior 70% ↓ ↓ 2015) 1 -1 FFM (treadmill) 12-20% RH 4 h post-exercise ·min ) VO2max

Trained -1 60 min @ (Zuhl et al., - 0.9 g∙kg Running 30 °C; GLUT may induce HSF-1 8 (51 ± 7 mL·kg -1 7 days 65-70% - ↓ 2014) 1 -1 FFM ∙day (treadmill) 12-20% RH and HSP-70 ·min ) VO2max

Trained CHO & 60 min @ (Lambert et 41 Running 22.4 °C; Multi-intervention design 17 (62 ± 2 mL·kg- GLUT mix 70% - No change al., 2001) 1 -1 millimolar (treadmill) 48% RH confounds results. ·min ) every 10 min VO2max

Glutamine Glutamine Intestinal Study Sample Age/Mass Modality Protocol Environment LPS Findings Dose Timing Permeability

0.5 g∙kg- Passive GLUT preserved GI barrier Soares (2014) 78 mice 4 weeks old 7 days 120 min 39 °C ↓ ↓ 1∙day-1 heating & ↓ severity of hyperthermia

Singleton 0.65 g∙kg-1 Passive 42 °C GLUT enhanced GI HSF-1 24 rats >300 g body mass 5 days 30 min ↓ ↓ (2006) every 12 h heating internally and HSP-70

LPS GLUT ↑ HSP-70 expression Wischmeyer 250-350 g body 10-20 min No 24 rats 0.75 g∙kg-1 injection 22 °C ↓ N/A & ↓ organ damage & (2001) mass prior exercise (5 mg∙kg-1) mortality from sepsis

LPS = lipopolysaccharide; Perm. = GI permeability; FFM = fat free mass; RH = relative humidity; I-FABP = intestinal fatty acid binding protein; TNF-α = tumour necrosis factor α; GLUT = glutamine; HSF-1 = Heat shock factor 1; HSP-70 = Heat shock protein 70

40 Chapter 2: Literature Review

2.5.2 Heat Acclimation The human body is remarkably capable of adjusting to stressful environmental conditions. Given time, a healthy individual can adapt to tolerate hot environments and maintain thermoregulatory homeostasis (Taylor, 2014). This is primarily due to numerous functional adaptations, such as increased sweat rate, which reduce physiological strain and the deleterious effects of heat stress on the body (Hori, 1995;

Périard et al., 2015; Robinson, Turrell, Belding, & Horvath, 1943; Sawka et al., 2011).

Repeated exposure to artificially-induced, hot environmental temperatures, a process known as heat acclimation, elicits numerous physiological adaptations that attenuate the negative effects of heat stress (Chalmers, Esterman, Eston, Bowering, & Norton,

2014; Ely, Lovering, Horowitz, & Minson, 2014; Nielsen et al., 1993; Périard et al.,

2015). These biochemical, perceptual and physiological improvements, such as greater skin blood flow, enhanced cardiovascular stability and expansion in blood plasma volume, cumulatively increases exercise capacity (Guy, Deakin, Edwards, Miller, &

Pyne, 2015; Périard et al., 2015; Sawka et al., 2011; Taylor, 2014). Table 2.4 provides a summary of the biological adaptations that result from heat acclimation.

Chapter 2: Literature Review 41

Table 2.4. Adaptations associated with heat acclimation.

Thermal comfort Improved Maximum aerobic power Increased Submaximal aerobic performance Improved

Core temperature Reduced Thirst Improved Rest (temperature) Electrolyte losses Reduced Exercise Total body water Increased Sweating Improved Plasma volume Increased Earlier onset Cardiac output Better sustained Higher rate Heart rate Lowered Skin temperature Reduced Stroke volume Increased Skin blood flow Improved Blood pressure Better defended Earlier onset Myocardial compliance Increased Higher rate (tropic) Myocardial efficiency Increased Muscle glycogen Cardio-protection Improved Lactate threshold Increased Heat shock proteins Increased Muscle and plasma lactate Lowered Acquired thermal tolerance Increased Skeletal muscle force generation Increased Whole-body metabolic rate Lower

Recreated from “Integrated physiological mechanisms of exercise performance, adaptation, and maladaptation to heat stress,” by M. N. Sawka, 2011, Comprehensive Physiology, 1(4), pg. 1883-1928.

42 Chapter 2: Literature Review

2.5.2.1 Physiological adaptations resulting from heat acclimation One of the key thermoregulatory adaptations of heat acclimation is the alteration in sweat characteristics (Garrett, Rehrer, & Patterson, 2011; Périard et al., 2015; Sawka et al., 2011). This is most obviously demonstrated by an enhanced sweat rate, as well as earlier initiation of sweating at a lower core temperature (Nadel, Pandolf, Roberts,

& Stolwijk, 1974; Nielsen, Strange, Christensen, Warberg, & Saltin, 1997). Other adaptions include an increased sweat sensitivity and a higher-rate of output from the sweat glands. These adaptions work to improve evaporative cooling, reduce skin temperature and blood flow, and attenuate heat storage (Périard et al., 2015). Sweat composition is also altered following acclimation, as the release of aldosterone causes a higher concentration of electrolytes to be retained, resulting in more diluted sweat

(Allan & Wilson, 1971; Chinevere, Kenefick, Cheuvront, Lukaski, & Sawka, 2008).

This not only conserves sodium but also enhances heat dissipation, as dilute sweat widens the water vapour gradient between the skin and air, resulting in faster evaporation (Taylor, 2014).

Heat acclimation has been suggested to lower metabolic rate and glycogen utilisation during exercise (King, Costill, Fink, Hargreaves, & Fielding, 1985; Sawka, Pandolf,

Avellini, & Shapiro, 1983). Consequently, lactate accumulation during submaximal exercise is reduced while the lactate threshold and related power output is increased

(Febbraio et al., 1994; Lorenzo & Minson, 2010). This has been speculated to occur due to the larger plasma volume, which increases blood flow through the splanchnic circulation, resulting in enhanced lactate removal (Périard et al., 2015; Rowell,

Brengelmann, Blackmon, Twiss, & Kusumi, 1968). Alternatively, improvements in oxygen delivery due to enhanced cardiac output may alter the metabolic efficiency, and thus delay lactate accumulation during exercise in the heat (Corbett, Neal, Lunt,

Chapter 2: Literature Review 43

& Tipton, 2014). Regardless, the exact mechanism/s behind a heat acclimation alteration in lactate metabolism remains to be completely identified (Minett et al.,

2016; Périard & Racinais, 2015).

Exercise in the heat presents a significant challenge for the cardiovascular system

(Casa, 1999; González-Alonso et al., 2008; Périard, Cramer, et al., 2011a; Rowell,

2011); however, heat acclimation has been found to effectively reduce this strain through a variety of mechanisms (Minett et al., 2016; Nielsen, 1998; Pandolf, 1998;

Périard et al., 2015; Sawka et al., 2011; Sawka et al., 2010). For example, increased plasma volume, redistribution of blood volume and enhanced venous tone from vascular beds, all contribute to improved cardiac stability (Périard et al., 2015; Sawka et al., 2011). Expanded plasma volume, ranging from 4-15%, has been found to occur rapidly following several days of heat exposure (Périard et al., 2015). However, the magnitude of the increase appears highly variable and is affected by skin temperature, training status and hydration (Nielsen et al., 1993; Patterson, Stocks, & Taylor, 2004;

Périard et al., 2015; Sawka et al., 1983; Senay, Mitchell, & Wyndham, 1976).

Nevertheless, plasma volume expansion improves cardiovascular stability by increasing vascular filling, with a resultant decrease in heart rate, and also reduces skin blood flow responses (Périard et al., 2015; Sawka et al., 2011).

2.5.2.2 Time course of heat acclimation Heat acclimation primarily consists of repeated exposure to heat stress through passive or exercise-induced means (Sawka et al., 2011; Taylor, 2014). While passive heat acclimation, administered via climate-controlled chambers, vapour barrier suits or water baths (Fox, Goldsmith, Kidd, & Lewis, 1963; Scoon, Hopkins, Mayhew, &

Cotter, 2007; Zurawlew, Walsh, Fortes, & Potter, 2016) reportedly incur some functional adaptations, the most widely used and effective method for developing heat

44 Chapter 2: Literature Review

acclimation involves aerobic exercise conducted in a hot environment (Brazaitis &

Skurvydas, 2010; Périard et al., 2015). Traditionally, heat acclimation research has divided protocols into three separate categories based on the duration of the exposures:

≤ 7 days (short-term heat acclimation), 8-14 days (medium-term heat acclimation), and

≥ 15 days (long-term heat acclimation) (Garrett et al., 2011). Physiological adaptations from heat acclimation develop relatively quickly with the majority of these enhancements, with approximately 75-80% occurring within the first 4-7 days

(Pandolf, 1998; Shapiro, Moran, & Epstein, 1998). For example, while initial exposure to heat stress results in elevations in heart rate, core and skin temperatures, these decrease during subsequent heat exposures, with the majority of cardiovascular adaptations achieved by 7 days (Périard et al., 2015; Sawka et al., 2010). Most physiological adaptations are considered to be completed after 10-14 days, although heat tolerance may continue to improve throughout further exposures (Périard et al.,

2015; Sawka et al., 2010). The extent of the functional adaptations elicited by heat acclimation are variable and have been shown to depend on exercise intensity as well as the duration and frequency of heat exposures, along with a significant scope of individual variability (Houmard et al., 1990; Périard et al., 2015; Racinais et al., 2014;

Racinais et al., 2012; Sawka et al., 2010).

As the concurrent physiological strain from repeated heat exposures may impair training quality, researchers have also investigated the efficacy of intermittent heat acclimation training. However, the majority of studies have shown that consecutive heat acclimation training is arguably more time-efficient, and effective, at inducing physiological adaptations, when compared to intermittent exposures (Barnett &

Maughan, 1993; Gill & Sleivert, 2001; Périard et al., 2015). Recently, short-term heat acclimation training has become a popular alternative to both intermittent and

Chapter 2: Literature Review 45

traditional heat acclimation training, particularly for high-performance athletes

(Garrett et al., 2011; Garrett et al., 2012; Garrett, Goosens, Rehrer, Patterson, & Cotter,

2009). Although this shorter time course does not induce complete physiological adaptations, it still improves exercise performance while remaining achievable for time-restricted athletes (Périard et al., 2015; Racinais, Alonso, et al., 2015).

Functional adaptations arising from heat acclimation training are relatively brief and decay with cessation of heat exposure training (Garrett et al., 2011). Research into the loss of heat acclimation adaptations is sparse. Most studies report conflicting results due to differences in project design, participant fitness levels and duration of exposures

(Garrett et al., 2011; Périard et al., 2015; Sawka et al., 2011). In general, the fastest adaptations to occur, such as heart rate, are also the fastest to decay, with reports of between a 29-92% loss of heat rate acclimation following approximately three weeks of decay (Pandolf, Burse, & Goldman, 1977; Saat, Sirisinghe, Singh, & Tochihara,

2005; Williams, Wyndham, & Morrison, 1967). In comparison, core temperature appears to have a slower rate of decay, lasting as long as 26 days after a previous heat stress event (Weller, Linnane, Jonkman, & Daanen, 2007). However, a recent review by (Daanen, Racinais, & Périard, 2018) highlighted that for every day without heat exposure, a ~2.5% decay in heart rate and core temperature adaptation occurs.

2.5.2.3 Heat acclimation and exercise-induced endotoxemia Importantly, heat acclimation may prove to be an effective intervention against exercise-induced endotoxemia, through three possible mechanisms: 1) improved cardiovascular stability resulting in preservation of splanchnic blood flow, and hence the intestinal barrier; 2) stimulation of the heat stress response, leading to a protective upregulation of heat shock protein 70 (HSP-70); and 3) enhanced endotoxin tolerance, thanks to increased neutralisation and clearances rates.

46 Chapter 2: Literature Review

As previously discussed, improved cardiovascular stability is one of the key beneficial changes which occurs during heat acclimation, attributed to a combination of factors

(i.e., expansion in plasma volume, stroke volume and cardiac output) (Lorenzo,

Halliwill, Sawka, & Minson, 2010; Nielsen, 1998; Périard et al., 2015). Studies investigating changes in stroke volume and cardiac output following heat acclimation have produced mixed evidence, ranging from a minimal difference (Nielsen et al.,

1997; Wyndham, Rogers, Senay, & Mitchell, 1976) to a significant improvement

(Lorenzo et al., 2010; Nielsen et al., 1993). While the magnitude of each of these adaptations appears to vary depending on acclimation duration, exercise intensity and humidity; in general, heat acclimation has been observed to improve central haemodynamics (Périard et al., 2015).

As impaired blood flow to the gut is a primary cause of intestinal mucosal death (Hall,

Baumgardner, Oberley, & Gisolfi, 1999; Lambert, 2008), maintenance of splanchnic blood flow, due to heat acclimation, may preserve intestinal tract integrity (Lambert,

2004). This hypothesis is supported in animal models, with aerobically trained sheep demonstrating greater maintenance of ileal blood flow during exercise (Sakurada &

Hales, 1998). Similarly, endurance-trained athletes experience lower circulating endotoxins during exercise in the heat, when compared to untrained individuals, potentially attributable to the increased cardiac output associated with aerobic training adaptations (Selkirk et al., 2008).

An organism that is repeatedly exposed to heat stress, such as heat acclimation training, will develop a host of cellular adaptations to better protect against subsequent heat exposure (Sawka et al., 2011). The driving force behind this cytoprotective effect, or acquired cellular thermotolerance, is an enhanced upregulation of heat shock proteins (HSP) as part of the overall heat shock response (Moseley, 1997). HSP play

Chapter 2: Literature Review 47

a pivotal role in protein management, through the refolding of denatured proteins following a heat stress event and preventing a destructive apoptotic cascade (De Maio,

1999). Unsurprisingly, increased expression of HSP is associated with improved cytoprotection (McClung et al., 2008). HSP are segregated into numerical groups based on their molecular mass, and of these, the expression and stress response of the

HSP-70 family are arguably the most widely studied (Ding, Fernandez-Prada,

Bhattacharjee, & Hoover, 2001; Dokladny et al., 2010; McClung et al., 2008; Shi et al., 2006; Singleton et al., 2005; Wischmeyer et al., 1997; Zuhl et al., 2015). Research into both animal and human subjects have noted an increase in HSP-70 following heat acclimation, and indeed, HSP expression is upregulated after a single heat exposure

(De Maio, 1999; McClung et al., 2008; Moseley, 1997).

Intriguingly, HSP-70 upregulation has also been found to strengthen the intestinal barrier against heat exposure and therefore reduce permeability to endotoxins

(Moseley, Gapen, Wallen, Walter, & Peterson, 1994). Work by Dokladny, Moseley, and Ma (2006) showed increased HSP expression resulted in an upregulation of occludin proteins, which play a key role in maintaining the intestinal epithelial tight junctions. Conversely, inhibition of this HSP response markedly reduced tight junction stability and increased gut permeability (Dokladny, Wharton, Lobb, Ma, & Moseley,

2006). HSP-70 expression also appears to provide a protective cross-tolerance against endotoxic shock, as heat-acclimated animals were found to survive following an injection of otherwise lethal endotoxins (Ryan, Flanagan, Moseley, & Gisolfi, 1992).

Further, HSP-70 upregulation is associated with a blunting of endotoxin-induced inflammatory cytokine production, potentially through inhibition of the NF-κB pathway (Ding et al., 2001; Dokladny et al., 2010). In conclusion, heat acclimation stimulates the expression of HSP-70 (Sawka et al., 2011), which has been shown to

48 Chapter 2: Literature Review

attenuate a rise in the permeability of the gastrointestinal barrier to endotoxins and the subsequent release of inflammatory cytokines (Ding et al., 2001; Kuennen et al., 2011;

Moseley, 1997; Moseley et al., 1994; Shi et al., 2006). Thus heat acclimation may prove to be an effective intervention strategy, aimed at reducing exercise-induced endotoxemia (Lambert, 2008).

The third and final mechanism by which heat acclimation may reduce exercise- induced endotoxemia is through enhanced endotoxin clearance (Lim & Mackinnon,

2006). This may be through higher levels of circulating anti-LPS immunoglobulin G

(IgG) and/or anti-LPS immunoglobulin M (IgM) antibodies (Camus et al., 1998; Lim

& Mackinnon, 2006). Over 30 years ago, Brock-Utne et al. (1988) detected a negative correlation (r = 0.38; P < 0.001) between plasma endotoxin levels and circulating anti-

LPS IgG. Similarly, Bosenberg et al. (1988) reported a moderate correlation (r = 0.61;

P < 0.02) between training status and prerace anti-LPS IgG levels, with well-trained athletes exhibiting higher anti-LPS concentrations. It was postulated that the intensive training undertaken by athletes might result in a small number of endotoxins leaking into circulation, causing the production of associated antibodies (Camus et al., 1998;

Lim & Mackinnon, 2006). Similar data have also been reported for racehorses, where pre-race anti-LPS IgG levels were significantly higher than in untrained horses (Baker,

Gaffin, Wells, Wessels, & Brock-Utne, 1988).

In contrast, the self-immunisation hypothesis was not supported by Jeukendrup et al.

(2000), as trained triathletes were found to have lower pre-race anti-LPS IgG levels than healthy, untrained controls. While the authors of this paper suggested that this reduced antibody concentration may have been due to endotoxin leakage from training sessions in the preceding weeks, this is difficult to substantiate. Interestingly, runners who undertook two weeks of training were found to have an increased concentration

Chapter 2: Literature Review 49

of anti-LPS IgM, whereas anti-LPS IgG levels were unchanged (Lim et al., 2009). It has been argued that the increased presence of IgM in circulation following endotoxic stress is evidence that it is the first antibody produced, is more responsive than IgG, and is more active in clearing endotoxins from circulation (Camus et al., 1998; Lim et al., 2009). The efficiency of another clearance mechanism, the reticuloendothelial system (RES) of the liver, may also be enhanced by heat acclimation (Haglind, Wang,

& Klein, 1996; Lim & Mackinnon, 2006). It has been suggested that sub-lethal leakage of endotoxins during training and/or heat stress induce an over-compensation of the

RES, leading to increased scavenging and clearance of endotoxins from circulation

(Haglind et al., 1996). However, due to the absence of research investigating exercise and RES function, the specific effect of heat acclimation training on this mechanism remains to be elucidated.

Although the exact mechanism(s) by which heat acclimation may attenuate exertional- endotoxemia is currently equivocal, previous research has highlighted the potential beneficial effects of this training intervention on exertional-endotoxemia (Lambert,

2004). Only three studies have previously investigated the effect of heat acclimation on GI damage, inflammation and plasma endotoxin levels (Barberio et al., 2015; Guy et al., 2015; Kuennen et al., 2011; Lambert, 2004). Further, all have methodological issues which reduce the relevance of the findings to an applied context (Barberio et al.,

2015; Guy et al., 2015; Kuennen et al., 2011). To the best of our knowledge, the earliest heat acclimation and endotoxemia research, conducted by Kuennen et al. (2011), reported no change in endotoxin levels following 7 days of training in hot, dry conditions (~47 °C; 21% RH). However, no significant elevation in circulating endotoxins, inflammatory cytokines or intestinal permeability was observed following the initial, baseline heat tolerance test. This suggests that the fixed intensity task was

50 Chapter 2: Literature Review

not stressful enough to induce endotoxin translocation, so it is unsurprisingly that no biological adaptation occurred after additional heat acclimation training.

Conversely, a study by Barberio et al. (2015), reported a rise in plasma endotoxins and intestinal damage after running in a hot environment. Yet after 5 days of heat acclimation training, there was no decrease in the endotoxemic response or level of intestinal damage following exercise in the heat, although there was a downward trend.

The authors suggest this may have been due to the short acclimation period (5 days), and that a longer time frame may be necessary for the beneficial adaptations to occur

(Barberio et al., 2015). Another potential factor may have also been the short duration of the exercise sessions, between 23-26 minutes, which may not have been long enough to incur the appropriate physiological heat acclimation enhancements, as previously highlighted (Racinais, Alonso, et al., 2015). Recently, Guy et al. (2016) reported that short-term heat acclimation training did not improve cycling time trial performance over that of a workload-matched control group (~6%), and neither group reported changes in endotoxin or anti-LPS concentrations. These findings suggest that the training stimulus itself, as opposed to the hot environmental conditions, were the driving factor behind the performance improvements, probably due to the low aerobic capacity of the participants (45 ± 5 mL·kg-1·min-1) (Guy et al., 2016).

Although the current literature addressing heat acclimation and its effect on GI damage, inflammation and endotoxemia are inconclusive, greater consideration for the acute training variables included (i.e., exercise intensity and cumulative heat exposure) warrant further attention. Further, the absences of longer duration physical performance trials (>30 min) or neuromuscular assessment in the previous studies require these factors to be addressed.

Chapter 2: Literature Review 51

2.6 SUMMARY AND IMPLICATIONS

Thermally stressful environments impair splanchnic blood flow, damaging the intestinal barrier integrity and increasing permeability to endotoxins (Lambert, 2008).

Impairments in GI barrier functions permits the translocation of endotoxins into the systemic circulation, eliciting an inflammatory cytokine response that may modulate voluntary neuromuscular activation (Lim & Mackinnon, 2006; Vargas & Marino,

2014). Although the logic of these proposed mechanisms is sound, only limited human research has investigated these possible mechanistic links. Nevertheless, prophylactic interventions (i.e., glutamine or heat acclimation) that preserve the GI barrier integrity and maintain central drive could potentially improve exercise performance in the heat

(Lambert, 2004; van Wijck et al., 2012).

This PhD program aims to:

1. Examine the effect of exercise in the heat on GI damage, endotoxin release,

inflammation and neuromuscular function;

2. Assess the protective effects of glutamine supplementation against the

development of endotoxemia and neuromuscular fatigue during exercise in hot

conditions; and

3. Investigate the effect of a short-term heat acclimation protocol on GI

inflammation, endotoxin release and neuromuscular fatigue during exercise in

the heat.

Chapter 2: Summary & Implications 53

Chapter 3: Study 1 – The effects of cycling in the heat on gastrointestinal inflammation and neuromuscular fatigue.

As published in the European Journal of Applied Physiology (accepted 2019). (Osborne, Stewart, Beagley, & Minett, 2019).

Osborne, J.O., Stewart, I.B., Beagley, K.W., & Minett, G. M. (2019). The effect of cycling in the heat on gastrointestinal-induced damage and neuromuscular fatigue. Eur. J. Appl. Physiol, 119(8): 1829-1840. DOI: 10.1007/s00421-019-04172-z.

3.1 ABSTRACT

Aims: This study investigated the effect of exercise in the heat on neuromuscular

function, gastrointestinal damage, endotoxemia and inflammatory cytokines.

Methods: Eight male cyclists completed two 60 min cycling trials in both hot (HOT:

34.5 ± 0.1 °C and 53 ± 1% relative humidity) and temperate environments (CON:

20.2 ± 0.3 °C and 55 ± 3% relative humidity). The cycling task comprised of

alternating 3 min intervals at a moderate-vigorous intensity (50% and 70% of

maximum power output; Pmax) for 30 min, followed by 30 min at moderate intensity

(40-50% Pmax). Neuromuscular function was assessed at pre-, post-exercise and 60

min post-exercise. Circulating levels of endotoxins, inflammatory cytokines and

markers of gut permeability and damage were also collected at these time points.

Heart rate, core temperature, skin temperature, perceived exertion, thermal sensation

and comfort were also measured.

54 Chapter 3: Study 1

Results: Post-exercise voluntary activation of HOT (87.9% [85.2, 90.8]) was

statistically lower (mean difference: -2.5% [-4.5, -0.5], d = 2.50) than that of CON

(90.5% [87.8, 93.2]). The HOT trial resulted in statistically elevated (+69%) markers

of gastrointestinal damage compared to CON (mean difference: 0.424 ng∙mL -1

[0.163, 0.684, d = -3.26]), although this was not observed for endotoxin, other

inflammatory markers, or gastrointestinal permeability.

Conclusion: This research provides evidence that short-duration cycling in the heat

results in sub-optimal neuromuscular activation and increased expression of

gastrointestinal damage markers, without a simultaneous elevation in circulating

endotoxins or pro-inflammatory cytokines.

3.2 INTRODUCTION

Endurance exercise is markedly impaired by heat stress, with a reduction in athletic performance observed in both laboratory and real-world settings (Racinais, Périard, et al., 2015; Tatterson et al., 2000). The aetiology of heat-related fatigue appears to be multi-factorial and dependent on a complex interplay of different factors including exercise intensity, hydration, health and fitness (Nybo et al., 2014). Exacerbated cardiovascular strain during exercise has traditionally been used to explain declines in exercise performance under heat stress (Nybo et al. 2014). However, this suggestion contrasts the reduced physical capacity observed with a high core temperature (Tc), irrespective of cardiac supply to exercising muscle (Cheung & Sleivert, 2004b; Nielsen et al., 1993). While there is an apparent inverse relationship between a rising Tc and attenuated CNS drive to the motor neurone pool (Cheung, 2007), the proposed critically limiting Tc appears overly simplistic given the evidenced tolerance of >40.0

Chapter 3: Study 1 55

°C Tc without ill consequence (Byrne et al. 2006; Ely et al. 2009; Racinais et al. 2019).

Accordingly, alternative mechanistic processes explaining evoked central fatigue experienced with exercise- and environment-induced hyperthermia remain to be identified (Nybo & González-Alonso, 2015; Nybo et al., 2014).

Exercise in the heat results in a redistribution of blood flow from splanchnic regions, leading to death of mucosal epithelial cells that line the intestinal tract (Lambert,

2008). Heat also directly effects the integrity of the gastrointestinal barrier and results in the opening of tight junctions (Lambert, 2004; Marshall, 1998; Moseley et al.,

1994). As a consequence of this increased permeability, endotoxins translocate through the weakened intestinal barrier, causing the clinical condition, endotoxemia

(Lambert, 2008). Endotoxemia elicits a strong immune response and the resultant release of inflammatory signalling cytokines, including tumour necrosis factor alpha

(TNF-α) (Lambert, 2009). Increased circulating levels of these cytokines have been implicated in signalling the brain to produce symptoms of sickness, such as nausea or malaise (Dantzer, 2004). Thus, Vargas and Marino (2014) argue that the production of similar cytokines during exercise may act as a neuro-modulator and result in a manifestation of transient perceptual fatigue, similar to the response that occurs during a disease state.

The exact relationship between this string of factors (i.e., exercise and heat stress, intestinal damage, endotoxin translocation, cytokine release, and the development of fatigue) is currently unknown. However, the neuroinflammatory model of fatigue may explain the decreased exercise performance when competing in a hot environment

(Lambert, 2008; Nybo et al., 2014; Vargas & Marino, 2014). While elevated

56 Chapter 3: Study 1

concentrations of endotoxins and inflammatory cytokines have been evidenced after prolonged exercise in hot conditions (Bosenberg et al., 1988; Camus et al., 1997; Gill,

Teixeira, et al., 2015), the main interest in exertional-endotoxemia and the resultant inflammation has concerned the aetiology of heat illness and injury (Lim &

Mackinnon, 2006). Instead, we propose that transient endotoxemia and the release of inflammatory cytokines during prolonged exercise in the heat could downregulate neural drive by directly, or indirectly, altering the electrophysiology of central neurons as has been previously demonstrated (Vezzani & Viviani, 2015; Vitkovic, Bockaert,

& Jacque, 2000). Vargas and Marino (2017a) recently showed neuroinflammation to alter cortical activity and contribute to fatigue during a 60 min cycle protocol in 35 ºC conditions. Unfortunately, the proposed preceding changes in gut permeability and endotoxin concentrations were not reported.

The present study aimed to examine the effect of exercise in the heat on neuromuscular function and exercise-induced endotoxemia. It was hypothesised that cycling in a hot environment would induce central fatigue, observable through reduced post-exercise voluntary activation. Further, exercise in the heat would increase the level of circulating endotoxins, inflammatory cytokines and markers of gastrointestinal permeability and damage, when compared to the control condition.

3.3 METHODS

Participants

Eight trained male cyclists (age: 28 ± 5 years; height 182 ± 7 cm; body mass: 76 ± 10

-1 - kg; relative VO2max: 58 ± 7 mL∙kg ∙min 1; Pmax: 394 ± 35 W), volunteered to participate in this study. All participants cycled at least twice per week and were

Chapter 3: Study 1 57

classified as trained or well-trained athletes (performance level 3 or 4) (De Pauw et al.,

2013). Participants were non-smokers, free of any injury or illnesses and reported no history of gastrointestinal issues or diseases. All participants were informed of all study requirements before they provided verbal and written consent. University

Human Research Ethics Committee approval of this project was attained before the commencement of any testing.

Participants completed an initial familiarisation trial followed by two experimental trials. During the initial familiarisation trial, participants were extensively familiarised with all neuromuscular assessments and procedures that would be undertaken during subsequent experimental trials. Participants also completed a maximal aerobic capacity (VO2max) test via expired gas analysis (TrueOne 2400; ParvoMedics, Salt

Lake City, USA) on a cycle ergometer (Excalibur Sport; Lode, Groningen,

Netherlands) using a 25 W.min-1 step protocol. The power of the highest, completed stage was taken as the participant’s maximal power output (Pmax).

Exercise Protocol

Experimental trials consisted of a 60 min cycling task on a cycle ergometer (Excalibur

Sport, Lode, Netherlands), comprised of 10 x 3 min intervals, alternating between 50% and 70% Pmax, followed by 30 min at a self-selected power between 40-50% Pmax. This

-1 exercise protocol was utilised as it rapidly elevated Tc (∆ 0.039 °C∙min ) when compared to a traditionally fixed load at 50% for 60 min (∆ 0.020 °C∙min-1; unpublished observations). Similar duration cycling protocols have previously reported elevated markers of gut damage and permeability (van Wijck et al., 2011) and

58 Chapter 3: Study 1

increased levels of inflammatory cytokines (Gray et al., 2009; Vargas & Marino,

2017a).

While the exercise protocol was identical, environmental conditions differed between the two trials, with one trial undertaken in the heat (HOT: 34.5 ± 0.1 °C and 53 ± 1% relative humidity) and the other in a temperate environment (CON: 20.2 ± 0.3 °C and

55 ± 3% relative humidity). The trials were completed in a counter-balanced and randomised order, matched for time of day (± 2 h), and separated by ≥ 7 days.

Participants abstained from caffeine and alcohol for 12 h, and strenuous exercise for

48 h, before each experimental trial. Physical activity, food and fluid intake were diarised for the 24 h before the first experimental trial and replicated for the subsequent trial.

Upon arrival for each experimental trial, participants provided urine and blood samples before undertaking a pre-exercise neuromuscular assessment. Participants were asked to nude weigh with a voided bladder, before being instrumented (i.e., rectal and skin thermistors, heart rate monitor) and dressing in cycling apparel (i.e., bib, socks and cycling shoes). Participants were seated for 5 min in a climate-controlled laboratory to provide baseline data before entering the climate chamber and undertaking the cycling task. Upon completion of the cycling task, participants immediately completed a neuromuscular assessment and provided a blood sample. A towel-dried post-exercise nude weigh was completed to allow for calculation of fluid loss. Participants then consumed 250 mL of water and rested for 60 min in a temperature laboratory environment before a final neuromuscular assessment and blood sample collection.

Physiological Measures

Chapter 3: Study 1 59

Heart rate (HR) values were recorded using a chest strap (Polar Electro Oy, Kempele

Finland) and software (Polar Team2, Kempele Finland). Core temperature was measured via a flexible rectal thermistor (449H; Henleys Medical, Hertfordshire, UK) inserted ~12 cm past the anal sphincter. Skin temperature was measured from conductive skin thermistors (EU-UU-VL5–0; Grant Instruments, Cambridge, UK) affixed to the skin (Leuko Sportstape Premium; Beiersdorf, Hamburg, Germany) at four separate sites: right shin, right scapula, posterior neck, and posterior left hand.

Mean skin temperature (Tsk) was calculated according to the four-site formula published by the International Organization for Standardization (ISO 9886, 2004). Tc and Tsk thermistors were connected to a data logger (Squirrel SQ2020; Grant

Instruments, Cambridge, UK) and computer, which recorded every 5 min while cycling. Hydration status was assessed upon arrival using a mid-stream urine sample to measure urine specific gravity (PAL-10S; Atagi Ci. Ltd, Tokyo, Japan) and fluid loss was calculated via pre- to post-exercise nude body mass changes using a set of calibrated scales (WB-110AZ; Tanita Corp., Tokyo, Japan).

Perceptual measures

Borg’s Rating of Perceived Exertion (RPE) scale (Borg, 1970) was used to measure perceived exercise intensity. Thermal sensation was recorded using a 16 point scale (0

= ‘unbearably cold’ to 8 = ‘unbearably hot’) (Young, Sawka, Epstein, Decristofano,

& Pandolf, 1987), and thermal comfort was measured using a 4 point scale (1 =

‘comfortable’ to 4 = ‘very uncomfortable’) (Gagge, Stolwijk, & Hardy, 1967). These perceptual measures were recorded at baseline and then every 5 min during the cycling task.

60 Chapter 3: Study 1

Neuromuscular function and voluntary activation

The neuromuscular function of the right knee extensors were assessed pre-, post- and

1 h post-exercise using a Biodex Systems 3 Dynamometer (Biodex Medical Systems,

Shirley, New York, USA). Participants were secured in an upright position via straps across the chest, waist and right thigh. The lever fulcrum was aligned with the right lateral epicondyle and the right lower leg attached to the lever arm via a strap positioned 10 mm proximal to the lateral malleolus.

Assessment of voluntary activation (VA) was achieved via stimulation of the right femoral nerve using reusable self-adhesive gel electrodes (Pals; Axelgaard

Manufacturing Co. Ltd., Fallbrook, CA). An anode electrode (3.2 cm diameter) was placed 3 cm distal to the inguinal ligament bordering the femoral triangle, and a cathode electrode (5 x 9 cm) was positioned on the medio-posterior aspect of the right upper thigh, on the border of the gluteal fold. The current applied to the nerve was driven by a Digitimer DS7AH stimulator (Digitimer Ltd., Welwyn Garden City,

Hertfordshire, England) using a single square-wave pulse with a 500 μs width. A resting twitch ramp was undertaken at the start of each experimental trial to determine the required current for maximal stimulation. The final current was then increased by an additional 20% to ensure supramaximal stimulation of the femoral nerve (Saboisky et al., 2003).

Participants performed a standardised warm-up and rested for two minutes before completing a set of 5 x 5 s maximum voluntary isometric knee extensor contractions

(MVC) at 90° knee flexion, with a 30 s rest between each repetition. During each

MVC, participants received strong verbal encouragement and exhortation as well as

Chapter 3: Study 1 61

visual feedback of torque production. Upon observing a plateau in voluntary torque during each MVC, the primary investigator manually triggered a single femoral nerve stimulation. A further stimulation was also triggered following the completion of each

MVC, providing evoked twitch torque properties (Shield & Zhou, 2004). This neuromuscular assessment was repeated immediately post- and 1 h post-exercise.

VA was calculated using the twitch interpolation technique as previously described by

Allen, Gandevia, and McKenzie (1995). Before calculating VA, each MVC torque trace was visually assessed to ensure it demonstrated a clear plateau in voluntary contraction torque. MVCs that failed this criterion were rejected and removed from the mean VA calculations. VA was assessed using the formula: VA (%) = (1 – interpolated twitch torque/resting control twitch torque) * 100. Peak voluntary isometric torque was taken as the mean torque value of the 25 ms preceding delivery of the electric stimulus. The peak torque value recorded in the 100 ms period following the stimulus was considered the superimposed torque value. These neuromuscular assessments of MVC torque and VA are reliable, with calculated ICCs of 0.94 and

0.94, respectively, in our laboratory.

Evoked twitch contractile properties

Evoked twitch properties were assessed from the resting twitch delivered following the completion of each MVC, as described previously (Cannon, Kay, Tarpenning, &

Marino, 2008; Shield & Zhou, 2004). Twitch data were averaged across the five repetitions for each time point and analysed for peak twitch torque (Pt; defined as the peak torque during the evoked twitch); time to peak torque (TPt; time from the first rise in torque above baseline to peak torque); half relaxation time (½ RT; time taken

62 Chapter 3: Study 1

for torque to reduce by half of the peak torque value); contraction duration (CD; time to peak torque plus half relaxation time); rate of torque development (RTD; slope of twitch-torque curve from onset to peak torque); rate of relaxation (RR; slope of twitch- torque curve peak torque to half relaxation time).

Surface electromyography (EMG)

Surface knee extensor EMG data were recorded from the vastus medialis (VM) and vastus lateralis (VL) of the right leg during MVC assessments. The 30 x 22 mm surface electrodes (Ambu Blue Sensor N-00-S; Ambu A/S, Ballerup, Denmark) were positioned over the visual mid-point of the respective muscle bellies, and an additional earth electrode was attached to the lateral femoral epicondyle. The electrodes were positioned with an inter-electrode distance of 20 mm and orientated parallel with the muscle fibres. All sites were shaved, abraded and cleaned before electrode placement.

Electrodes remained attached to the participant during the length of the testing trial to ensure consistency in placement. Raw EMG data was sampled with dynamometer data at 1 kHz through a 16-bit PowerLab 26T AD unit (AD Instruments, Sydney, Australia)

(amplification=1000; common mode rejection ratio=110 dB), bandpass filtered between 20-500 Hz and stored for later analysis. EMG data were quantified via the root-mean-square method with a triangular Bartlett sliding window of 100 ms (100 data points) using LabChart 8.0 software (AD Instruments, New South Wales,

Australia). The mean EMG signal amplitude over the 500 ms period before stimulation was calculated and averaged across all repetitions for each time point (Cannon, Kay,

Tarpenning, & Marino, 2007). Mean post- and 1 h post-exercise EMG amplitudes were then normalised to mean values obtained during pre-exercise MVC (Burden & Bartlett,

Chapter 3: Study 1 63

1999). The 60 ms before each peak root-mean-square EMG value was excluded to remove the confounding effect of the stimulation artefact.

Biochemical markers and analysis

Pre-, post- and 1 h post-exercise samples were drawn from an antecubital venipuncture using a butterfly needle (21G, BD, North Ryde, Australia) and serum

Vacutainer tubes (BD, North Ryde, Australia). After clotting at room temperature, samples were centrifuged at 3500 revolutions∙min-1 for 10 min at 4 ˚C, pipetted into pyrogen-free microtubes, and frozen at -20 ˚C for a maximum of 8 months.

Participants were seated upright in a phlebotomy chair for all venepunctures. Serum concentrations of circulating cytokines (i.e., TNF-α and interleukin 1 beta; IL-1β) were determined using quantitative sandwich enzyme-linked immunosorbent assays

(ELISA) as per the manufacturer instructions (elisakit.com, Melbourne, Australia).

Intra-assay precision was calculated for IL-1β (CV = 13.2%) and TNF-α (CV=

5.6%). Absorbance was read at 450 nm with a 570 nm wavelength subtraction and corrected for blank control wells (SpectroStar Nano; BMG Labtech, Germany).

Intestinal fatty acid binding protein (I-FABP) serum levels were quantified using a commercially available ELISA kit (CV= 7.0%; Human FABP2, RayBiotech,

Norcross, GA, USA) and read at 450 nm with correction for blank control wells.

Serum claudin 3 (CLDN-3) concentration was assessed with a commercially available ELISA kit (CV= 17.1%; Human Claudin-3, Cusabio, Wuhan, China).

Absorbance was read at 450 nm with a 570 nm wavelength subtraction.

Serum endotoxin levels were detected using a commercially-available, quantitative kinetic chromogenic Limulus amebocyte lysate (LAL) assay kit (CV= 12.7%; Lonza,

64 Chapter 3: Study 1

Walkersville, MD, USA). Samples were analysed according to the manufacturer’s instructions. Briefly, the E. coli 055:B5 endotoxin standard (Lot Number:

0000547137) was reconstituted with reagent water to the volume stated on the supplied

Certificate of Analysis and vigorously vortexed for 15 minutes to yield a 50 EU∙mL -1 stock solution. Tenfold serial dilutions of this standard (5, 0.5, 0.05, 0.005 EU∙mL -1) were prepared in duplicate using the provided sterile water (Lonza, Walkersville, MD,

USA). Each solution was vortexed for at least 1 min between dilutions and sterile tips

(Biopur epTIPS; Eppendorf AG, Germany) were used for all pipetting to reduce carry- through contamination of the standards. Samples were thawed, diluted at 1:5 with sterile water (Lonza, Walkersville, MD, USA) in endotoxin-free microtubes

(GoldenGate Bioscience, Claremont, CA, USA) and heated in a heat block for 15 min at 75 °C to inactivate endotoxin-neutralising agents. 100 μL of each sample, standard and negative controls were then pipetted into 96-well plate and incubated in a microplate reader (SpectroStar Nano; BMG Labtech, Germany) for 10 min at 37 °C before the 100 μL of LAL reagent was added to each well. The spectrophotometer immediately read the optical density of the plate at 405 nm and then every 61 s for 118 cycles. The reaction time in seconds for each well to increase absorbance 0.2 units above baseline was determined. A log/log (reaction time/concentration) linear correlation of each standard was calculated (r = -0.991). The slope and y-intercept from this formula were then used to calculate the unknown endotoxin concentration of the samples, corrected for the dilution factor.

Statistical analysis

Data were analysed using linear mixed modelling in a Bayesian framework

(Mengersen, Drovandi, Robert, Pyne, & Gore, 2016). Exploratory data plots were

Chapter 3: Study 1 65

inspected for normality before Markov chain Monte Carlo (MCMC) procedures were used to generate posterior predicted values, using the ‘rjags’ and ‘R2jags’ packages in the statistical software, R. Specifically, 50k iterations with an initial 1k burn-in were thinned by a factor of 10 to generate 5k posterior values. Models utilised vague prior distributions (mean 0, precision 0.001) with a random intercept for each participant.

Fixed model parameters included time, condition, and time x condition and the final model for each analysis was chosen following comparison of extracted deviance information criterion values.

Mean and 95% credible interval [CI] values were calculated for each posterior parameter, with a comparison between conditions or time points calculated as the mean difference (MD) and associated 95% CI. Evidence of statistical effect or difference was accepted when a 95% CI did not include 0. Cohen’s d effect sizes were used to evaluate the magnitude of all statistically different comparisons and were categorised as small (0.2), moderate (0.5) and large (0.8) (Cohen, 1988).

3.4 RESULTS

Exercise protocol

Workload changes (i.e., a reduction from 50% Pmax constant exercise load to 40% Pmax between the 30-45 min) were replicated each participant during the subsequent experimental trial. No difference in the completed workload was observed between conditions.

Neuromuscular function

No statistical difference in pre-exercise MVC torque (MD: 3 N∙m [-11, 18]) or VA

(MD: 0.3% [-1.3, 1.9]) was observed between conditions (Table 3.1). Both conditions

66 Chapter 3: Study 1

reported statistically lower post-exercise MVC torque than respective pre-exercise values (CON MD: -24 N∙m [-38, -8], d = 3.09; HOT MD: -27 N∙m [-43, -11], d = 1.44;

Figure 3.1), although no difference in MVC torque was observed between conditions at this time point (MD: 7 N∙m [-9, 23]). Post-exercise VA followed a similar pattern, with reductions compared to pre-exercise values (CON MD: -1.9% [-3.6, -0.1], d =

2.08; HOT MD: -4.8% [-6.7, -2.8], d = 4.82) observed for both conditions. Further, post-exercise HOT VA values were found to be statistically lower than CON (MD: -

2.5% [-4.5, -0.5]; d = 2.50). There was some evidence of a statistical condition difference in MVC torque 1 h post-exercise, with HOT impairing MVC torque values compared to CON (MD: -19 N∙m [-35, -3]; d = 2.42), and both conditions were decreased compared to pre-exercise values (MD: -35 to -19 N∙m [-51, -4], d = 2.46 –

4.47). There was some evidence that VA remained impaired in HOT compared to CON

1 h post-exercise (MD: -1.7% [-3.5, 0.1], d = 1.92), and was also lower than within- condition pre-exercise values (MD: -3.5% [-5.3, -1.6], d = 3.75).

Evoked twitch contractile properties and EMG

Cycling for 60 min, regardless of environmental condition, resulted in a statistical reduction in Pt torque compared to pre-exercise values (MD: -12 to -12 N∙m [-19, -6], d = 3.42 – 3.80; Table 3.1), although this difference disappeared 1 h post-exercise for both conditions (MD: -5 to -2 N∙m [-11, 4]). There was no evidence of a statistical condition difference at any time point for Pt. Cycling in HOT resulted in statistically lower post-exercise ½ RT and CD values compared to baseline (d = 2.01 – 2.75), although no other time, condition or interaction effects were observed (Table 3.1).

There was no evidence of any time, condition or time x condition interaction factor effects for TPt, RTD or RR (Table 3.1).

Chapter 3: Study 1 67

Relative %EMG output for post-exercise VL and VM was statistically reduced compared to baseline values, regardless of condition (d = 1.92 – 4.55, Table 1).

Further, VM EMG remained impaired 1 h post-exercise for both conditions (MD: -13 to -12 [-24, 0], d = 2.06 – 2.31). While no evidence of a condition difference was observed for VL at any time point, post-exercise VM EMG values for HOT were found to be statistically lower than CON (MD: -11% [-23, 0]; d = 1.95).

68 Chapter 3: Study 1

Figure 3.1. Posterior predicted mean [95% CI] for MVC torque (N∙m), VA (%) and Pt torque (N∙m) at pre, post and 1 h post-exercise in HOT and CON. * indicates a statistical time difference compared to pre-exercise values within a condition; † indicates a statistical condition difference at a respective time point.

Chapter 3: Study 1 69

Biochemical markers and analysis

A rise in markers of gastrointestinal damage markers (I-FABP) was observed in HOT, with statistically higher (+140% mean) post-exercise values compared to baseline

(MD: 0.608 ng∙mL-1 [0.345, 0.861], d = -4.67; Table 3.2). Evidence of a condition difference at this time point was also observed, with post-exercise HOT reporting statistically higher (+69% mean) levels of I-FABP than CON (MD: 0.424 ng∙mL-1

[0.163, 0.684]; d = 3.26), although this difference disappeared by 1 h post-exercise

(Figure 3.2). In contrast, the exercise in CON was not found to alter circulating I-FABP levels above pre-exercise levels (Table 3.2). No parameter (i.e., time, condition or interactions) was statistically different for either inflammatory cytokine (IL-1β and

TNF-α), intestinal permeability (CLDN-3) or endotoxin.

70 Chapter 3: Study 1

Figure 3.2. Posterior predicted mean [95% CI] serum concentrations of inflammatory cytokines (IL-1β and TNF-α), markers of intestinal damage (I- FABP) and permeability (CLDN-3) at pre, post and 1 h post-exercise in HOT and CON. * indicates a statistical time difference compared to pre-exercise values within a condition; † indicates a statistical condition difference at a respective time point

Physiological and perceptual measures

Evidence of time, condition and time x condition interaction effects were observed for all physiological (HR, Tc and Tsk) and perceptual (RPE, thermal sensation and thermal comfort) measures (Table 3.3). While both HR and Tc were similar at baseline, a statistical difference between conditions was observed from 5 min and 10 min, respectively, and this condition difference persisted for the remainder of the exercise task. Statistically higher Tsk, RPE, thermal sensation and thermal comfort values were

Chapter 3: Study 1 71

observed for every time point in HOT, when compared to CON. Analysis of pre- exercise USG (βCONDITION: 0.005 [-0.032, 0.040]) and body mass (βCONDITION: 0.30 [-

0.27, 0.88]) revealed no statistical differences between conditions at baseline. Body mass loss was statistically greater following HOT, compared to CON (MD: -0.66 kg

[-0.86, 0.44], d = -6.28).

3.5 DISCUSSION

This study investigated the effect of exercise in the heat on neuromuscular function, exertional-endotoxemia and inflammation. It was hypothesised that the development of mild endotoxemia and the subsequent transient release of inflammatory cytokines could potentially downregulate neural drive, and therefore voluntary torque output, during exercise in the heat (Gill, Hankey, et al., 2015; Lambert, 2008; Vargas &

Marino, 2014). However, 60 min of cycling was found to attenuate MVC torque, regardless of the environmental condition, equally. While central fatigue was observed following exercise in both conditions, HOT resulted in a greater impairment in post- exercise VA, compared to CON (d = 2.50; Figure 1). Peripheral fatigue was similar between conditions, despite a faster half-relaxation time and shorter contraction duration in the heated muscles compared to baseline. Intestinal damage was observed to be statistically higher following exercise in HOT than CON (d = -3.26). In contrast to the study hypothesis and the reported intestinal damage, there was no evidence of a statistical difference for either inflammatory cytokine (Table 3.2). Nevertheless, these findings align with the observed absence of endotoxin translocation, which was postulated as the initial driving mediator of the pro-inflammatory cytokine response.

In summary, the findings from this study suggest that reductions in MVC torque following cycling in the heat appear to principally stem from diminished CNS drive to

72 Chapter 3: Study 1

the musculature with partial contribution from peripheral fatigue; i.e., distal to the neuromuscular junction.

Périard and colleagues (2011b) reported similar reductions in voluntary force production to that seen here (Figure 1) following a 40 km cycling time trial in hot and cool temperatures. The authors noted that exercise in the heat resulted in only a small decrease in VA, but considerable peripheral fatigue, highlighting the failure of the muscular contractile properties, in particular, a shorter half-relaxation time. The present study observed a similar finding for peripheral fatigue, with evidence of a statistical reduction in post-exercise peak twitch torque for both conditions, despite alterations in HOT muscle properties (i.e., faster ½RT and CD; Table 3.1). One possible explanation may be that temperature-induced changes in muscle contractile properties could be briefly surmounted by a transient increase in motoneurone discharge rates to preserve fusion (Todd et al., 2005). We speculate that a brief upregulation in motor unit firing during the short-duration of an MVC (5 s) in the current study could potentially explain the similar level of voluntary torque production between conditions.

The termination of fixed-pace exercise at a specific critical Tc or the anticipatory reduction in power output to avoid reaching this critical threshold has been suggested in multiple studies (Nielsen et al., 1993; Tucker, Rauch, Harley, & Noakes, 2004b).

However, Nybo and González-Alonso (2015) recently argued that a causative association between elevated Tc and fatigue may be overly simplistic, as studies have reported some athletes continuing to exercise without adverse performance or health complications, despite a Tc that exceeds the ‘cut-off’ of 40°C (Byrne et al., 2006; Ely

Chapter 3: Study 1 73

et al., 2009). While it should be recognised that elevated Tc may be a driving factor behind exercise cessation and the development of fatigue in certain situations, alternate mechanisms and process that may induce fatigue during exercise in the heat should also be considered. Therefore, by building upon work by Lambert (2008) and Marshall

(1998), we hypothesised that the development of exertional endotoxemia and subsequent release of pro-inflammatory cytokines could potentially down-regulate neural drive to the skeletal muscle.

Prolonged exercise in the heat has been shown to place considerable strain on the circulatory system and sufficiently impair blood flow to the gut to cause hypoxia of the epithelial tissue (Lambert, 2009). The subsequent cell death compromises the integrity of the intestinal wall and permits translocation of endotoxins into the circulation, mediating an immune response and the production of inflammatory cytokines (Jeukendrup et al., 2000; Pals et al., 1997). Inflammatory cytokines such as

IL-1β and TNF- α, have been linked with the manifestation of transient fatigue and effort-related motivation (Vargas & Marino, 2014). The exact mechanism(s) by which inflammatory cytokines may produce these symptoms is currently equivocal.

However, it has been theorised that peripherally-produced cytokines could influence afferent feedback via neuronal modulation of the vagus nerve or by signalling communication cells located in the circumventricular organs of the brain (Dantzer,

2004; Vezzani & Viviani, 2015). The blood-brain barrier normally blocks the translocation of peripherally-produced inflammatory cytokines. However, saturable transport systems (Dantzer, 2004) or a rise in hyperthermia-induced barrier permeability (Sharma & Hoopes, 2003) may permit cytokines to cross into the brain and directly regulate neuronal output (Vezzani & Viviani, 2015; Vitkovic et al., 2000).

74 Chapter 3: Study 1

While the role of pro-inflammatory cytokines on brain neural output is complex and multifaceted, elevated levels of these molecules following exercise, or in patients with neuro-immune diseases (Morris, Berk, Galecki, Walder, & Maes, 2016), have been evidenced to underpin mental fatigue and the sensations of illness (Ament & Verkerke,

2009; Robson-Ansley, Milander, Collins, & Noakes, 2004).

Although intestinal blood flow was not directly measured, the cycling protocol resulted in considerable cardiovascular and thermal strain (Table 3.3). In contrast to previous research, intestinal permeability was unchanged between conditions and time points

(van Wijck et al., 2011). One possible explanation for this unexpected result may be the use of serum biomarker CLDN-3 to assess intestinal permeability, as opposed to the more commonly utilised technique of ingestible sugar probes (Marchbank et al.,

2011; Pals et al., 1997) that requires prolonged post-trial urine collection (i.e. 4-5 h post-exercise). While claudin-3 expression has been proposed as a non-invasive method of assessing intestinal barrier dysfunction (Grootjans et al., 2010; Thuijls et al., 2010), only one previous exercise science study has utilised this marker (Yeh et al., 2013), with the authors reporting considerably higher expression of CLDN-3 (i.e.,

6500-8500 pg∙mL-1) than observed in the present study (19-29 pg∙mL-1). The exact cause of this difference is difficult to ascertain, particularly as clinical research utilising cardiac surgery patients has previously reported urinary CLDN-3 levels in the range of 20-170 pg∙mL-1 (Habes et al., 2017).

Exercise in HOT was found to result in increased (+140% mean) serum levels of I-

FABP, a marker of gastrointestinal damage, indicating that the intestinal barrier was compromised. This observation was similar to the gastrointestinal damage recently

Chapter 3: Study 1 75

reported (March et al., 2017; Pugh, Impey, et al., 2017), despite differing environmental conditions and exercise protocols (i.e., running) compared to the present study. While running has been linked with increased mechanical tearing of tight junctions within the intestinal tract (Lim & Mackinnon, 2006; van

Nieuwenhoven, Brouns, & Brummer, 2004), the observed increase in I-FABP (Figure

2) during the present study suggests that sufficient thermal and physiological strain was experienced to induce intestinal damage in-line with previous literature (March et al., 2017; Pugh, Impey, et al., 2017), despite the use of cycling.

Despite this rise in intestinal damage following exercise in the heat, endotoxin concentrations were not observed to increase for either condition statistically, or reach a level classified as mild endotoxemia (>5 pg∙mL-1, corresponding to approximately

0.05-0.10 EU∙mL-1). Similarly, no evidence supported a statistical effect of time, condition or interaction for any of the inflammatory cytokines. This may have been due to the comparatively short exercise duration (60 min) or modality (cycling) which are in contrast to the ultra-endurance events (Gill, Hankey, et al., 2015; Ng et al., 2008) or treadmill protocols (Shing et al., 2014; Yeh et al., 2013) utilised in previous exertional-endotoxemia research. Alternatively, the variable half-life, detection and expression of endotoxins and cytokines in the blood may not have aligned with the venepuncture collection time points. Further, Gnauck, Lentle and Kruger (2016) identified multiple confounding issues which can influence the quantification of endotoxin, including the vacutainer material, false positives from β-glucan contamination, and neutralisation by inhibitory components of a blood sample.

Dilution (1:5) and heating of samples (75 °C for 15 min) was utilised in the current study to reduce the sequestration of endotoxins by binding proteins in serum. However,

76 Chapter 3: Study 1

a spike recovery of only 23-26% indicated continued inhibition of the samples. Thus, we suggest that the low level of endotoxins found here (undetectable to 0.011 EU∙mL-

1) may not accurately reflect the true endotoxin response during cycling in the heat.

3.6 CONCLUSION

The present study proposed that exercise in the heat, and the resultant translocation of endotoxins into systemic circulation, could potentially alter motivation and drive the development of central fatigue via the release of pro-inflammatory cytokines. In alignment with the study hypothesis, strenuous exercise in hot environmental conditions resulted in increased intestinal damage. However, the low endotoxin concentration and attenuated inflammatory cytokine response in HOT was an unexpected finding and in contrast to previous research (Jeukendrup et al., 2000).

Exercise in the heat was also observed to reduce knee extensor torque and VA, indicating the development of central fatigue, when compared to a temperate condition, which aligns with the findings of Nybo and Nielsen (2001a). Interestingly, this attenuation in central drive occurred during brief MVCs, as opposed to the sustained contractions utilised in other studies (Périard, Cramer, et al., 2011b; Todd et al., 2005). How impairments in maximal isometric contraction force translate to submaximal, dynamic exercise performance remains to be elucidated. Future investigations should reassess the possibility of an association between transient exertional-endotoxemia and hyperthermia-induced fatigue, with a view to implementing prophylactic strategies that may preserve neural drive and potentially resulted in improved exercise performance.

Chapter 3: Study 1 77

3.7 ACKNOWLEDGMENTS

The authors sincerely thank Mr Logan Trim (Institute of Health and Biomedical

Innovation, Queensland University of Technology, Brisbane, Queensland, Australia) for his technical assistance with the immunoassay analysis.

78 Chapter 3: Study 1

Table 3.1. Posterior predicted mean [95% credible interval] for pre, post and 1 h post-exercise for neuromuscular variables in HOT or CON.

Pre-Exercise Post-Exercise 1 Hour Post-Exercise Variable CON HOT CON HOT CON HOT

MVC Torque 232 [208, 255] 229 [204, 253]; 208 [183, 233]* 201 [176, 226]*; 213 [188, 237]* 193 [168, 218]*†;

(N∙m) d: 0.40 [-1.53, 2.39] d: 0.85 [-1.15, 2.83] d: 2.42 [0.42, 4.41]

VA (%) 92.3 [89.8, 95.0] 92.7 [90.0, 95.3]; 90.5 [87.8, 93.2]* 87.9 [85.2, 90.8]*†; 91.0 [88.4, 93.7] 89.2 [86.6, 92.0]*†;

d: -0.41 [-2.36, 1.54] d: 2.50 [0.48, 4.45] d: 1.92 [-0.09, 3.87]

EMG VL (%) - - 81 [71, 92]* 86 [75, 97]*; 94 [83, 100] 84 [73, 95];

d: -0.64 [-2.58, 1.31] d: 1.41 [-0.56, 3.36]

EMG VM (%) - - 86 [78, 95]* 75 [66, 83]*†; 88 [79, 96]* 87 [78, 96]*;

d: 1.95 [0.00, 3.94] d: 0.03 [-1.97, 1.97]

Pt (N∙m) 66 [57, 74] 67 [58, 75]; 55 [45, 63]* 55 [46, 64]*; 64 [55, 72] 62 [53, 70];

d:-0.26 [ -2.24, 1.72] d: -0.26 [-2.22, -1.74] d: -0.65 [-1.35, 2.59]

TPT (ms) 65 [59, 72] 62 [55, 69]; 65 [58, 71] 57 [50, 64]; 64 [57, 71] 57 [51, 64];

d: 1.14 [-0.85, 3.31] d: 2.53 [0.55, 4.51] d: 2.18 [0.25, 4.15]

Chapter 3: Study 1 79

1/2 RT (ms) 52 [41, 63] 50 [39, 61]; 42 [31, 52] 36 [25, 47]*; 61 [50, 72] 55 [44, 66];

d: 0.32 [-1.67, 2.28] d: 0.83 [-1.17, 2.77] d: 0.83 [-1.19, 2.75]

CD (ms) 113 [96, 128] 109 [92, 124]; 103 [86, 119] 90 [72, 105]*; 122 [105, 137] 108 [90, 123];

d: 0.67 [-1.28, 2.58] d: 1.95 [-0.01, 3.93] d: 2.03 [0.01, 3.96]

RTD (N∙m∙s-1) 961 [914, 1010] 979 [920, 1041]; 939 [846, 1004] 967 [887, 1047]; 974 [910, 1038] 991 [916, 1069];

d: -0.72 [-2.64, 1.26] d: -0.79 [-2.75, 1.24] d: -0.52 [-2.44, 1.44]

RR (N∙m∙s-1) 704 [654, 756] 712 [644, 782]; 705 [632, 778] 719 [622, 819]; 688 [616, 759] 690 [592, 790];

d: -0.28 [-2.25, 1.74] d: -0.36 [-2.31, 1.58] d: -0.06 [-2.01, 1.93]

* indicates a statistical time difference compared to pre-exercise values in the same condition; † indicates a statistical condition difference at the same time point. Cohen’s d effect size [95% credible interval] is presented for condition parameter comparisons.

MVC: maximum voluntary contraction; VA: voluntary activation; EMG: electromyography; VL: vastus lateralis; VM: vastus medialis; Pt; peak twitch torque; TPt; time to peak torque; ½ RT: half relaxation time; CD: contraction duration; RTD: rate of torque development; RR: rate of relaxation.

80 Chapter 3: Study 1

Table 3.2. Posterior predicted mean [95% credible interval] for pre-, post and 1-h post-exercise for blood markers in HOT or CON.

Pre- Post- 1 Hour Post-Exercise Variable Exercise Exercise

CON HOT CON HOT CON CON

Endotoxin (EU∙mL-1) 0.009 [n.d., 0.009 0.010 [n.d., 0.011 [n.d., 0.046]; 0.011 [n.d., 0.045] 0.010 [n.d., 0.044];

0.042] [n.d., 0.045] d: -0.12 [-2.07, 1.90] d: 0.08 [-1.86, 2.04]

0.043];

d: -0.07

[-1.99,

1.91]

IL-1β (pg∙mL-1) 14.6 [7.2, 22] 15.9 [6.6, 16.2 [8.6, 14.8 [7.1, 22.7]; 14.3 [6.6, 22.1] 15.5 [6.3, 24.5];

25.2]; 23.6] d: 0.26 [-1.71, 2.27] d: -0.20 [-2.09, 1.82]

d: -0.25

[-2.23,

1.65]

TNF-α (pg∙mL-1) 48.2 [16.9, 66.2 55.4 [20.0, 61.4 [27.3, 88.9]; 38.9 [6.9, 63.8] 53.4 [16.9, 84.6]; 73.1] [32.4, 84.2] d: -0.50 [-2.53, 1.47] d: -1.05 [-3.03, 0.92]

95.1];

Chapter 3: Study 1 81

d: -1.61

[-3.56,

0.32]

I-FABP (ng∙mL-1) 0.481 [0.226, 0.435 0.619 [0.371, 1.043 [0.786, 1.298]*†; 0.587 [0.341, 0.846] 0.633 [0.375, 0.883]; 0.733] [0.188, 0.872] d: -3.26 [-5.26, -1.26] d: -0.34 [-2.31, 1.58]

0.691];

d: 0.35 [-

1.60,

2.35]

CLDN-3 (pg∙mL-1) 24 [19, 29] 29 [22, 26 [20, 32] 24 [18, 30]; 25[19, 30] 19 [14, 25];

36]; d: 0.57 [-1.38, 2.53] d: 1.52 [-0.40, 3.50]

d: -1.26

[-3.21,

0.74]

* indicates a statistical time difference compared to pre-exercise values in the same condition; † indicates a statistical condition difference at the same time point. Cohen’s d effect size [95% credible interval] is presented for condition parameter comparisons.

EU: endotoxin units; n.d.: not detected; IL-1β: interleukin 1 beta; TNF-α: tumour necrosis factor alpha; I-FABP: intestinal fatty acid binding protein; CLDN-3: claudin 3. Table 3.3. Posterior predicted values [95% credible interval] for physiological and perceptual measures during exercise.

82 Chapter 3: Study 1

Variable CON HOT

Mean HR (beats∙min-1) 139 [132, 145] 159 [152, 165]

Peak HR (beats∙min-1) 165 [157, 172] 182 [174, 188]

38.6 [38.5, 38.8] Mean Tc (°C) 38.2 [38.0, 38.3]

39.5 [39.3, 39.7] Peak Tc (°C) 38.5 [38.4, 38.7]

35.5 [35.3, 35.7] Mean Tsk (°C) 31.1 [30.9, 31.4]

35.9 [35.7, 35.2] Peak Tsk (°C) 32.0 [31.7, 32.3]

Mean RPE (AU) 12 [12, 13] 16 [16, 17]

Mean Thermal sensation (AU) 4 [4, 5] 7 [6, 7]

Mean Thermal comfort (AU) 2 [1, 2] 3 [3, 3]

HR: heart rate; Tc : core temperature; Tsk: skin temperature; RPE: rating of perceived exertion; AU = arbitrary units.

Chapter 3: Study 1 83

3.8 LINKING SECTION

The research presented in Chapter 3: Study 1 demonstrated that exercise in hot, as opposed to temperate, environments resulted in increased central fatigue and gastrointestinal damage. These findings tentatively support the proposed neuroinflammatory theory of hyperthermic fatigue, and highlight a possible link between exercise in the heat, intestinal damage and reduced voluntary activation of skeletal muscle. While speculative, intervention strategies that protect the gastrointestinal tract from damage may disrupt this inflammatory cascade, and could conceivably preserve central drive and therefore maintain athletic performance during exercise in the heat (Guy & Vincent, 2018; Lambert, 2009).

Therefore, Study 2 aimed to investigate the ergogenic potential of an acute intervention strategy, specifically glutamine supplementation, which could be rapidly employed to preserve athletic performance during hot environmental conditions. Glutamine has been previously reported to preserve intestinal barrier integrity in clinical patients and fixed-intensity exercise protocols (De-Souza & Greene, 2005; Pugh, Sage, et al., 2017;

Zuhl et al., 2015). Thus, supplementation with this amino acid before an externally- valid self-paced exercise task was theorised to protect against intestinal damage, reduce endotoxin translocation and preserve skeletal muscle recruitment, resulting in a faster time trial performance than a placebo.

Several issues that occurred during Study 1 were evaluated and resulted in methodology alterations for implementation in Study 2. For example, blood sample collection, storage and analysis was considerably revised in an attempt to improve sample sensitivity and quality. This included different blood collection tubes (serum to EDTA); clotting temperature (room temperature to 4 °C); centrifuge settings (3500

84 Chapter 3: Study 1

revolutions∙min-1 at 4 °C for 10 min to 15 min); and freezer storage temperature (-20

°C to -80 °C). Different sensitivity ELISA kits from different manufactures were also used, for example I-FABP (RayBiotech to Thermo Scientific), and samples diluted to reduce any matrix effect (TNF-α: 1:3; I-FABP: 1:2). Further, endotoxin analysis in

Study 1 was diluted 1:5 with sterile water, heated for 15 min at 75 °C, whereas Study

2 diluted samples 1:10 with magnesium chloride (MgCl) and heated for 45 min at 75

°C. Additional changes between Study 1 and Study 2 included the implementation of a more sensitive thermal comfort scale (4 point to 10 point scale), the use of a 40-point profile of mood states questionnaire at the start of each trial, and a change from wired thermistors collected via a Squirrel logger to wireless data loggers for core (T-Tec 7

RF 7-3E) and skin temperature (iButton thermocrons).

Chapter 3: Study 1 85

Chapter 4: Study 2 – Glutamine supplementation does not improve 20 km cycling time trial performance in the heat.

4.1 ABSTRACT

Aims: This study investigated the effect of acute glutamine supplementation on 20 km

cycling performance in the heat and neuromuscular function, as well as endotoxin,

gastrointestinal (GI) damage and markers of inflammation.

Methods: Twelve, well-trained male cyclists completed two, 20 km cycling time trials

(20TT) in an environmental chamber set at 35 °C and 50% RH. One hour before each 20TT,

participants ingested either glutamine (GLUT; 0.9 g∙kg-1 of fat-free mass) or placebo

(CON). Heart rate (HR), rectal temperature (Tre), skin temperature (Tsk) and perceptual

measures were recorded during the 20TT. Neuromuscular function (torque, voluntary

activation, evoked twitch) and recruitment (EMG) of the quadriceps were assessed pre- and

post-exercise. Venous blood samples were also collected at these time points to determine

the level of endotoxins, GI damage (intestinal fatty acid binding protein; I-FABP) and

inflammatory cytokines (interleukin-6, IL-6; tumour necrosis factor alpha, TNF-α).

Statistical analysis was undertaken using mixed models in a Bayesian framework.

Results: Completion time of the 20TT was not different between conditions (mean

difference: 11 s [-23, 44]). Relative to CON, glutamine supplementation did not affect HR,

Tre, Tsk or perceptual measures during the 20TT. However, GLUT was associated with

maintenance of voluntary torque at pre-exercise levels, while CON resulted in a significant

Chapter 4: Study 2 87

reduction (-9.3%) in this measure post-20TT. Although no change was observed in the

plasma concentration of endotoxins, post-20TT IL-6 levels were significantly elevated for

both conditions (4- to 5-fold increase over baseline).

Conclusion: Glutamine supplementation does not improve 20 km cycling performance in

the heat when compared to a placebo. However, glutamine appears to maintain gut integrity

and preserve knee extensor torque. Therefore, athletes competing in hot environmental

conditions may benefit from glutamine to reduce inflammation and gut damage.

4.2 INTRODUCTION

Hot environmental conditions hinder endurance exercise performance (Galloway & Maughan,

1997; Tatterson et al., 2000), and elevations in core body temperature may attenuate descending central drive to the exercising skeletal muscle (Cheung, 2007). While it has been theorised that this reduction in voluntary activation could limit exercise performance in the heat, a direct causal link between core temperature and central fatigue may be overly simplistic (Nybo &

González-Alonso, 2015). The ‘neuroinflammatory model’ presents a possible alternate explanation, emphasising the reduction in gastrointestinal (GI) blood flow during hyperthermic exercise, which may lead to the death of GI epithelial cells and disruption of the intestinal barrier (Lambert, 2008; Pires et al., 2016). This increased intestinal permeability and damage allow the translocation of endotoxins (lipopolysaccharides; LPS) into systemic circulation, initiating a pro-inflammatory cytokine release (e.g., interleukin-6 (IL-6) and tumour necrosis factor-alpha (TNF-α)) (Lim & Mackinnon, 2006). Elevated circulating levels of these inflammatory cytokines have been linked with transient fatigue-like symptoms and altered

88 Chapter 4: Study 2

motivational states (Dantzer, 2004). Accordingly, it is suggested that these signalling molecules could mediate the development of hyperthermic central fatigue, and thus, alter exercise performance in the heat (Vargas & Marino, 2014).

Given the alterations in gut barrier function and subsequent inflammatory response associated with exertional heat strain, prophylactic strategies capable of maintaining exercise performance in the heat are of interest (Guy & Vincent, 2018; Lambert, 2009). One such proposed solution is glutamine, a non-essential amino acid that serves as the primary energy source for gut mucosa

(Camilleri et al., 2012; Lambert et al., 2001). Glutamine supplementation augments the intestinal barrier and protects against septic shock in clinical populations (Castell, 2003; De-

Souza & Greene, 2005). Similar outcomes have been observed in exercise models, with glutamine supplementation (i.e., acute and 7-day loading) reducing intestinal damage and permeability after running in the heat (Pugh, Sage, et al., 2017; Zuhl et al., 2015; Zuhl et al.,

2014). Glutamine’s role in stabilising the intestinal lining is likely multifactorial, and research has emphasised two distinct pathways by which it may regulate barrier function: providing an essential energy source for gut mucosal epithelial cells; and enhanced expression of transmembrane proteins which form tight junctions of the intestinal barrier (Rao & Samak,

2012).

Previous literature has primarily investigated the mechanistic application of glutamine on gut function, as opposed to ergogenic sports performance, and thus have exclusively utilised fixed- pace exercise protocols (Lambert et al., 2001; Pugh, Sage, et al., 2017; Zuhl et al., 2015; Zuhl et al., 2014). To date, no study has examined the efficacy of glutamine on self-paced exercise performance in the heat. It could be postulated that the previously demonstrated protective effect of glutamine supplementation against exertional-endotoxemia might attenuate the

Chapter 4: Study 2 89

occurrence of a pro-inflammatory cytokine cascade. As elevated levels of inflammatory cytokines have been implicated in fatigue-like sensation and impaired efferent drive (Vargas &

Marino, 2017b; Vargas & Marino, 2014), a glutamine-mediated reduction in these cytokines may preserve voluntary activation of skeletal muscle and, therefore, maintain athletic performance.

The current study investigated the effects of acute glutamine supplementation on 20 km time trial cycling performance in the heat. A secondary aim was to investigate the effect of glutamine supplementation on neuromuscular function, inflammation and endotoxemia. It was hypothesised that glutamine would improve exercise performance (i.e., faster time trial completion) and preserve neuromuscular function in association with an attenuation of the development of central fatigue. Further, glutamine supplementation would maintain intestinal barrier integrity, observable via a diminished level of circulating endotoxins, inflammatory cytokines, markers of GI damage and subjective GI symptoms.

4.3 METHODS

Twelve male cyclists (age: 32 ± 6 y, body mass: 78 ± 8 kg, fat-free mass: 65 ± 6 kg, VO2max:

-1 -1 -1 61.0 ± 6.2 mL∙kg ∙min , Pmax: 430 ± 48 W; HRmax: 189 ± 8 beats∙min ) volunteered to participate in this study. All participants cycled at least twice per week (distance: 225 ± 93 km∙wk-1) and were classified as trained or well-trained athletes (performance level 3 or 4) (De

Pauw et al., 2013). Participants were non-smokers, free of any injury or illnesses, reported no history of GI or kidney diseases and were consuming no other supplements. All participants were informed of the study requirements and procedures before obtaining written and verbal

90 Chapter 4: Study 2

consent. The University Human Research Ethics Committee approved this project before the commencement of any testing.

Experimental overview

Participants visited the laboratory on three separate occasions and completed equipment and procedural familiarisation (i.e., 20 km cycling time trial and neuromuscular testing) followed by two experimental sessions, involving a glutamine supplementation trial (GLUT) and a placebo control trial (CON). The experimental sessions were completed in a double-blind, randomised, crossover manner, each separated by ≥ 7 days. All testing sessions involved a 20 km cycling time trial (20TT) that was completed in mean ± standard deviation environmental conditions of 35.1 ± 0.5 °C and 51 ± 4% RH. Participants were asked to diarise their physical activity, food and fluid intake for the 24 h before the first experimental trial and to replicate this for the subsequent session. Participants were also asked to abstain from caffeine and alcohol for 12 h, and strenuous physical activity for 48 h, before each experimental trial. Compliance to these requests was assessed before each session.

Experimental trials

Each experimental trial was matched for time of day (±2 h), and upon arrival, participants ingested a GLUT or CON solution from an opaque drink bottle. The GLUT solution compromised of powdered glutamine (0.9 g∙kg-1 of fat-free mass; L-Glutamine; Bulk Nutrients,

Grove, Australia) mixed with 450 mL of room-temperature water and 50 mL of sugar-free lemon cordial (Diet Rite, Tru Blu Beverages, Bundamba, Australia). This dosage has been previously demonstrated to attenuate GI permeability when compared to a placebo (Pugh, Sage, et al., 2017; Zuhl et al., 2015). The CON beverage contained the matched fluid (i.e., water and cordial) but was void of any supplement. Body mass, fat mass and fat-free mass were measured

Chapter 4: Study 2 91

during the familiarisation session to calculate the supplement dose using multi-frequency bioelectrical impedance analysis (MC-780MA; Tanita Corp., Tokyo, Japan). All solutions were independently mixed immediately before ingestion, to limit glutamine degradation in the aqueous solution.

After resting measures and instrumentation, participants donned their cycling apparel (i.e., bib, socks and cycling shoes). Participants were then seated in a climate-controlled laboratory (24.2

± 1.0 °C, 60 ± 4% RH) for 5 min to collect baseline physiological and perceptual data, including an abbreviated profile of mood states (POMS) questionnaire (Grove & Prapavessis, 1992), before entering the climate chamber (35.1 ± 0.5 °C, 51 ± 4% RH) and undertaking a 20TT. All

20TTs were completed on a Velotron Pro cycle ergometer (RacerMate Inc., Washington, USA), commencing from a seated, stationary position. Velotron 3D software (Version NB04.1.0.2101,

RacerMate Inc., Washington, USA) provided elapsed distance feedback over a computer- simulated flat cycling course. A constant background scene was shown, and no simulated wind resistance or computer opponents were displayed. Participants were instructed to complete the

20 km distance as quickly as possible and could freely alter gearing using a hood-mounted toggle-switch. No encouragement was provided during the 20TT and participants did not receive any performance data until they had completed both experimental sessions. Highly reliable 20TT performance results (intra-class correlation (ICC) = 0.93) are reported using familiarised, trained cyclists in our laboratory (Borg et al., 2018). The time delay between completion of the supplement consumption and commencement of the 20TT was 61 ± 7 min.

This lead in time was chosen to align with peak plasma glutamine concentrations (Ziegler et al., 1990). Fluid was withheld during the 20TT and until after the collection of post-exercise measures.

92 Chapter 4: Study 2

Thermophysiological measures

Heart rate (HR) was recorded using a chest strap (Polar Electro Oy, Kempele Finland) and

2 computer software (Polar Team , Kempele Finland). Rectal temperature (Tre) was assessed via a thermistor (449H; Henleys Medical, Hertfordshire, UK) inserted 12 cm past the anal sphincter and connected to a wireless data logger (T-Tec 7 RF 7-3E) set to record every 2 s. No participant reached the Tre termination threshold of 40 °C during the exercise bout.

Skin temperature was recorded every 5 s with wireless iButton thermocrons (DS1922L-F50 iButtons, Maxim Integrated, San Jose, USA) attached with adhesive tape (3.8 cm width, Leuko

Sportstape Premium; Beiersdorf, Hamburg, Germany) to four sites: posterior neck, right scapula; posterior left hand; and mid-anterior shin. Mean skin temperature (Tsk) was calculated according to the published four-site formula (ISO 9886, 2004).

A mid-stream urine sample was collected before each experimental session to assess urine specific gravity (PAL-10S; Atago Co. Ltd, Tokyo, Japan) and urine colour (scale: 1-8 AU)

(Armstrong et al., 2010). Body fluid loss was calculated via pre- to post-exercise nude weight change over time, using a set of calibrated scales (WB-110AZ; Tanita Corp., Tokyo, Japan).

Perceptual measures were recorded every 2 km during the 20TT. These included perceived effort (Borg, 1970), a 16-point thermal sensation scale (scale: 0 = ‘unbearably cold’ to 8 =

‘unbearably hot’) (Young et al., 1987) and a modified 10 point thermal comfort scale (1 =

‘comfortable’ to 5 = ‘extremely uncomfortable’) (Gagge et al., 1967). Participants completed a post-trial GI distress form (adapted from Pfeiffer, Cotterill, Grathwohl, Stellingwerff, and

Jeukendrup (2009)) to assess the incidence of gut symptoms (abdominal cramps, nausea, urge to vomit, urge to defecate, side ache, flatulence) and severity on a 100 point scale (0 = none;

100 = severe), and also rated the session exertion (sRPE) (Borg, 1998).

Chapter 4: Study 2 93

Neuromuscular function and voluntary activation (VA)

The neuromuscular function of the right knee extensor muscles was assessed pre- and immediately post-exercise using a Biodex System 3 dynamometer (Biodex Medical Systems,

Shirley, New York, USA). Participants were seated in an upright position (back of the chair adjusted to 95° from horizontal) and the chair adjusted so the fulcrum of the lever was aligned with the right lateral epicondyle of the femur. Participants were then secured using chest, waist, right thigh and ankle straps.

Voluntary activation of the knee extensors was assessed using the twitch interpolation technique on the right femoral nerve. Briefly, self-adhesive gel electrodes (Pals; Axelgaard

Manufacturing Co. Ltd., Fallbrook, CA) were positioned on the right femoral triangle and the border of the right gluteal fold. The current was applied to the nerve by a Digitimer DS7AH stimulator (Digitimer Ltd., Welwyn Garden City, Hertfordshire, England) using a single, 100

μs, square-wave pulse. A twitch ramp procedure was completed at the start of each session with stimulus of increasing current delivered to a resting muscle until a plateau in twitch torque was observed. The current was then increased by an additional 10% to guarantee supramaximal stimulation.

Participants completed a warm-up of eight, 5 s isometric knee extensions before resting for 2 min and then completing a set of five, 5 s maximal isometric knee extensions at 90° knee flexion, with a 30 s rest between each repetition. To ensure each repetition was a maximal effort, participants received loud verbal encouragement and visual feedback of their voluntary torque.

The primary investigator manually triggered a femoral nerve stimulation when the participant demonstrated a plateau in maximal voluntary torque during each MVC. An additional stimulation was triggered upon the completion of each MVC to assess peripheral contractile

94 Chapter 4: Study 2

properties. Another five, 5 s MVCs with interpolated twitch were also completed immediately post-exercise.

The twitch interpolation formula used to calculate VA was as follows: VA (%) = (1 – amplitude/resting control twitch torque) * 100 (Allen et al., 1995). Only MVC repetitions that demonstrated a torque plateau and were within 90% of the maximal voluntary torque of that set were included in the mean VA calculations. Maximal voluntary isometric torque was considered the mean torque value of 25 ms preceding the interpolated twitch, and the peak torque recorded in the 100 ms following the stimulus was considered the superimposed torque.

The difference between the maximal voluntary isometric torque and the superimposed torque value was considered the amplitude. Participants were extensively familiarised with the neuromuscular testing procedures during the initial laboratory visit and were able to consistently reproduce MVC torque values. These neuromuscular assessments of maximal voluntary torque and VA by our laboratory have been found to be highly reliable, with a calculated ICC’s of 0.94 and 0.94, respectively.

Evoked twitch contractile properties

Evoked twitch contractile data were averaged for each time point and analysed for peak twitch torque (Pt; maximum torque during the evoked twitch); time to peak torque (TPt; time interval between the first rise in torque above baseline to peak torque); half relaxation time (½ RT; time interval for torque to drop by half of peak torque); contraction duration (CD; time to peak torque plus half relaxation time); rate of torque development (RTD; slope of twitch-torque curve from onset to peak torque); rate of relaxation (RR; slope of twitch-torque curve from peak torque to half relaxation time).

Chapter 4: Study 2 95

Surface electromyography (EMG)

Muscle activation of the right vastus medialis (VM) and vastus lateralis (VL) were measured using surface EMG during the neuromuscular assessments. All sites were shaved, abraded and swabbed with alcohol before electrode placement. Electrodes (Ambu Blue Sensor N-00-S;

Ambu A/S, Ballerup, Denmark) were positioned in line with the muscle fibres for each muscle, and an earth electrode was attached to the lateral femoral epicondyle. Electrodes remained attached during the 20TT to ensure consistency between the pre- and post-neuromuscular assessments.

EMG data were sampled at 1 kHz through a 16-bit PowerLab 26T AD unit (AD Instruments,

Sydney, Australia) (amplification=1000; common mode rejection ratio=110 dB) and band-pass filtered (20-500 Hz). Raw EMG data were smoothed using the RMS method (100 ms window) through LabChart 8.1.5 software (AD Instruments, New South Wales, Australia). Voluntary muscle activation was considered the mean smoothed EMG value of a 500 ms period preceding each interpolated stimulus. Post-exercise EMG amplitudes were then compared as a relative change to the mean values obtained during pre-exercise MVCs and presented as a percentage.

Blood markers and biochemical analysis

Pre- and post-exercise blood samples were drawn from an antecubital venipuncture using a sterile butterfly needle (21G, BD, North Ryde, Australia) into EDTA vacutainer tubes (BD,

North Ryde, Australia) and immediately centrifuged at 3500 RPM for 15 min at 4˚C. Following centrifugation, pyrogen-free aliquots of plasma were frozen at -80˚C for a maximum of 6 months before analysis.

Circulating IL-6 (HS600B; R&D Systems, Minneapolis, USA), TNF-α (EK-0001; elisakit.com,

Melbourne, Australia), and I-FABP (EHFABP2; Thermo Scientific, Fredrick, USA)

96 Chapter 4: Study 2

concentrations were determined using quantitative sandwich enzyme-linked immunosorbent assays (ELISA), prepared in accordance with the manufacturer’s protocols. All samples were adequately diluted (IL-6, 1:4; TNF-α, 1:3; I-FABP, 1:2) to reduce interference and avoid a matrix effect. Absorbance was quantified using a SpectroStar Nano (BMG Labtech, Offenburg,

Germany) with appropriate wavelength subtraction to correct for optical imperfections in the plates. Intra-assay coefficient of variation (CV) was calculated as 8.0% (IL-6), 8.0% (TNF-α) and 7.1% (I-FABP), respectively.

Endotoxin concentrations were quantified using a kinetic chromogenic Limulus amebocyte lysate (LAL) assay kit (50-650U), as instructed by the manufacturer (Lonza, Walkersville,

USA). In short, the E. coli 055:B5 endotoxin standard was reconstituted with LAL reagent water and vortexed for 15 min to produce a 50 EU∙mL-1 stock solution. Serial dilutions of this stock standard (5, 0.5, 0.05, 0.005 EU∙mL-1) were prepared in duplicate, with each solution vortexed for at least 2 min between dilutions. Plasma samples were heated in a heat block for

45 min at 75 °C to inactivate endotoxin-neutralising agents and diluted at 1:10 with magnesium chloride (MgCl) to overcome the chelating effects of EDTA. Plates were incubated for 20 min at 37 °C in a SpectroStar Nano (BMG Labtech, Offenburg, Germany) before the addition of

100 μL of reconstituted Kinetic-QCL™ Reagent to each well. The plates were read at 405 nm, which was then repeated every 61 s for 118 cycles. A log/log (time for the sample to increase

0.2 absorbance units/concentration) linear correlation of each standard was calculated (r = -

0.999) and used to determine the unknown concentration of the samples. Sterile tips (Biopur epTIPS; Eppendorf AG, Germany) were used for all pipetting to reduce the risk of any endotoxin contamination.

Chapter 4: Study 2 97

Statistical Analysis

Data were analysed using mixed modelling in a Bayesian framework, as previously described

(Mengersen et al., 2016). Exploratory raw data plots were inspected for normality before

Markov chain Monte Carlo (MCMC) procedures were used to generate expected posterior values, using the ‘rjags’ package through RStudio Desktop (RStudio v 1.1.447, Boston, USA).

Specifically, an initial burn-in of 1k followed by 50k iterations and thinned by a factor of 10 to generate 5k posterior samples. All model beta values utilised vague informative priors (mean =

0; precision = 0.001), apart from neuromuscular variables which integrated data from a research project employing the same protocol and participants of similar training status (Osborne PhD

Study 1, unpublished). The final model for each analysis was dependent on the goodness of fit based off an extracted deviance information criterion value.

The mean and 95% credible interval [CI] values were calculated for each posterior parameter, and interval-based testing was utilised to identify parameter differences, with statistical difference accepted when 95% CI ≠ 0. Parameters were further explored to identify within- or between-group mean differences (MD), with statistical difference again accepted when 95% CI

≠ 0. Cohen’s d effect sizes were used to evaluate the magnitude of each difference and were categorised as small (0.2), moderate (0.5) and large (0.8) (Cohen, 1988).

4.4 RESULTS

No differences in psychological state upon arrival or baseline hydration measures (USG and urine colour) were found between conditions (Table 4.1). Pre-exercise body mass, resting Tre and Tsk were also similar between GLUT and CON. The absence of any statistical difference in

98 Chapter 4: Study 2

baseline measures indicates that participants commenced each 20TT in a similar psychological and physiological state.

Post-exercise fluid loss, both absolute and relative, were also not statistically different between conditions (GLUT: 1.3 kg loss [1.0, 1.3] and 1.5 % loss [1.3, 1.7]; CON: 1.3 kg loss [1.1, 1.4] and 1.6 % loss [1.4, 1.8]; Table 4.1). While the incidence of GI issues was not found to be statistically different between conditions, a small-to-moderate trend for higher severity of reported symptoms was seen in CON (Nausea: d = 0.26; Vomit: d = 0.48; Stitch: d = 0.27;

Table 4.2). Supplementation did not affect the participant rating of perceived sessional exertion following each 20TT (Table 4.1). Participants were successfully blinded to the intervention with a calculated James’s Blind Index of 0.67 [0.48, 0.85] (James, Bloch, Lee, Kraemer, &

Fuller, 1996).

Exercise

Glutamine supplementation did not affect 20TT completion time, with a mean difference of 11 s [-23, 44] between conditions (CON: 33:22 min [31:52, 34:54]); GLUT: 33:33 min [32:02,

35:06]). No statistical differences were observed in mean 20TT power, cadence and speed, between the two conditions (Table 4.3).

Statistical analysis revealed only a significant time parameter difference, without a significant condition parameter or interaction effect, for HR, Tre, Tsk, RPE, thermal sensation and thermal comfort over the 20TT (Figure 4.1).

Chapter 4: Study 2 99

200 39.5

39.0 175 AB C) q q 38.5

) 150 -1 min ˜ 38.0 125

HR (beats 37.5 100 Core temperature (

37.0 75

CON GLUT CON GLUT

36.5 50 0 2 4 6 8 10 12 14 16 18 20 0 2 4 6 8 10 12 14 16 18 20

Distance (km) Distance (km)

38 20

CD19 37

18 C)

q 36 17

35 16 RPE (AU) 15 34 Skin ( temperature 14

33 13 CON GLUT CON GLUT

32 12 0 2 4 6 8 101214161820 0 2 4 6 8 101214161820

Distance (km) Distance (km)

8 5 EF 7 4

6 3

5 2 Thermal Comfort (AU) Thermal Sensation (AU)

4 1

CON GLUT CON GLUT

3 0 0 2 4 6 8 10 12 14 16 18 20 0 2 4 6 8 10 12 14 16 18 20

Distance (km) Distance (km)

Figure 4.1. Physiological and perceptual measures during 20 km TT in the heat following ingestion of glutamine (GLUT) or placebo (CON). (A) HR; (B) Tre; (C) Tsk; (D) RPE; (E) Thermal sensation; (F) Thermal Comfort. Data displayed as posterior predicted.

100 Chapter 4: Study 2

Neuromuscular

The 20TT resulted in VA impairment, from pre-exercise values, for both conditions (CON: -

5.5 % [-1.7, -9.1], d = 2.9 [0.9, 4.9]; GLUT: -5.2 % [-1.5, -8.8], d = 2.8 [0.8, 4.8]). Time was also a significant parameter for the remaining muscle contractile properties of Pt, TPt, ½ RT and CD, which were all reduced from initial values (Table 4.4). However, glutamine was not observed to alter any of these measures. In contrast, modelling revealed a time effect for MVC torque (βTIME = -22 [-37, -7]), with evidence supporting statistical differences pre- to post- exercise for CON (MD = -22 N∙m [-37, -7]), but this was not observed in GLUT (MD = -12

N∙m [-27, 4]). No statistically different parameter effects were observed for muscle activation

(relative %EMG of vastus lateralis and vastus medialis), RTD or RR (Figure 4.2; Table 4.4).

Chapter 4: Study 2 101

·

280 † 260

m) 240 ˜ ˜

220

200

MVC Torque (N 180

160

CON GLUT

100

†† 95

90

85 Voluntary Activation (% )

80 Pre-Exercise

Post-Exercise 0 CON GLUT

Figure 4.2. The neuromuscular function of the knee extensors pre and post-20TT following ingestion of glutamine (GLUT) or placebo (CON). Data displayed as posterior predicted mean with ± 95% CI and overlayed with raw individual responses. † indicates a significant difference from pre-exercise values in the same condition.

102 Chapter 4: Study 2

Bloods

Cycling in the heat resulted in a significant rise in the concentration of IL-6 compared to pre- exercise levels for GLUT (MD = 2.7 pg∙mL-1 [1.4, 4.0], d = 4.1 [2.1, 6.1]) and CON (MD = 3.2 pg∙mL-1 [1.9, 4.6], d = 4.7 [2.7, 6.6]), although no condition differences were observed (Figure

4.3; Table 4.5). Plasma TNF-α levels also significantly increased above pre-exercise values for

CON (MD = 1.8 pg∙mL-1 [0.5, 3.1], d = 2.8 [0.8, 4.8]). A similar finding was also observed for

I-FABP, with evidence to support a statistical rise pre- to post-exercise for CON (MD = 0.441 ng∙mL-1 [0.235, 0.656], d = 4.10 [2.18, 6.10). However, GLUT appeared to attenuate this post- exercise increase in TNF-α and I-FABP, with no clear statistical difference observed between time-points (MD = 0.5 pg∙mL-1 [-0.7, 1.8]; and 0.206 ng∙mL-1 [-0.025, 0.418], respectively). No statistically different main effects for condition or time were identified for endotoxin.

Chapter 4: Study 2 103

8 †

6 ) -1 mL ˜

(pg 4 D D TNF- 2

0 CON GLUT

7 †† 6

5 ) -1

4 mL ˜

3 IL-6 (pg 2

1

0 CON GLUT

† 1.25

1.00 ) -1

mL 0.75 ˜

0.50 I-FABP (ng

0.25

Pre-Exercise Post-Exercise 0.00 CON GLUT

Figure 4.3. Plasma concentration of TNF-α, IL-6 and I-FABP at pre- and post-20TT timepoints. Data displayed as posterior predicted mean with ± 95% CI and overlayed with raw individual responses. † indicates a significant difference from the pre-exercise value of the same condition.

104 Chapter 4: Study 2

4.5 DISCUSSION

The main finding of this study was that acute glutamine supplementation did not improve completion time, mean power, speed or cadence for a 20 km self-paced cycling trial in the heat

(Table 4.3). This outcome was reflected in the comparable values of physiological and perceptual strain for CON and GLUT over the 20 km distance (Figure 4.1). Furthermore, glutamine supplementation was not found to preserve voluntary activation more than a placebo, with a statistically significant reduction in this measure seen in both conditions following strenuous cycling in a hot environment (Table 4.4). These data suggest that athletes competing in shorter-duration events (i.e., criterium racing) under hot conditions are unlikely to obtain a performance benefit from acute glutamine supplementation. However, this study may provide preliminary evidence that glutamine supplementation may maintain knee extensor torque

(Figure 4.2) and protect against GI damage (i.e., I-FABP; Figure 4.3) and blunt increases in certain inflammatory cytokines (i.e., TNF-α; Figure 4.3).

To the best of our knowledge, this is the first study to investigate the effect of glutamine supplementation on self-paced exercise performance in the heat. Previous literature has focused primarily on glutamine supplementation during fixed-paced exercise tasks in the heat (Lambert et al., 2001; Pugh, Sage, et al., 2017; Zuhl et al., 2015; Zuhl et al., 2014). Collectively, this research suggests that increased GI permeability and damage, stemming from strenuous exercise in the heat, can be ameliorated by glutamine. Similar outcomes have been observed in animal studies (Singleton & Wischmeyer, 2006). Although the current study did not specifically assess GI permeability, glutamine supplementation was observed to attenuate the rise in I-

FABP compared to a placebo, which is in agreement with the dose-response findings of Pugh,

Sage, et al. (2017). This data supports the concept that glutamine may provide a stabilising and

Chapter 4: Study 2 105

protective effect for the gut mucosa that comprises the intestinal barrier (Castell, 2003). As intestinal barrier dysfunction has been linked to increased inflammatory stress, glutamine may play a crucial role in reducing gut damage and therefore diminish the resultant pro- inflammatory cascade.

While glutamine supplementation was observed to reduce the severity of GI damage, no significant change in endotoxin concentration was seen at any time-point or condition. This finding was surprising given the previously reported increase in circulating LPS following strenuous exercise in the heat (Lambert, 2008). This outcome could be explained by the modality (cycling), or relative brevity (~33 min) of the current exercise protocol that contrasts with the ultramarathon (Brock-Utne et al., 1988; Gill, Hankey, et al., 2015) and heat stress trials

(i.e., running till voluntary exhaustion) (Lim et al., 2009) previously examined. In contrast, our data support the findings of van Wijck et al. (2011), who reported that cycling for 60 min in the heat increased gut damage (I-FABP) and permeability, but did not result in detectable endotoxemia. This discrepancy in exercise duration between laboratory protocols (van Wijck et al., 2011) and prolonged, field-based competitions (Gill, Hankey, et al., 2015) could, at least partially, rationalise these inconsistent findings in exertional endotoxemia. Thus, it could be argued that LPS translocation may have occurred in the current study if a longer exercise task was utilised. Due to this potential limitation of the present study, the previously-hypothesised effect of glutamine supplementation on endotoxin translocation remains to be elucidated

(Lambert, 2009).

Although there was little evidence to support a heat-mediated increase of circulating endotoxins, the current study did observe that TNF-α concentrations were only significantly elevated in the control condition (Figure 4.3). These findings suggest that glutamine

106 Chapter 4: Study 2

supplementation may modulate the release of the pro-inflammatory cytokine, TNF-α, which is in agreement with previous research using in-vitro human PBMC models (Wischmeyer et al.,

2003) as well as fixed-intensity protocols (Zuhl et al., 2015). Notably, however, this attenuated

TNF-α response did not translate into cycling performance improvements (Table 4.5). One explanation for this lack of ergogenic benefit may be due to the similar rise in IL-6 concentrations between conditions (Figure 4.3), with no significant differences observed post- exercise. The increased levels of IL-6 in both conditions may have occurred due to the multi- origin nature of the molecule, which is considered both a cytokine and a myokine due to its expression during skeletal muscle contractions (Pedersen et al., 2003). The non-significant difference in 20TT completion time between conditions would suggest a comparable level of muscle contractile activity, and thus a similar level of IL-6 released. Further, as IL-6 has been implicated in contributing to the feeling of fatigue (Vargas & Marino, 2014), it could be argued that the comparable post-exercise levels of this molecule may explain the similar cycling performance, session RPE and post-exercise voluntary activation that were observed between conditions (Table 4.1, Table 4.3 & Table 4.4).

Cycling in a hot environment for 20 km was observed to reduce maximal voluntary knee extensor torque in CON significantly (MD: 22 N∙m [7, 37]), but not GLUT (MD: 12 N∙m [-4,

27]). This finding is further confounded by the similar decreases in VA and peak twitch torque, suggesting a comparable development of central and peripheral fatigue in both conditions

(Table 4.4). Research by Périard and colleagues (2011b) evidenced impairments in voluntary torque following a 40 km cycling time trial in the heat, with the authors implicating the prolonged dynamic muscular contractile activity of the exercise task. In particular, high muscle temperatures have been shown to result in more responsive contractile properties, namely a

Chapter 4: Study 2 107

faster TPt and ½RT (Todd et al., 2005). While a similar occurrence was seen in the current study, a faster TPT and ½RT were found in both conditions, and so cannot fully explain the observed post-exercise torque difference (Table 4.4).

One potential explanation for the difference in knee extensor torque may be due to the increased levels of TNF-α detected in the post-exercise CON group (+41% increase over resting values;

Table 4.5). As previous research has demonstrated that elevated TNF-α results in inhibited contractile function of myofilaments in animals via an oxidant cascade (Hardin et al., 2008;

Reid et al., 2002), we speculate that a similar occurrence in the current study could potentially explain the observed reduction in CON torque. However, we also recognise that the post- exercise impairment in torque for both conditions could be considered marginal (CON: 9% reduction; GLUT: 5% reduction) and may not reflect a meaningful physiological change.

Further research should aim to elucidate the mechanistic effects of glutamine supplementation on hyperthermic exercise, specifically neuromuscular function.

4.6 CONCLUSION

In summary, this study established that acute glutamine supplementation does not improve self- paced cycling performance in the heat. This intervention did not result in physiological or perceptual changes during the 20 km exercise bout when compared with a placebo. Despite strenuous exercise (mean HR: 88% HRmax) and elevated core temperatures (final temperature:

~39.0°C), there was no apparent development of exertional endotoxemia. However, glutamine supplementation was observed to potentially provide a protective effect for gut mucosa against ischaemic injury and may have reduced circulating levels of the pro-inflammatory cytokine,

TNF-α. Glutamine may also attenuate peripheral fatigue, and thus improve muscle contractile

108 Chapter 4: Study 2

force. Future studies should investigate the application of glutamine in reducing inflammation and gut damage in prolonged duration exercise or multi-day sporting competitions.

4.7 ACKNOWLEDGMENTS

The authors sincerely thank Mr Logan Trim (Institute of Health and Biomedical Innovation,

Queensland University of Technology, Brisbane, Queensland, Australia) for his technical assistance with the immunoassay analysis.

Chapter 4: Study 2 109

4.8 TABLES

Table 4.1. Baseline, post and pre-post posterior predicted data (mean [95% credible interval]).

Variable Control Glutamine

POMS 7 [1, 14] 8 [2, 13] 37.1 [36.9, 37.3] Resting Tre (°C) 37.0 [36.8, 37.2] 33.2 [33.0, 33.5] Resting Tsk (°C) 33.1 [32.9, 33.4]

USG 1.009 [0.969, 1.048] 1.016 [0.977, 1.055]

Urine colour 2.8 [2.0, 3.7] 3.3 [2.4, 3.7]

Pre-post body mass loss (kg) 1.3 [1.1, 1.4] 1.3 [1.0, 1.3]

Pre-post body mass loss (%) 1.6 [ 1.4, 1.8] 1.5 [1.3, 1.7]

Sessional RPE 8.5 [7.8, 9.2] 8.5 [7.8, 9.2]

Chapter 4: Study 2 111

Table 4.2. Gastrointestinal distress post-exercise questionnaire (N = response rate/12; mean of responders (range))

GI Distress Symptom Control Glutamine

Cramp N = 0; 0 (0 – 0) N = 1; 50 (0 – 50)

Nausea N = 4; 19 (0 – 25) N = 4; 13 (0 – 12.5)

Urge to defecate N = 0; 0 (0 – 0) N = 0; 0 (0 – 0)

Urge to vomit N = 4; 27 (0 – 50) N = 3; 12 (0 – 25)

Stitch/pain in gut N = 3; 33 (0 – 75) N = 3; 16 (0 – 30)

Flatulence N = 0; 0 (0 – 0) N = 0; 0 (0 – 0)

112 Chapter 4: Study 2

Table 4.3. Posterior predicted mean data for 20TT exercise parameters (mean [95% credible interval])

Variable Control Glutamine

Completion Time (min:sec) 33:22 [31:52, 34:54] 33:33 [32:02, 35:06]

Power (W) 256 [226, 285] 251 [221, 280]

Cadence (RPM) 97 [91, 103] 98 [92, 104]

Speed (km∙h-1) 36.3 [34.7, 37.8] 36.1 [34.5, 37.6]

Heartrate (beats∙min-1) 167 [166, 173] 166 [165, 172]

Core temperature (°C) 38.0 [37.9, 38.2] 38.0 [38.0, 38.2]

Skin temperature (°C) 36.0 [36.0, 36.3] 36.1 [36.0, 36.3]

RPE (AU) 16.0 [15.3, 16.8] 16.2 [15.5, 16.9]

Thermal sensation (AU) 6.2 [6.0, 6.5] 6.3 [6.1, 6.7]

Thermal comfort (AU) 3.5 [3.1, 3.9] 3.6 [3.3, 4.0] AU = arbitrary units.

Chapter 4: Study 2 113

Table 4.4. Posterior predicted data for pre- and post-20TT neuromuscular function parameters (mean [95% credible interval]).

Pre-Exercise Post Exercise Variable Control Glutamine Control Glutamine

MVC Torque (N∙m) 233 [ 201, 267] 231 [198, 264] 211 [ 178, 246]† 220 [187, 253] † VA (%) 91.5 [87.1, 95.7] 92.4 [87.9, 96.7] 86.0 [81.6, 90.4]† 87.3 [82.8, 91.5]

EMG VL (%) - - 76.9 [64.3, 88.8] 81.5 [69.0, 93.8]

EMG VM (%) - - 71.4 [60.9, 81.3] 78.4 [67.8, 88.4] † Pt (N∙m) 64.8 [53.3, 76.5] 60.5 [48.7, 72.4] 55.2 [43.7, 67.1]† 53.1 [41.6, 65.0] † TPT (ms) 76 [70, 81] 80 [75, 85] 68 [62, 73]† 69 [64, 76] † 1/2 RT (ms) 61 [49, 72] 60 [48, 72] 40 [28, 52]† 44 [32, 55] † CD (ms) 132 [115, 145] 138 [121, 152] 105 [87, 118]† 110 [93, 124]

RTD (N∙m∙s-1) 834 [673, 998] 790 [628, 957] 826 [665, 996] 778 [611, 949]

RR (N∙m∙s-1) 590 [443, 733] 556 [408, 702] 631 [481, 774] 597 [450, 745] * indicates a statistical difference between conditions at the same time-point; † indicates a statistical difference between pre- and post-exercise values of the same condition.

114 Chapter 4: Study 2

Table 4.5. Posterior predicted data for blood markers (mean [95% credible interval]).

Pre-Exercise Post Exercise Variable Control Glutamine Control Glutamine

LPS (EU∙mL-1) 0.22 [0, 1.19] 0.04 [0, 0.82] 0.87[ 0.04, 1.72] 0.08 [0, 0.85] † IL-6 (pg∙mL-1) 1.0 [0, 2.0] 0.7 [0, 1.7] 4.2 [3.2, 5.2]† 3.3 [2.3, 4.3]

TNF-α (pg∙mL-1) 3.8 [2.9, 4.8] 3.3 [2.3, 4.2] 5.6 [4.6, 6.6]*† 3.8 [2.8, 4.7]*

I-FABP (ng∙mL-1) 0.533 [0.337, 0.728] 0.610 [0.413, 0.808] 0.973 [0.780, 1.163]† 0.815 [0.621, 1.017] * indicates a statistical difference between conditions at the same time-point; † indicates a statistical difference between pre- and post-exercise values of the same condition.

Chapter 4: Study 2 115

4.9 LINKING SECTION

The glutamine supplementation protocol for Study 2 was contextualised as an acute intervention strategy that could be rapidly completed before exercise in a hot environment. For example, athletes in competition during a sudden heat-wave or military personnel being rapidly deployed to hot environmental locations. A time- sensitive ergogenic aid would be critical in this context of sudden heat exposure; however, many competitive events are regularly scheduled to occur in summer months and/or are well known for the thermally-stressful environmental conditions (i.e., UCI

Tour Down Under). This knowledge provides athletes with an extended timeframe in which to implement intervention strategies that improve exercise performance in heat.

Therefore, the aim of Study 3 was to investigate the effect of short-term heat acclimation training on self-paced cycling performance in the heat. Heat acclimation training has been previously found to result in improved cardiovascular stability, partially arising from an expansion in plasma volume (Périard et al., 2015). This enhanced cardiac output has been proposed to attenuate the reliance on splanchnic blood redistribution and therefore preserving gut blood flow and intestinal barrier stability (Lambert, 2004). Conceivably, this may reduce endotoxin translocation and the associated pro-inflammatory cascade (Kuennen et al., 2011). Only a limited number of studies have investigated the effect of heat acclimation training on endotoxemia and exercise performance, and to our knowledge, none have also measured central fatigue and neuromuscular recruitment. Thus, it was hypothesised that heat acclimation training would attenuate intestinal damage and inflammation, while maintaining central drive and enhancing self-paced cycling performance.

116 Chapter 4: Study 2

Following the completion of Study 2, the methodology was reappraised and improved to enhance the sensitivity and validity of data collection during Study 3. For example, the pre-20TT baseline was extended from 5 min to 20 min, to better capture possible cardiovascular adaptations arising from heat acclimation training in Study 3. Blood lactate collection was implemented pre- and post-20TT to provide additional data around exercise intensity. Verbal feedback from participants in Study 2 also highlighted that the 40-point abbreviated POMS questionnaire was confusing and poorly-received. In addition, this questionnaire did not specifically assess certain areas of interest for Study 3, such as fatigue or sleep quality. Therefore, for Study 3 a different pre-exercise POMS questionnaire was utilised; the 5-point psychological wellbeing sheet (McLean, Coutts, Kelly, McGuigan, & Cormack, 2010). Study 3 also retained the similar changes made between Study 1 and Study 2 such as; more sensitive thermal comfort scale, use of wireless thermistor data loggers and altered blood sample storage and analysis. For more information on these changes between Study 1 and the subsequent studies, please see Section 3.8.

Chapter 4: Study 2 117

Chapter 5: Study 3 – Short-duration heat acclimation training improves 20 km cycling performance in the heat and enhances knee extensor strength.

5.1 ABSTRACT

Aims: This study examined the effect of 5 days of heat acclimation (HA) training on 20

km cycling (20TT) performance in the heat, neuromuscular function and inflammation.

Methods: Eight recreationally-trained males completed a baseline 20 km cycling time trial

-1 (20TTINITIAL) followed by 5-day cycling training block (60 min∙day at 50% Pmax) and a

final 20 km time trial (20TTFINAL), in a counterbalanced, cross-over design. Training was

conducted in either a hot (HA: 34.9 ± 0.7 °C, 53 ± 4% relative humidity (RH)) or

thermoneutral (CON: 22.2 ± 2.6 °C, 65 ± 8% RH) environment, whereas both 20TT’s were

completed in the heat (35.1 ± 0.5 °C, 51 ± 4% RH). Neuromuscular assessment of the knee

extensors (5 x 5 s maximum voluntary contractions) was completed pre- and immediately

post-exercise for both 20TT’s and Day 1 and Day 5 of each training block. Blood samples

were also collected at these same time points and analysed for endotoxins, inflammation

and markers of gut damage.

Results: A statistical improvement in 20TTFINAL performance was observed for HA (62 s

[18-104]) but not for CON, compared to 20TTINITIAL baselines. HA training was found to

increase knee extensor strength compared to CON. Despite the faster 20TTFINAL for HA,

no condition difference was found for central fatigue, circulating endotoxin levels,

inflammation, or markers of gut damage.

118 Chapter 5: Study 3

Conclusion: Short-term HA training improves subsequent 20TT cycling performance in

the heat by 2.9% [0.8-4.9] without an associated increase in intestinal damage or

inflammation. These findings suggest that HA training may be a time-efficient training

method to maintain neuromuscular function and improve exercise performance in hot

conditions.

5.2 INTRODUCTION

Exercise performance in hot conditions is compromised by a reduction in avenues for metabolic heat to be lost to the surrounding environment, with subsequent thermal strain evoking premature fatigue (Galloway & Maughan, 1997). While causal mechanism(s) for hyperthermia-induced fatigue remain elusive, considerable efforts have been made to elucidate possible factors which may influence fatigue, such neurobiological changes or psychological aspects (Nybo et al., 2014). Declines in central nervous system (CNS) function resulting in a reduced voluntary drive to exercising skeletal muscle has been linked with elevated core temperatures, though experimental evidence of a direct relationship has been inconclusive

(Nybo & González-Alonso, 2015). A possible intermediary mechanism linking these two factors may be the development of exertional-endotoxemia, as reduced intestinal blood flow during exercise in the heat leads to increased gastrointestinal (GI) damage and the subsequent translocation of endotoxins (lipopolysaccharides; (LPS)) into systemic circulation (Lambert,

2004; Pires et al., 2016). Elevations in circulating levels of LPS prompt a strong immune response by the body and the release of pro-inflammatory cytokines (Lim & Mackinnon, 2006), which have been hypothesised to facilitate an increased sensation of fatigue, and possibly downregulate voluntary activation (Vargas & Marino, 2014).

Chapter 5: Study 3 119

If exertional-endotoxemia, and the associated pro-inflammatory cascade, are causal factors behind the development of central fatigue, interventions that attenuate this response may prove ergogenic. One potential strategy is heat acclimation (HA) training, which provides advantageous physiological and perceptual adaptations that alleviate decrements in exercise capacity (Racinais, Alonso, et al., 2015). Enhanced cardiovascular stability, due in part to an expansion in plasma volume, and lower exercising core temperature, are key physiological improvements arising from HA (Périard et al., 2015). These functional HA adaptations may preserve GI barrier integrity during hyperthermic exercise, as superior cardiac output is thought to better maintain gut blood flow (Lambert, 2004; Sakurada & Hales, 1998), and lower core temperatures may reduce thermal stress to intestinal epithelial cells (Moseley et al., 1994). HA- mediated preservation of the gut barrier has been proposed to attenuate endotoxin translocation and release of pro-inflammatory cytokines (Kuennen et al., 2011; Lambert, 2004). While the exact neurophysiological role that inflammatory cytokines might play in the modulation of central fatigue is poorly understood, several authors have highlighted that cytokine-mediated alterations in neurotransmitter levels could influence CNS drive, and therefore exercise performance (Davis & Bailey, 1997; Vargas & Marino, 2014). We propose that HA training may improve exercise through maintenance of gut integrity and minimise central fatigue during heat-stress.

Few studies have investigated the effects of HA training on endotoxemia, and of these, circulating LPS levels were observed to remain stable from pre- to post-HA (Barberio et al.,

2015; Guy et al., 2016; Kuennen et al., 2011). Further, only Guy et al. (2015) utilised a control group or a self-paced exercise task to determine athletic improvement (~6%). As this enhanced performance following HA was similar in magnitude to that observed in thermoneutral training

(~6%), improvements could have occurred due to task-familiarisation or the training effect, as opposed to HA per se. Similarly, only limited research has investigated the effect of HA on

120 Chapter 5: Study 3

neuromuscular function, with conflicting results suggesting that although voluntary activation is unaffected, skeletal contractile function is either improved (Racinais, Wilson, & Périard,

2017), or unchanged (Brazaitis & Skurvydas, 2010), following repeated passive HA trials. To our knowledge, the only active (exercise-induced) HA study that has explored this area did not undertake post-exercise neuromuscular assessments or utilise a control group, confounding the evaluation of their findings (Wingfield, Gale, Minett, Marino, & Skein, 2016).

No previous research has specifically examined the interplay between HA training, exertional endotoxemia, inflammation and neuromuscular function. Therefore, the purpose of this investigation was to determine the efficacy of short-term HA training on 20 km self-paced cycling performance in the heat, compared to a thermoneutral-matched control group. An additional aim was to identify if adaptations arising from HA would protect against exertional- endotoxemia and inflammation while maintaining neuromuscular function. It was hypothesised that HA would attenuate GI damage, LPS translocation and the release of pro-inflammatory cytokines, reducing central fatigue and resulting in improved cycling performance.

5.3 METHODS

Participants

Eight, recreationally-trained males (age: 27 ± 2 y; body mass: 82 ± 12 kg; height: 181 ± 9 cm;

-1 -1 VO2peak: 49.3 ± 4.9 mL∙kg ∙min ; Pmax: 347 ± 56 W; weekly training sessions: 3.6 ± 1.3; weekly training duration: 191 ± 63 min, performance level 2 - 3; (De Pauw et al., 2013)), volunteered to participate in this study. All participants were non-smokers, free of any injury or illnesses and reported no history of gastrointestinal diseases. Participants were informed of the study procedures and requirements before obtaining verbal and written consent. University

Human Research Ethics Committee approval of this project was attained before the

Chapter 5: Study 3 121

commencement of any testing. Each participant completed 17 separate laboratory visits and received AUD $500 in remuneration.

Experimental overview

Participants were randomly assigned to experimental trials involving a 5-day cycling protocol

-1 (60 min·day at 50% of Pmax) in either hot (heat acclimation; HA: 35 °C; 50% relative humidity

(RH)) or thermoneutral (control; CON: 24 °C; 50% RH) conditions in a counter-balanced, crossover order (Figure 5.1). Environmental conditions were regulated via a climate-controlled chamber (4.7 km∙h-1 wind speed) and monitoring equipment (3M QUESTemp, Quest

Technology, Wisconsin, USA). Familiarisation to all testing procedures, including a practice

20TT and neuromuscular function tests, was completed before commencing data collection.

Figure 5.1. Experimental schematic of a training block. NM: neuromuscular assessment.

The efficacy of each intervention was primarily assessed via a 20 km cycling time trial (20TT) in the heat (35 °C; 50% RH) and associated neuromuscular assessments and biochemical analyses administered before and after the 5-day cycling protocol. Participants abstained from water consumption during all cycling sessions. Participants were also asked to abstain from caffeine and alcohol for 12 h, and strenuous exercise for 24 h, before every performance time trial.

122 Chapter 5: Study 3

Assessments of peak oxygen uptake (VO2peak; TrueOne 2400; ParvoMedics, Salt Lake City,

USA) and maximum aerobic power (Pmax) were conducted before each 5-day training block to inform exercise prescription using a 25 W·min-1 step protocol on a cycle ergometer (Excalibur

Sport; Lode, Groningen, Netherlands).

Heat acclimation intervention

The intervention protocol required participants to complete 5 consecutive days of 60 min·day-

1 cycling at 50% Pmax in controlled environmental conditions of either 34.9 ± 0.7 °C, 53 ± 4%

RH (HA) or 22.2 ± 2.6 °C, 65 ± 8% RH (CON).

The procedures herein were replicated for each testing day. Participants nude weighed and instrumented (rectal thermistor, heart rate monitor, skin thermistors) before dressing in cycling apparel (i.e., cycling bib or exercise shorts without a shirt) and undertaking 60 min of cycling

(Wattbike Pro; Wattbike, Nottingham, UK). Bike set-up was adjusted for each participant and then replicated to ensure consistency across sessions and intervention periods. Resistance and cadence were altered as desired to maintain the required 50% Pmax target power output for the cycle. Participants were free to listen to music or watch a movie if desired to ensure compliance with the cycle task. Perceptual measures of RPE, thermal sensation and thermal comfort were recorded every 10 min during the cycle. Upon completion of the cycle, participants towelled dry, re-weighed and provided a sessional rating of perceived exertion (sRPE; Borg (1998)).

Performance time trials

For each experimental trial, participants presented at the lab at the same time of day (±2 h) and provided baseline venous blood and urine samples before completing a pre-exercise neuromuscular assessment. Participants were then nude weighed and instrumented (i.e., rectal thermistor, heart rate monitor, skin thermistors) before dressing in cycling apparel (i.e., cycling bib or exercise shorts without a shirt). A 20 min baseline period was recorded while participants

Chapter 5: Study 3 123

lay prone in a darkened room (21.2 ± 0.9 °C, 60 ± 4% RH), before undertaking a 20TT

(Velotron Pro, RacerMate, Washington, USA) in 35.1 ± 0.5 °C, 51 ± 4% RH. Participants did not complete a warm-up before the 20TT to reduce variation in initial core temperature. Bike setup (i.e., reach, saddle height, bar height and pedal type) were adjusted to the participants’ preference during familiarisation and were replicated during subsequent experimental sessions.

All 20TTs were undertaken using Velotron 3D software (Version NB04.1.0.2101, RacerMate

Inc., Washington, USA) on a flat course with a constant background scene and without simulated wind resistance or computer opponents. Participants commenced all TTs from a stationary, seated position and were blinded to all feedback, apart from an elapsed distance and the selected gear in gear-inches. Participants completed the 20 km distance as quickly as possible, and no investigator encouragement was provided. We have previously reported recreationally-active athletes (runners, cyclists and team-sport) produce highly reliable 20TT results following the above approach (ICC = 0.80-0.93; (Borg et al., 2018)). Perceptual measures (RPE, thermal sensation and thermal comfort) were recorded every 2 km during the

20TT. Lactate was collected from the left earlobe at before and immediately following the

20TT using a Lactate Scout+ (EKF Diagnostics, Cardiff, Wales). All participants abstained from consuming water until after the 20TT and collection of post-exercise measures (i.e., neuromuscular assessment, bloods, nude weigh, GI form, sRPE). Physical activity, food and fluid intake were diarised for the 24 h before the initial performance trial (20TTINITIAL) and replicated for the subsequent, post-training performance trial (20TTFINAL).

Physiological measures

Core temperature (Tre) was measured using a wireless data logger (T-Tec 7 RF 7-3E) connected to a flexible rectal thermistor (449H; Henleys Medical, Hertfordshire, UK), inserted 12 cm past the anal sphincter. Data were sampled every 30 s during training days and 5 s during 20TT

124 Chapter 5: Study 3

sessions. No participant reached the core temperature termination limit of 40 °C during any session.

Heart rate (HR) was recorded with a heart rate chest strap (Polar Electro Oy, Kempele Finland) and software (Polar Team2, Kempele Finland). Skin temperature was recorded (sample rate: 30 s for training days; 5 s for 20TT sessions) using wireless iButton thermocrons (DS1922L-F50 iButtons, Maxim Integrated, San Jose, USA) taped (3.8 cm width, Leuko Sportstape Premium;

Beiersdorf, Hamburg, Germany) to four different sites: posterior neck, right scapula; posterior left hand; and mid-anterior shin. Mean skin temperature (Tsk) was calculated using a previously published four-site formula (ISO 9886, 2004).

Participant hydration was assessed from the first void of the day and before each trial via a mid-stream urine sample to measure urine colour (scale: 1-8 AU) (Armstrong et al., 2010) and urine specific gravity (PAL-10S; Atagi Ci. Ltd, Tokyo, Japan). Sweat rate and fluid loss were calculated via pre- to post-exercise nude body mass change using calibrated scales (WB-

110AZ; Tanita Corp., Tokyo, Japan).

Perceptual measures

Perceptual measures collected during the training days and performance 20TTs included perceived effort (RPE; Borg (1970)), a 16-point thermal sensation scale (0 = ‘unbearably cold’ to 8 = ‘unbearably hot’) (Young et al., 1987) and a modified 10 point thermal comfort scale (1

= ‘comfortable’ to 5 = ‘extremely uncomfortable’) (Gagge et al., 1967). The psychological wellbeing of participants (i.e., abbreviated profile of mood states; POMS) was collected at the start of each trial (McLean et al., 2010). Participants also completed a GI form to identify the incidence of gut issues (e.g. nausea, cramps, flatulence, urge to defecate or vomit) and severity on a 100 point scale (0 = none; 100 = severe).

Neuromuscular function and voluntary activation (VA)

Chapter 5: Study 3 125

Neuromuscular function (maximal torque, VA, muscle activity and peripheral contractile properties) of the right vastus lateralis (VL) and vastus medialis (VM) were assessed pre- and post-exercise for every 20TT and on the first (Day 1) and final (Day 5) days of each training block (Figure 5.1). Participants were strapped into a Biodex Systems 3 dynamometer (Biodex

Medical Systems, Shirley, New York, USA) with the fulcrum of the lever positioned at the right lateral epicondyle, the chair back adjusted to 95° and securing straps over the chest, waist, thigh and ankle.

To assess VA, the right femoral nerve was stimulated with a single, 100 μs, square-wave pulse from a Digitimer DS7AH stimulator (Digitimer Ltd., Welwyn Garden City, Hertfordshire,

England) delivered via gel electrodes (Pals; Axelgaard Manufacturing Co. Ltd., Fallbrook, CA) positioned on right femoral triangle and gluteal fold. Stimulation amplitude was determined through a twitch ramp protocol. Briefly, femoral nerve stimulations of increasing current were delivered until a reduction in twitch torque was observed. The final current utilised to assess

VA was increased by a further 10% to ensure supramaximal stimulation.

Participants undertook a standardised warm-up of eight isometric knee extensions (90° knee flexion) of increasing intensity, rested for 3 min, then undertook a set of five, 5 s maximal voluntary isometric knee extensions (MVCs), with a 30 s rest between each repetition.

Investigators provided loud verbal encouragement to ensure each contraction was a maximal effort, and participants were also provided visual feedback of their torque production on a screen. Nerve stimulation was manually triggered upon observing a plateau in voluntary torque during each repetition and then again within 2-3s the cessation of muscular contraction. An identical post-exercise procedure (5 x 5 s MVCs) was repeated following every 20TT and the first and final training days of both interventions.

126 Chapter 5: Study 3

All MVC repetitions were re-assessed to ensure a plateau in torque occurred immediately before the interpolated stimulation (Shield & Zhou, 2004). Only repetitions which passed these criteria were included in mean VA calculations. Voluntary activation was calculated using the twitch interpolation formula: VA (%) = (1 – interpolated twitch torque/evoked control twitch torque) * 100 (Allen et al., 1995). The mean of 25 ms preceding the triggered twitch was considered a participant’s maximal voluntary torque, the peak torque in the 100 ms after stimulation was the superimposed torque, and the difference between these two values was considered as the interpolated twitch torque (amplitude). This laboratory has found that familiarised participants demonstrate high reliability with these neuromuscular assessments, seen in the calculated ICCs of 0.94 (VA) and 0.94 (MVC torque).

Evoked twitch contractile properties

Various muscular contractile properties were extracted from resting muscle stimulation that was delivered immediately following each MVC. These included peak twitch torque (Pt; the peak torque during evoked twitch); time to peak torque (TPt; time from the first rise in torque to peak torque); half relaxation time (½ RT; time for torque to reduce by half of the peak torque); contraction duration (CD; TPt + ½ RT); rate of torque development (RTD; slope of twitch-torque curve from onset to peak torque); rate of relaxation (RR; slope of twitch-torque curve peak torque to half relaxation time).

Surface electromyography (EMG)

Surface EMG electrodes (Ambu Blue Sensor N-00-S; Ambu A/S, Ballerup, Denmark) were positioned over the VM and VL, oriented in parallel with the muscle fibres. An earth electrode was attached to the lateral femoral epicondyle on the right leg. All sites were prepared (i.e., shaved, abraded, and cleaned with an alcohol swab) before attaching the electrodes, which remained in place during the exercise bout to ensure consistency within the session.

Chapter 5: Study 3 127

Raw EMG data were recorded using a 16-bit PowerLab 26T AD unit (AD Instruments, Sydney,

Australia) (sample rate=1 kHz; amplification=1000; common mode rejection ratio=110 dB), band-pass filtered (20-500 Hz) and processed with RMS (100 ms window) via LabChart 8.1.5 software (AD Instruments, New South Wales, Australia). Muscle activation was considered a

500 ms period of mean smoothed EMG, preceding a stimulation. This was calculated for each repetition, averaged and normalised to the mean values obtained during pre-exercise MVCs.

Biochemical analysis of blood markers

Blood samples were collected at the same points as the neuromuscular assessment; pre- and post each 20TT, as well as the first and final training intervention days (Day 1 and 5; Figure

5.1). Participants were seated and blood was drawn from an antecubital venipuncture with a sterile butterfly needle (21G, BD, North Ryde, Australia) into 2 x EDTA vacutainer tubes (BD,

North Ryde, Australia). A small (900 μL) amount of whole blood was pipetted into a separate

1.5 mL tube for determination of haemoglobin, haematocrit and blood glucose concentrations.

The EDTA tubes were then centrifuged at 3500 RPM for 15 min at 4˚C, the plasma aliquoted into pyrogen-free 1.5 mL tubes and frozen at -80˚C for a maximum of 6 months. An additional

SST tube was also collected before each 20TT to assess serum osmolality using an Osmomat

030 (Gonotec, Berlin, Germany).

To determine haematocrit, capillary tubes in triplicate were filled with whole blood, centrifuged (12,000 RPM for 10 min at 24 °C) and measured with a digital calliper. Blood glucose and haemoglobin were measured in duplicate from whole blood using an Accu-Chek

Performa (Roche Diagnostics Pty Ltd, West End, Australia) and a HemoCue analyser (Model

201+, Angelholm, Sweden), respectively. Mean values of haematocrit and haemoglobin were then used to calculate changes in plasma volume, as previously described (Dill & Costill,

1974).

128 Chapter 5: Study 3

Blood concentrations of TNF-α (EK-0001; elisakit.com, Melbourne, Australia), IL-6

(HS600B; R&D Systems, Minneapolis, USA) and I-FABP (EHFABP2; Thermo Scientific,

Fredrick, USA) were determined using quantitative sandwich enzyme-linked immunosorbent assays (ELISA). Intra-assay precision was calculated for TNF-α (CV = 8.0%), IL-6 (CV =

8.0%) and I-FABP-α (CV= 7.1%). All samples were diluted to reduce interference (TNF-α,

1:3; IL-6, 1:4; I-FABP, 1:2) and the absorbance read using a SpectroStar Nano (BMG Labtech,

Offenburg, Germany) at the specified wavelength. Wavelength correction for plate imperfections was undertaken when recommended by the manufacturer.

Plasma endotoxin levels were analysed with a kinetic chromogenic Limulus amebocyte lysate assay kit produced by Lonza (Walkersville, USA). Samples were diluted 1:10 with magnesium chloride (MgCl) and heated at 75 °C for 45 min to reduce endotoxin-neutralising components and improve detection sensitivity. Standards were prepared in duplicate at 50 EU∙mL-1 and serially diluted by 1:10 to 0.005 EU∙mL-1. Samples and standards were added to the plates and incubated at 37 °C for 20 min in a SpectroStar Nano (BMG Labtech, Offenburg, Germany).

The required kinetic reagent was then added to every well, and the plates were read at 405 nm, every 61s for 118 cycles (120 min). The unknown endotoxin concentration of samples was calculated from a log/log (time for the sample to increase 0.2 absorbance units/concentration) linear correlation of each standard (r = -0.999), corrected for dilution, as specified by the manufacturer. Sterile, endotoxin-free pipette tips (Biopur epTIPS; Eppendorf AG, Germany) were utilised for all pipetting during blood collection, storage and analysis.

Statistical Analysis

Data were statistically analysed using mixed modelling in a Bayesian framework with the

‘rjags’ package of RStudio Desktop (RStudio v 1.1.447, Boston, MA, USA), as previously described (Mengersen et al., 2016). Exploratory plots of raw data were generated to visually

Chapter 5: Study 3 129

determine the appropriate model and suitability of each model was assessed through a comparison of deviance information criterion values. Certain priors were informed from previous data collected in this laboratory where appropriate (e.g., neuromuscular properties), while novel measures (e.g., performance improvement) utilised vague, uninformed priors

(prior mean = 0; prior precision = 0.001). Markov chain Monte Carlo (MCMC) procedures were used to generate posterior predicted values (1k burn-in, 50k iterations, thinned by 10x resulting in 5k samples). Ergometer data (power, speed, and cadence) over the 20TT was averaged into 0.5 km bins for analysis.

Modelling with random intercepts was employed for 20TT and training day analysis. Main factors (β) for 20TT were test (20TTINITIAL and 20TTFINAL), condition (CON and HA) and either time (pre- and post-20TT) or distance (distance cycled in 20TT). Factors for training days were day (Day 1 and Day 5), condition (CON and HA) and time (pre- and post-cycling or minutes cycled). Modelling also incorporated all factor interactions relevant to the specific model. A factor was considered statistically different when the corresponding 95% credible interval (CI) excluded 0. Further exploration of statistically different factors provided within- or between- group mean differences (MD) with statistical difference again accepted when 95% CI excluded

0. Magnitude-based effect sizes (Cohen’s d) were also calculated for relevant group MD and categorised as small (0.2-0.4), moderate (0.5-0.7) or large (>0.8) (Cohen, 1988). All data is presented herein as mean [95% CI] unless otherwise stated.

5.4 RESULTS

20TT Performance Trials

Pre-trial measures

130 Chapter 5: Study 3

Resting physiological, perceptual and hydration variables collected before the commencement of each 20TT are displayed in Table 5.1. No statistical differences were observed between conditions for relative VO2peak or Pmax, indicating that participants started each training block at a similar level of fitness. Mean washout between conditions (i.e., 20TTFINAL to 20TTINITIAL) was 37 days (range: 25 – 63) and was not found to be associated with a change in 20TT completion time (Spearman’s ρ: 0.07).

Physiological

No condition differences were detected for resting HR, Tre, Tsk or blood glucose at the

20TTINITIAL time-point. Furthermore, no evidence of a statistically different test or condition factor was found for these measures. 20TTFINAL plasma volume was observed to increase from

20TTINITIAL values by 2.2% [-3.0, 7.6] for CON and 7.2% [2.7, 11.7] for HA, with evidence trending towards a condition difference (MD: 5.0% [-1.6, 11.5]; d = 1.48 [0.47-3.41]).

Perceptual

No factors were found to be statistically different for TSensation or TComfort. Similarly, no evidence of statistically different test or condition factors were observed for the majority of POMS categories (i.e., fatigue, soreness, stress and mood). However, the POMS score for sleep in

CON was found to be statistically higher at 20TTFINAL when compared to the 20TTINITIAL time- point in the same condition (MD: 0.6 [0.1, 1.2]). No other statistical differences were observed for test or condition.

Hydration

Pre-trial body mass, first void USG, arrival USG and serum osmolality were consistent across all conditions and tests, with no statistically different factors revealed (Table 5.1). Thus, all participants were considered to have started each 20TT in a similar state of hydration.

20TT Performance

Chapter 5: Study 3 131

Statistical differences in both test and condition factors were found for 20TT completion time

(Figure 5.2; Table 5.2). Further analysis revealed that HA training resulted in a statistically faster 20TTFINAL completion time, compared to HA 20TTINITIAL (MD: 62 sec [18, 104], 2.9% improvement, d = 2.86 [0.82, 4.75]) and CON 20TTFINAL (55 sec [11, 98], d = 2.51 [0.49,

4.46]). In contrast to the improvement seen in HA, CON did not observe a statistically faster completion time, with a mean difference of 30 sec ([-6, 67], 1.4% improvement). No difference was observed between conditions for 20TTINITIAL completion time, suggesting adequate washout between trials and no training effect.

Statistically different test and condition factors were found for mean 20TT power and speed but not cadence (Table 5.2). Specifically, HA 20TTFINAL power (228 W [189, 273]) and speed

-1 (34.3 km∙h [31.7, 36.9] were statistically higher than HA 20TTINITIAL (d = 2.79-2.9) or CON

20TTFINAL (d = 2.41-2.49) values. Like completion time, no statistical differences in power or speed were observed between CON 20TTINITIAL and 20TTFINAL.

132 Chapter 5: Study 3

20TTFINAL Performance Improvement

10

8

6

4

HA Mean

Improvement (%) 2 CON Mean FINAL 0 20TT -2 CON

HA -4

Figure 5.2. Individual 20TTFINAL completion time improvement for CON and HA.

Power output relative to 20TTINITIAL

30

20

10

R elative pow er (% ) 0

CON 20TTFINAL

HA 20TTFINAL -10 0 2 4 6 8 10 12 14 16 18 20 Distance (km)

Figure 5.3. Mean power output of 20TTFINAL (relative to 20TTINITIAL values) for CON and HA.

Chapter 5: Study 3 133

20TT Physiological

A main effect of condition and distance was found for HR (βcondition: 4.7 [2.1, 7.4]) and Tre

(βcondition: 0.15 [0.07, 0.23]), with both variables observed to be statistically higher in HA than

CON for both 20TTINITIAL and 20TTFINAL tests (Table 5.2). No difference in test factor was found for HR or Tre. Statistically different factors for distance, test and condition were revealed for Tsk, with a trend for lower values in CON for both tests (20TTINITIAL MD: -0.1 °C;

20TTFINAL MD: -0.1 °C). Both conditions saw a decrease in skin temperature during 20TTFINAL compared to the initial time trial (MD: -0.1 to -0.2 °C [-0.2 to 0.0]).

20TT Perceptual

No factor was found to be statistically different for RPE, thermal sensation or comfort.

Post-trial measures

A test effect was revealed for post-exercise blood lactate (βTEST: 2.8 [0.5, 4.9]), which was found to be statistically higher following 20TTFINAL, compared to 20TTINITIAL (Table 5.2). This test difference in post-exercise blood lactate was seen in both conditions. No evidence of a statistical difference for any factor was found for body fluid loss (absolute or relative) or sRPE.

Incidence and severity of GI symptoms were not found to be statistically different for any factor

(Table 5.3).

134 Chapter 5: Study 3

200

190 A

180 )

-1 170 min ˜ ˜ 160

150 HR (beats

140 CON 20TTINITIAL

HA 20TTINITIAL

130 CON 20TTFINAL

HA 20TTFINAL 120 0 2 4 6 8 101214161820

Distance (km)

39.0 B 38.5 C) q

38.0

37.5 oetmeaue( Core temperature CON 20TTINITIAL 37.0 HA 20TTINITIAL

CON 20TTFINAL

HA 20TTFINAL 36.5 02468101214161820

Distance (km)

37 C

36 C) q

35 Skin ( temperature

34 CON 20TTINITIAL

HA 20TTINITIAL

CON 20TTFINAL

HA 20TTFINAL 33 0 2 4 6 8 10 12 14 16 18 20

Distance (km)

Figure 5.4. Mean and 95% CI for (A) HR, (B) core and (C) skin temperature during CON and HA 20TT tests.

Chapter 5: Study 3 135

MVC and VA

Analysis revealed statistically different time, test, condition and interaction factors for MVC torque (Table 5.4). Post-exercise MVC torque was found to be lower than pre-exercise values for all conditions (MD: 19 – 36 N∙m; d = 3.28 – 6.64; Figure 5.5). HA saw an increase to

20TTFINAL pre-exercise values (MD: 11 N∙m [2, 21; d = 2.22 [0.28, 4.21]), whereas CON

20TTFINAL was found to reduce compared to 20TTINITIAL pre-exercise torque (MD: -13 N∙m [-

24, -3]; d = 2.46 [0.49, 4.49]). Evidence of a condition effect was observed for 20TTFINAL measures, with a statistically higher MVC torque in HA for both pre- (MD: 17 N∙m [7, 27]; d

= 3.29 [1.29, 5.23]) and post-cycling (MD: 16 N∙m [5, 27]; d = 2.75 [0.83, 4.68]) time-points.

A time and test x time interaction factor were found to be statistically different for VA. All conditions and tests saw a reduction in VA from pre- to post-exercise (MD: -3.4 to -6.5%; d =

3.16 – 6.04). Further analysis of the interaction factor revealed that post-exercise VA for CON was statistically higher after the 20TTFINAL trial when compared to 20TTINITIAL (MD: 3.3%

[1.1, 5.5]; d = 2.88 [0.95 – 4.83]). No evidence of a condition factor was found for VA.

Contractile properties and EMG

Analysis of peak twitch torque found statistical differences for test and time factors. Further analysis revealed that this variable was reduced from 20TTINITIAL to 20TTFINAL at the pre- cycling time-point for both conditions. Furthermore, all conditions and tests saw a reduction in

Pt from pre- to post-cycling (MD: -15.6 to -20.3 N∙m, d = 7.40 - 10.47). A time factor was found for TPt, ½RT and RTD, with faster post-exercise measures for all conditions and tests.

A test factor was also found for the first two variables, which revealed a slower TPt but a faster

½RT at the 20TTFINAL HA pre-cycling time-point verse 20TTINITIAL (Table 5.4). A faster pre- cycling rate of relaxation was revealed for both conditions at 20TTFINAL, when compared to

-1 20TTINITIAL values (MD: 47-55 N·m·s ; d = 1.29 – 1.92). No statistically different condition

136 Chapter 5: Study 3

factor was found for any contractile variable. No test, condition or test x condition factor was seen for either EMG variable.

Chapter 5: Study 3 137

300 20TTINITIAL Day 1 Day 5 20TTFINAL * #

275 # * † # † m)

˜ † † † † 250 † †

225 MVC (N Torque

200

CON

HA 175 Pre Post Pre Post Pre Post Pre Post

Figure 5.5. Mean and 95%CI for MVC torque throughout each training block. * indicates HA is statistically different to CON at same time-point. † indicates statistically different pre- to post-exercise torque in the same condition; # indicates torque is statistically different from same time-point and condition in 20TTINITIAL.

138 Chapter 5: Study 3

Biochemical analysis of blood markers

TNF-α was found to have a statistically different time factor, resulting in a large increase pre-

-1 to post-cycling during CON 20TTFINAL (MD: 2.2 pg·mL [0.1, 4.2]; d = 2.08 [0.1, 4.05]; Table

5.5). No other time differences or factors were detected. Analysis of I-FABP, a marker of gut damage, was found to have statistically different test and time factors (Figure 5.6). Between- test differences were observed for both HA timepoints and CON post-exercise (d = 2.12 –

2.26), but not for the CON pre-exercise timepoint (MD: 0.023 ng·mL-1 [-0.411, 0.458]). A similar outcome was found for time, with the 20TT resulting in an increased post-exercise I-

FABP compared to pre-exercise values for both HA tests, and CON 20TTFINAL (MD: 0.518 –

0.665 ng·mL-1; d = 2.50 – 3.04]). No condition differences were observed. Only post-exercise

IL-6 concentrations were measured, so the factor of time was excluded from modelling.

However, the analysis revealed some evidence of a statistically different test factor, with a lower IL-6 concentration detected post-exercise of HA 20TTFINAL, when compared to the same time-point in the previous test (MD: 2.4 pg·mL-1 [-0.3, 5.3]; d = 1.7 [-0.2, 3.7]). A similar trend was seen for condition, with evidence supporting a probable difference between 20TTINITIAL post-exercise values (MD: 2.7 pg·mL-1 [-0.2, 5.6]; d = 1.88 [-0.12, 3.88]), although this was not observed at 20TTFINAL. No main factors were found to be statistically different for LPS.

Chapter 5: Study 3 139

Training Days (Day 1 and 5)

Pre-trial measures

Statistical analysis of pre-exercise thermal sensation, comfort, body mass and POMS found no effect of day, condition or day x condition interaction factors. First-void and arrival hydration measures were also found to be not statistically different between days and conditions, confirming that participants began each training session in a similar state of hydration (Table

5.6).

Performance

A statistical difference between conditions was observed on Day 5, with a mean difference of

4 W [4-6] (CON: 175 W [154, 196]; HA: 171 W [150, 191]). Day and condition effects were also observed for cadence, which revealed a slightly lower RPM in HA compared to CON regardless of day (MD: 1-2 RPM, [1-2]), as well as a small reduction in mean cadence from

Day 1 to 5 for both conditions (MD: 2 RPM [2-3]).

Physiological

Due to the difference in environmental temperature, mean HA was statistically higher than

CON for HR, Tre and Tsk, regardless of day. Adaptation to the task was observed for both conditions, with a significant reduction in mean HR for Day 5 compared to Day 1 (CON MD:

-3 beats∙min-1 [-4 to -2]; HA MD: -10 beats∙min-1 [-11 to -8]). In contrast, the change in mean

Tre between days was only found to be lower in HA (MD: -0.2 °C [-0.2 to -0.1]), whereas CON saw a slight increase in this measure over the course of the training block (MD: 0.1 °C [0.0 to

0.1]). Mean skin temperature followed a similar pattern to Tre, with a reduction seen by Day 5 of HA (MD: -0.1 °C [-0.2 to -0.0]) and a rise in CON (MD: 0.2 °C [0.2 to 0.3]).

140 Chapter 5: Study 3

Perceptual

A condition difference for RPE, thermal sensation and thermal comfort was observed for both days, with HA recording slightly higher mean values for all measures (MD: 1 [0-1]; 0.7-0.9

[0.5-1.2]; and 0.8-0.9 [0.6-1.1], respectively). Thermal comfort was also found to reduce across the training block, with a lower mean Day 5 value for both conditions (MD 0.2-03 [0-0.5]). No effect for day was found for either perceptual variable.

Post-trial measures

Body fluid loss (absolute and relative) and sweat rate was found to have a strong condition effect, with HA reported a statistically higher fluid floss and sweat rate (MD: 0.3 – 0.4 kg; 0.3

– 0.4%; 0.40 – 0.47 L·hour-1) than seen in CON, for both training days (d = 3.43 – 9.11).

However, no training day difference was observed for any of these variables (βDAY: -0.2 to 0.2;

-0.2 to 0.2; and -0.04 to 0.16, respectively). No statistically different factors were found for session RPE and post-exercise blood lactate (Table 5.7).

Neuromuscular properties

Analysis of MVC torque, VA and peak twitch found a statistically different time factor for both conditions (βTIME: -29 [-44 to -14]; -3.9 [-6.2 to -1.6]; -19.9 [-13.1 to 26.6]), with a reduction in all variables from pre-exercise values for both days (Table 5.8). Evidence supported a similar time factor for TPt and CD, with shorter time for both variables post-exercise for all conditions and days (d = 2.37–3.61), except TPt CON Day 5 (MD: 7 ms [-1 to -15]). No evidence of a time factor was found for ½RT, RTD, RR. Furthermore, no statistical difference in condition or day factors were found for any neuromuscular variable, including EMG measures of VM and VL.

Chapter 5: Study 3 141

Biochemical analysis of blood markers

No clear main factor was found for TNF-α; however, there was some evidence to support a statistically different interaction factor (βCONDITION*TIME: 1.8 [-0.2 to 3.7]). Further analysis revealed an increase in post-exercise concentrations of this cytokine for Day 1 of HA (MD:

2.9 pg∙mL-1 [1.3-4.3]; d = 3.77 [1.77-5.70]). Statistical analysis of I-FABP provided evidence supporting a similar condition x time interaction, specifically an increase following exercise on Day 1 of HA (MD: 0.459 ng∙mL-1 [0.109 – 0.821]; d = 2.52 [0.60-4.51]; Figure 5.6). No other statistical difference was observed for any other day or condition.

142 Chapter 5: Study 3

2.0 20TTINITIAL Day1 Day5 20TTFINAL

†# †#

1.5 †

) † -1 # mL ˜

1.0 I-FABP (ng 0.5

HA CON 0.0 Pre Post Pre Post Pre Post Pre Post

Figure 5.6. Mean and 95%CI for I-FABP for each 20TT and training days. † indicates statistically different pre- to post-exercise torque in the same condition; # indicates torque is statistically different from same time-point and condition in 20TTINITIAL.

Chapter 5: Study 3 143

5.5 DISCUSSION

To our knowledge, this study is the first to investigate the effect of short-duration HA on self-paced cycling performance, inflammation and neuromuscular function using a randomised, controlled, and cross-over design. The main findings were: 1) 5 days of moderate-intensity HA training (five, consecutive 60 min sessions at 50% Pmax in 35

°C) improves 20km self-paced cycling performance in the heat by 2.9% [0.8-4.9] without inducing additional inflammatory stress or exertional endotoxemia; 2) HA training increases knee extensor strength; and 3) 5 days of matched training in CON does not enhance 20TT performance in the heat and results in knee extensor strength impairments. Accordingly, these data suggest short-duration HA training benefits adaptations that result in improved neuromuscular function and cycling performance in the heat.

While the performance benefit arising from HA training is well accepted (Chalmers et al., 2014; Garrett et al., 2011; Périard et al., 2015), considerable methodological variations (e.g., exercise type, training duration and intensity, and environmental conditions) confounds interpretation of aggregate results. In the context of short- duration (<7 days) HA training, few studies have utilised an externally valid measure of self-paced exercise performance in the heat, with mixed results. Racinais, Périard, et al. (2015) reported that highly-trained cyclists improved cycling performance by

10.4% following 5 days of HA training. This 3.5-fold increase over the current study

(10.4% compared to 2.9%) may be due to the longer duration of the performance task

(43.4 km vs 20 km) and/or greater training duration (>240 min∙day-1 vs 60 min∙day-1).

A comparable improvement (9.6%) in work completed in 30 min was also seen following 10 days of 90 min heat-training (Keiser et al., 2015). However, participants

144 Chapter 5: Study 3

were immersed in a hot water bath for 20 min before commencing the time trial, which makes it difficult to draw parallels to the outcomes of the present study.

In contrast to medium and long-term HA training, the reduced time requirements of short-duration HA are appealing to many athletes and coaches. Crucially, this limited time course has been shown to still facilitate considerable physiological and perceptual adaptations to subsequent heat stress, such as plasma volume expansion (Périard et al.,

2015). In the present study, the observed increase in plasma volume for HA (7.2%

[2.7, 11.7]) aligns with previously reported values (4.2-9.8%) (Garrett et al., 2011) and this enhanced cardiovascular stability may potentially explain the observed improvement in cycling performance, as has been previously discussed (Périard et al.,

2015). However, the consistency in other measures of heat adaptation (i.e., resting HR,

Tre and Tsk) between 20TTs and conditions, suggests that 5 days of HA training may be insufficient to finalise these thermo-physiological changes. This evidence supports the findings of similar short-duration HA protocols, which did not observe a change in these physiological variables (Garrett et al., 2012; Willmott et al., 2017).

We hypothesised that adaptations arising from HA training, namely reductions in core temperature and maintenance of intestinal blood flow, could attenuate the development of central fatigue and result in enhanced exercise performance (Kuennen et al., 2011;

Lambert, 2004; Nybo & Nielsen, 2001a). Our data showed no difference in VA levels pre- or post-exercise between HA tests or when compared to CON. Thus, it could be argued that the improved HA exercise performance combined with the similar VA levels (Table 5.4), may indicate that HA does enhance neural recruitment of the skeletal muscle, as more work was completed without a detriment in VA. As far as we are aware, only three other HA studies have also examined neuromuscular responses, although methodological differences, such as the use of passive heating (hot water

Chapter 5: Study 3 145

immersion, (Brazaitis & Skurvydas, 2010); hot environmental temperature, (Racinais et al., 2017), or lack of a control arm and post-exercise measures (Wingfield et al.,

2016), does limit comparison. Regardless, only Racinais et al. (2017) reported HA- induced increases in torque for a given VA, while the other two studies observed no difference (Brazaitis & Skurvydas, 2010), or a reduction (Wingfield et al., 2016) in pre- to post-HA torque. Importantly, we observed a significant moderate association between certain inflammatory cytokines (i.e., TNF-α) and central fatigue (i.e., VA; ρ

= 0.43, P = 0.005), and speculate there could be a potential relationship between these two factors.

A novel discovery of the current study was the alteration in neuromuscular strength throughout the training block, with an increase in knee extensor strength for HA, while

CON reported impairment in this measure (Table 5.4 and Table 5.8). Notably, while the 20TT resulted in a similar reduction in MVC torque for both conditions (MD: 16-

17 N∙m [5-27]; d = 2.75-3.29), this separation in strength remained post-exercise. As pre-exercise VA was similar across conditions, suggesting a comparable level of descending drive, this strength disparity is assumed to arise from changes in peripheral fatigue (Périard, Caillaud, et al., 2011). However, both conditions reported a reduction in peak twitch torque pre-20TTFINAL compared to 20TTINITIAL (MD: 4 N∙m [0-8]; d =

1.39-1.59), indicating a similar level of cumulative fatigue from the 5-day training block, as has been previously identified (Wingfield et al., 2016). This fatigue may have stemmed from the increased training load associated with consecutive days of cycling, particularly as participants were only recreationally-active and arguably unaccustomed to this level of training. However, while this potentially explains the reduction in CON

MVC torque, the development of peripheral fatigue combined with stable VA level and EMG activity, confounds the observed HA strength gain.

146 Chapter 5: Study 3

Although speculative, one possible reason for the strength increase following HA may be due to heat-induced hypertrophy of the leg muscles involved in cycling (e.g., VM,

VL, gluteus maximus, gastrocnemius, biceps femoris). This aligns with previous research which has shown increased muscle cross-sectional area and torque in human participants following passive heat stress (Goto et al., 2011; Racinais et al., 2017), possibly arising from the release of heat shock protein 72. Further, 5 days of cycling may have enhanced neural recruitment patterns, specifically a reduction in co- activation of antagonist muscles (Enoka, 1997), potentially explaining the increased knee extensor torque without associated changes in VA or EMG activity of VM and

VL. Alternatively, as participants were unable to be blinded to the specific intervention, a psychological bias may have also contributed to the observed differences in 20TTFINAL pre-cycling torque. Participants who had completed HA training may have been more motivated in the pre-20TT neuromuscular testing due to perceived preparedness to complete the subsequent cycling task when compared to

CON. However, this was not reflected in pre-20TT POMS or VA level between conditions (Table 5.1 and Table 5.4).

Training in the heat was initially found to increase gut damage (i.e., I-FABP) and circulating inflammatory cytokines (i.e., TNF-α), probably due to the relative physiological strain experienced during HA as opposed to CON (Table 5.9). However, no change in these analytes was observed post-exercise on Day 5 in either condition.

This suggests that the increased thermotolerance arising from physiological HA adaptations, such as greater plasma volume, may have maintained perfusion of intestinal tissue during hyperthermic exercise, and therefore attenuated gut damage and inflammation (Lambert, 2004). A similar outcome for TNF-α following short- duration HA has been previously reported by both Barberio et al. (2015) and Hailes,

Chapter 5: Study 3 147

Slivka, Cuddy, and Ruby (2011). Thus, in the context of the current evidence and previous literature (Barberio et al., 2015; Selkirk et al., 2008), we propose that 5 days of moderate intensity HA training appears to blunt markers of inflammation and gut damage associated with exercise in the heat.

An interesting finding from the present study was that the faster completion time of

HA 20TTFINAL did not result in greater levels of circulating LPS, I-FABP, IL-6 or TNF-

α post-exercise when compared to CON (Table 5.5). However, as post-exercise gut damage (i.e., I-FABP) increased for both conditions compared to 20TTINITIAL, it appears that the shorter-duration but higher-intensity of the self-paced 20TT outstrips the previously discussed beneficial adaptations seen during the final day of HA training. Increased gut damage has been linked to the translocation of LPS into systemic circulation, resulting in endotoxin-mediated cytokinemia and development of exertional heat illness (Dokladny et al., 2016; Lim & Mackinnon, 2006). However, the stable level of endotoxins in the present study suggest that elevated gut damage following ~35 min of vigorous exercise is insufficient to induce LPS translocation, which is in agreement with most (Guy et al., 2016; Kuennen et al., 2011), but not all

(Barberio et al., 2015), previous HA research. A similar pattern was also seen in the measured cytokines, TNF-α and IL-6, which were unchanged and decreased, respectively, compared to 20TTINITIAL (Table 5.5). Taken together, the outcomes from this study provide evidence that 5 days of HA training protect against inflammation and do not induce endotoxemia during subsequent self-paced exercise in the heat.

Limitations of the present study included the use of recreationally-trained participants, which could have resulted in a larger magnitude of improvement compared to a more- highly trained cohort. However, our robust study design (repeated measures with the work-matched control group) supports that thermoneutral exercise does not provide

148 Chapter 5: Study 3

enough stimulus to improve 20TT performance. Furthermore, similar duration HA training has been found to still be ergogenic for highly-trained cyclists (Racinais,

Périard, et al., 2015). While important inflammatory biomarkers were measured during pre- and post-cycling, additional time-points and would have provided more information regarding the expression and clearance of these markers. The self-paced nature of the 20TT task, with associated fluctuations in exercise intensity, also made it challenging to identify the effect of HA adaptations of measured variables. Utilising a fixed-load task before the 20TT may have provided additional information around these changes. Although the training sessions were undertaken at a fixed intensity of

50% Pmax, participants struggled to maintain this workload on the final day of HA, with a between-condition difference for Day 5 of 4 W [4-6W]. However, this power fluctuation was only equal to 1.1% difference of the overall mean Pmax (Table 5.1) and both conditions remained within <0.7% of the 50% Pmax target power (CON: 50.6%

[49.5, 51.4]; HA: 49.5% [48.2, 50.3]). Therefore, it is unlikely that this mean power difference on the final training day would have altered the subsequent 20TTFINAL performance.

5.6 CONCLUSION

In summary, this is the first study to date which has examined short-duration HA training and its effect on self-paced exercise, neuromuscular fatigue and inflammatory biomarkers. Our data provide evidence that 5 days of moderate- intensity HA training favourably improves 20TT cycling performance and knee extensor strength, without conferring traditional HA adaptations observed in longer protocols. Importantly, the greater work-load achieved following HA training did not induce increased central fatigue, inflammation or the translocation of endotoxins. In

Chapter 5: Study 3 149

conclusion, our data support the ergogenic benefit of short-duration HA training, while remaining a feasible intervention for time-restricted athletes.

5.7 ACKNOWLEDGMENTS

The authors sincerely thank Mr Logan Trim (Institute of Health and Biomedical

Innovation, Queensland University of Technology, Brisbane, Queensland, Australia) for his technical assistance with the immunoassay analysis.

150 Chapter 5: Study 3

5.8 TABLES

Table 5.1. Pre-trial data for initial and final 20TT (mean [95% credible interval]). CONTROL HEAT ACCLIMATION Variable 20TTFINAL 20TTINITIAL 20TTFINAL 20TTINITIAL

Physiological

HR (beats∙min-1) 55 [49, 60] 50 [45, 55] 52 [47, 58] 49 [44, 55] 37.0 [36.9, 37.2] 37.1 [36.9, 37.3] 37.2 [37.0, 37.4] 37.1 [37.1, 37.3] Tre (°C) 33.1 [32.8, 33.3] 33.0 [32.8, 33.2] 33.1 [32.9, 33.3] 33.0 [32.8, 33.3] Tsk (°C)

Δ HR (beats∙min-1) - 4 [-1, 9] - 3 [-2, 8] - -0.03 [-0.21, 0.15] - 0.05 [-0.13, 0.22] Δ Tre (°C) - 0.05 [-0.08, 0.19] - 0.08 [-0.06, 0.21] Δ Tsk (°C)

Δ plasma volume (%) - 2.2 [-3.0, 7.6] - 7.2 [2.7, 11.7]*

-1 -1 49.2 [45.0, 53.5] - 50.4 [46.3, 54.6] - Relative VO2peak (mL∙kg ∙min )

Maximal aerobic power (W) 345 [311, 380] - 346 [312, 381] -

Pre-20TT blood glucose (mmol∙L-1) 4.8 [4.2, 5.4] 4.8 [4.1, 5.4] 4.5 [3.8, 5.1] 4.2 [3.6, 4.9]

Chapter 5: Study 3 151

Perceptual 3.3 [2.8, 3.7] 3.2 [2.7, 3.7] 3.2 [2.7, 3.7] 3.1 [2.6, 3.6] TSensation (AU) 1.1 [0.8, 1.4] 1.3 [1.0, 1.6] 1.1 [0.8, 1.4] 1.3 [1.0, 1.6] TComfort (AU)

POMS - Fatigue 3 [2, 3] 3 [3, 4] 3 [3, 4] 3 [3, 4] # POMS - Sleep 3 [3, 4] 4 [3, 5] 3 [3, 4] 4 [3, 4]

POMS - Soreness 3 [2, 4] 3 [3, 4] 3 [2, 4] 3 [3, 4]

POMS - Stress 3 [3, 4] 3 [3, 4] 3 [3, 4] 3 [3, 4]

POMS - Mood 4 [3, 4] 4 [3, 4] 4 [3, 4] 3 [3, 4]

Hydration

Pre-20TT body mass (kg) 79.0 [68.2, 88.0] 78.8 [67.9, 87.8] 78.0 [67.1, 87.1] 78.5 [67.6, 87.6]

USG – first void 1.021 [1.005, 1.033] 1.021 [1.004, 1.037] 1.015 [1.004, 1.036] 1.019 [1.002, 1.042]

USG – arrival 1.013 [1.000, 1.030] 1.008 [1.000, 1.032] 1.012 [1.000, 1.029] 1.008 [1.002, 1.032]

Serum Osmolality - arrival 0.294 [0.287, 0.301] 0.294 [0.285, 0.304] 0.294 [0.284, 0.304] 0.294 [0.280, 0.307] * indicates a statistical condition difference at the same time-point and test; # indicates a statistical test difference at the same condition (i.e., HA

20TTINITIAL to HA 20TTFINAL)

152 Chapter 5: Study 3

Table 5.2. Performance data for initial and final 20TT (mean [95% credible interval]). CONTROL HEAT ACCLIMATION Variable 20TTFINAL 20TTINITIAL 20TTFINAL 20TTINITIAL

Performance # Completion Time (min:sec) 36:07 [33:27, 38:18] 35:37 [32:50, 37:50] 35:44 [33:03, 37:55] 34:42 [31:53, 36:54]*

Δ completion time (sec) - 30 [-6, 67] - 62 [18, 104] # Power (W) 205 [165, 251] 213 [173, 258] 211 [171, 256] 228 [189, 273] ]* # Speed (km∙h-1) 32.8 [30.1, 35.4] 33.3 [30.7, 36.0] 33.1 [30.4, 35.8] 34.3 [31.7, 36.9]*

Cadence (RPM) 93 [86,100] 92 [85, 100] 94 [87, 100] 92 [85, 100]

Physiological

Mean HR (beats∙min-1) 158 [155, 161] 157 [155, 166] 162 [159, 170]* 161 [159, 169]* 37.7 [37.7, 37.9] 37.8 [37.8, 38.0] 37.7 [37.6, 37.9]* 37.8 [37.7, 38.0]* Mean Tre (°C) 35.9 [35.8, 36.1] 35.7 [35.6, 35.9]# 35.9 [35.6, 35.9] 35.8 [35.8, 36.1]*# Mean Tsk (°C)

Body fluid loss (kg) 1.0 [0.8, 1.2] 1.1 [0.9, 1.3] 1.0 [0.9, 1.2] 1.1 [0.9, 1.3]

Relative body fluid loss (%) 1.2 [1.0, 1.4] 1.3 [1.1, 1.5] 1.3 [1.1, 1.5] 1.4 [1.2, 1.6]

Chapter 5: Study 3 153

Post-20TT blood lactate (mmol∙L- 7.7 [4.3, 11.6] 11.2 [7.8, 15.1]# 7.4 [4.0, 10.5] 9.7 [6.3, 12.8]# 1)

Perceptual 15.7 [14.6, 16.9] 15.9 [14.8, 17.1] 15.7 [14.6, 16.9] 15.6 [14.5, 16.8] Mean RPE (AU)

5.9 [5.4, 6.4] 5.9 [5.4, 6.3] 6.0 [5.6, 6.5] 5.9 [5.4, 6.4] Mean TSensation (AU) 3.2 [2.6, 3.7] 3.0 [2.4, 3.6] 3.1 [2.5, 3.7] 3.0 [2.5, 3.6] Mean TComfort (AU)

Session RPE (AU) 7.5 [5.8, 9.1] 8.4 [6.7, 10] 7.5 [5.8, 9.1] 8.0 [6.4, 9.7] * indicates a statistical condition difference at the same time-point and test; # indicates a statistical test difference at the same condition (i.e., HA

20TTINITIAL to HA 20TTFINAL)

154 Chapter 5: Study 3

Table 5.3. GI symptom questionnaire data for initial and final 20TT.

CONTROL HEAT ACCLIMATION Variable 20TTFINAL 20TTINITIAL 20TTFINAL 20TTINITIAL

GI Symptoms

Cramp 1; 25 (N/A) 0 0 0

Nausea 2; 50 (25 – 75) 2; 25 (25 – 25) 4; 39 (25 – 55) 2; 25 (25 – 25)

Urge to defecate 0; 0 2; 33 (25 – 40) 0

Urge to vomit 3; 25 (25 – 25) 1; 25 (N/A) 2; 48 (45 – 50) 1; 25 (N/A)

Stitch/pain in gut 0; 0 0 0

Flatulence 0; 1; 50 (N/A) 0 1; 50 (N/A) Data displayed as: incidence; mean severity of symptoms (range of severity). Severity graded as 0 = none; 25 = mild; 50 = moderate; 75 = moderate/severe; 100 = severe.

Chapter 5: Study 3 155

Table 5.4. Neuromuscular properties for initial and final 20TT (mean [95% credible interval]).

20TTFINAL 20TTINITIAL Variable Condition Pre-Exercise Post-Exercise Pre-Exercise Post-Exercise

† MVC Torque CON 251 [218, 282] 216 [182, 247]† 238 [205, 269]# 220 [186, 251] (N∙m) #† HA 244 [211, 275] 214 [180, 246]† 256 [222, 287]*# 236 [202, 266]* †# VA (%) CON 95.9 [94.4, 97.4] 89.4 [87.7, 91.0]† 96.0 [94.5, 97.6] 92.6 [91.0, 94.3] † HA 95.2 [93.7, 96.6] 90.7 [89.0, 92.3]† 96.1 [94.7, 97.5] 92.2 [90.7, 93.8]

EMG VL (%) CON - 56.8 [38.9, 738.8] - 63.0 [42.8, 82.1]

HA - 67.3 [47.1, 86.5] - 68.1 [47.1, 88.7]

EMG VM (%) CON - 66.3 [47.0, 82.7] - 74.2 [54.2, 91.3]

HA - 62.0 [41.7, 80.3] - 80.2 [59.5, 98.6] † Pt (N) CON 68.1 [59.7, 76.0] 47.8 [39.3, 56.2]† 64.1 [55.5, 72.3]# 48.5 [39.9, 56.6] † HA 69.1 [60.7, 77.2] 49.5 [41.3, 57.7]† 65.3 [56.8, 73.3]# 46.6 [38.4, 54.8] † TPt (ms) CON 80 [75, 84] 69 [64, 73]† 76 [71, 81] 67 [63, 72] † HA 83 [79, 88] 67 [62, 71]† 76 [72, 81]# 65 [60, 69]

156 Chapter 5: Study 3

† 1/2 RT (ms) CON 70 [64, 85] 46 [35, 57]† 74 [64, 85] 48 [36, 58] † HA 65 [54, 76] 47 [36, 57]† 71 [60, 81]# 47 [36, 57] † CD (ms) CON 149 [136, 162] 115 [102, 128]† 151 [138, 164] 115 [102, 128] † HA 148 [136, 161] 113 [100, 126]† 147 [134, 159] 112 [99, 124] -1 † RTD (N∙s ) CON 815 [763, 867] 738 [674, 801]† 828 [767, 888] 719 [647, 794] † HA 822 [759, 882] 736 [665, 807]† 840 [771, 907] 717 [640, 795] -1 RR (N∙s ) CON 552 [502, 603] 570 [512, 630] 505 [445, 565]# 536 [470, 603]

HA 558 [501, 618] 575 [513, 641] 503 [438, 567]# 534 [468, 601] * indicates a statistical condition difference at the same time-point and test; † indicates a statistical time difference at same test and condition # (i.e., pre- to post-exercise); indicates a statistical test difference at the same time-point and condition (i.e., 20TTINITIAL post-exercise to

20TTFINAL post exercise)

Chapter 5: Study 3 157

Table 5.5. Plasma concentrations of inflammatory markers for initial and final 20TT (mean [95% credible interval]).

20TTFINAL 20TTINITIAL Variable Condition Pre-Exercise Post-Exercise Pre-Exercise Post-Exercise

-1 IL-6 (pg∙mL ) CON - 3.6 [1.4, 5.8] - 2.2 [0.1, 4.5] # HA - 6.3 [3.9, 8.7]* - 3.9 [1.8, 6.1] -1 # TNF-α (pg∙mL ) CON 3.2 [1.6, 4.7] 4.1 [2.5, 5.8] 3.3 [1.5, 5.1] 5.5 [3.8, 7.1]

HA 3.5 [1.9, 5.1] 4.3 [2.5, 6.0] 3.6 [1.9, 5.2] 4.8 [3.1, 6.4] -1 # I-FABP (ng∙mL ) CON 0.733 [0.385, 1.057] 0.932 [0.595, 1.256] 0.711 [0.353, 1.067] 1.376 [1.061, 1.68] † # HA 0.449 [0.131, 0.762] 0.973 [0.635, 1.311]† 0.902 [0.572, 1.215]# 1.420 [1.103, 1.738] † -1 LPS (EU∙mL ) CON 0.212 [0.048, 0.555] 0.388 [0.062, 0.874] 0.188 [0.083, 0.546] 0.552 [0.171, 0.811]

HA 0.216 [0.023, 0.648] 0.427 [0.060, 0.785] 0.245 [0.033, 0.579] 0.650 [0.082, 0.989] * indicates a statistical condition difference at the same time-point and test; † indicates a statistical time difference at same test and condition # (i.e., pre- to post-exercise); indicates a statistical test difference at the same time-point and condition (i.e., 20TTINITIAL post-exercise to

20TTFINAL post exercise)

158 Chapter 5: Study 3

Table 5.6. Baseline data for Day 1 and Day 5 of training (mean [95% credible interval]).

Training Day ∆ Day 1 to 5 Variable Condition Day 1 Day 5

Hydration

USG – first void CON 1.019 [1.005, 1.030] 1.023 [1.004, 1.033] 0.004 [-0.050, 0.041] HA 1.024 [1.003, 1.029] 1.022 [1.002, 1.034] 0.002 [-0.050, 0.052] USG – arrival CON 1.013 [1.000, 1.030] 1.023 [1.000, 1.029] 0.010 [-0.026, 0.047]

HA 1.017 [1.000, 1.030] 1.016 [1.000, 1.029] 0.000 [-0.037, 0.038] Body fluid loss (kg) CON 1.1 [0.8, 1.3] 1.1 [0.8, 1.3] -0.03 [-0.2, 0.2]

HA 1.4 [1.1, 1.7]* 1.4 [1.2, 1.7]* 0.02 [0.1, 0.2] Relative body fluid loss (%) CON 1.3 [1.1, 1.5] 1.3 [1.1, 1.5] 0.0 [-0.2, 0.3]

HA 1.7 [1.5, 2.0]* 1.7 [1.5, 2.0]* 0.0 [-0.2, 0.2] -1 Sweat rate (L∙hour ) CON 0.83 [0.60, 1.07] 0.89 [0.66, 1.13] 0.60 [-0.04, 0.16]

HA 1.30 [1.07, 1.54]* 1.30 [1.07, 1.54]* 0 [-0.10, 0.10] Perceptual Resting TSensation (AU) CON 3.2 [2.6, 3.7] 3.5 [2.9, 4.0] 0.3 [-0.1, 0.7]

Chapter 5: Study 3 159

HA 3.5 [2.9, 4.0] 3.7 [3.1, 4.2] [0.2, -0.2, 0.6] Resting TComfort (AU) CON 1.0 [0.8, 1.3] 1.2 [1.0, 1.4] 0.2 [-0.1, 0.4]

HA 1.0 [0.8, 1.2] 1.0 [0.8, 1.2] 0.0 [-0.3, 0.3] POMS - Fatigue CON 3 [3, 4] 3 [2, 4] 0 [-1, 0]

HA 3 [3, 4] 3 [3, 4] 0 [-1, 1] POMS – Sleep CON 4 [3, 4] 4 [3, 4] 0 [0, 1]

HA 4 [3, 5] 4 [3, 4] 0 [-1, 0] POMS - Soreness CON 3 [2, 4] 3 [2, 4] 0 [-1, 1]

HA 3 [3, 4] 3 [2, 3] -1 [-2, 0] POMS - Stress CON 3 [3, 4] 3 [2, 4] 0 [-1, 0]

HA 3 [3, 4] 3 [3, 4] 0 [-1, 0]

POMS - Mood CON 4 [4, 4] 4 [3, 4] 0 [-1, 0] HA 4 [3, 4] 4 [3, 4] 0 [0, 0] * indicates a statistical condition difference at the same time-point and day.

160 Chapter 5: Study 3

Table 5.7. Exercise data for Day 1 and Day 5 of training (mean [95% credible interval]).

Training Day ∆ Day 1 to 5 Variable Condition Day 1 Day 5

Performance

Power (W) CON 172 [152, 192] 175 [155, 196] 3 [1, 5] HA 172 [152, 193] 171 [151, 192]* -1 [-3, 1] Cadence (RPM) CON 80.0 [74, 85] 78 [71, 83]# -2 [-3, -2]

HA 78 [72, 84]* 76 [70, 82]*# -2 [-3, -2] Physiological -1 Mean HR (beats∙min ) CON 139 [130, 145] 135 [127, 142]# -3 [-4, -2]

HA 155 [146, 161]* 145 [137, 152]*# -10 [-11, -8] Mean Tre (°C) CON 37.9 [37.8, 38.1] 38.0 [37.9, 38.1]# 0.05 [-0.03, 0.08] HA 38.2 [38.1, 38.3]* 38.1 [37.9, 38.2]*# -0.15 [-0.18, -0.13] Mean Tsk (°C) CON 32.3 [31.9, 32.7] 32.5 [32.1, 32.9]# 0.23 [0.15, 0.32]

HA 36.0 [35.6, 36.4]* 35.9 [35.5, 36.3]*# -0.11 [-0.19, -0.03] -1 Post blood lactate (mmol∙L ) CON 3.4 [1.0, 5.9] 3.6 [1.0, 6.1] 0.2 [-2.5, 2.7]

Chapter 5: Study 3 161

HA 5.9 [3.1, 8.9] 3.6 [1.5, 5.9] -2.3 [-5.3, 0.4] PV (%) CON 54.6 [52.7, 56.5] 56.7 [49.3, 63.9] 2.1 [-5.4, 9.4]

HA 56.2 [49.9, 62.6] 60.6 [52.4, 68.9] 4.4 [-5.5, 14.1] Perceptual RPE (AU) CON 13 [13, 14] 13 [13, 14] 0 [-1, 0]

HA 14 [13, 15]* 14 [13, 15]* 0 [-1, 0] Mean TSensation (AU) CON 4.9 [4.6, 5.3] 4.9 [4.5, 5.3] 0. [-0.2, 0.2]

HA 5.9 [5.5, 6.3]* 5.6 [5.2, 6.0]* -0.3 [-0.5, 0.0] Mean TComfort (AU) CON 2.3 [1.9, 2.6] 2.0 [1.6, 2.3]# -0.3 [-0.5, -0.1]

HA 3.1 [2.7, 3.4]* 2.9 [2.6, 3.2]*# -0.2 [-0.4, 0.0] Session RPE (AU) CON 3.9 [2.4, 5.4] 3.8 [2.3, 5.3] -0.1 [-1.7, 1.5]

HA 4.9 [3.3, 6.4] 5.0 [3.5, 6.5] -0.2 [-1.5, 1.7] * indicates a statistical condition difference at the same time-point and day; # indicates a statistical day difference at the same time-point and condition (i.e., Day 1 post-exercise to Day 5 post exercise).

162 Chapter 5: Study 3

Table 5.8. Neuromuscular properties for Day 1 and Day 5 of training (mean [95% credible interval]).

Day 1 Day 5 Variable Condition Pre Post Pre Post

MVC Torque (N∙m) CON 251 [219, 281] 222 [189, 252]† 246 [213, 281] 223 [190, 252]†

HA 244 [210, 273] 218 [185, 248]† 252 [218, 281] 227 [193, 258]† VA (%) CON 93.3 [90.1, 96.7] 89.4 [86.0, 92.9]† 95.3 [92.0, 96.7] 91.0 [87.6, 92.9]†

HA 92.6 [89.2, 96.2] 89.5 [86.0, 93.0]† 94.0 [90.7, 97.4] 89.8 [86.4, 93.2]† EMG VL (%) CON - 77.9 [65.3, 90.1] - 70.9 [56.9, 85.0

HA - 72.1 [59.0, 85.2] - 87.9 [74.5, 100] EMG VM (%) CON - 72.8 [50.2, 90.2] - 72.2 [48.5, 91.6]

HA - 68.6 [45.6, 87.7] - 52.4 [29.0, 71.5] Pt (N) CON 69.3 [60.0, 78.7] 49.4 [39.8, 59.1]† 67.4 [57.7, 78.7] 48.8 [39.5, 59.1]†

HA 67.7 [58.1, 77.2] 55.0 [45.4. 64.8]† 68.5 [58.9, 78.2] 54.3 [44.6, 63.9]† TPt (ms) CON 77 [70, 85] 67 [59, 76]† 76 [68, 85] 69 [61, 76]

HA 74 [66, 81] 65 [57, 73]† 78 [69, 86] 68 [59, 76]† 1/2 RT (ms) CON 71 [57, 84] 61 [47, 75] 74 [59, 84] 51 [36, 75]

Chapter 5: Study 3 163

HA 66 [52, 79] 47 [33, 61] 67 [52, 81] 46 [31, 60] CD (ms) CON 149 [130, 168] 129 [109, 149]† 150 [130, 168] 119 [99, 149]†

HA 137 [116, 157] 111 [91, 132]† 144 [123, 165] 113 [92, 133]† -1 RTD (N∙s ) CON 827 [773, 881] 777 [707, 847] 823 [753, 881] 748 [661, 847]

HA 865 [794, 931] 816 [730, 897] 863 [779, 946] 788 [690, 887] -1 RR (N∙s ) CON 548 [498, 602] 546 [483, 611] 536 [473, 602] 542 [467, 611]

HA 559 [495, 628] 583 [503, 667] 555 [475, 637] 597 [501, 692] † indicates a statistical time difference at same condition and day (i.e., pre- to post-exercise).

164 Chapter 5: Study 3

Table 5.9. Plasma concentrations of inflammatory markers for Day 1 and Day 5 of training (mean [95% credible interval]). Day 1 Day 5 Variable Condition Pre Post Pre Post -1 TNF-α (pg∙mL ) CON 3.2 [1.6, 4.6] 4.2 [2.5, 5.9] 2.9 [1.3, 4.6] 4.0 [2.3, 5.9]

HA 2.0 [0.4, 3.5] 4.9 [3.3, 6.5]† 2.0 [0.2, 3.5] 3.5 [1.9, 5.1] -1 I-FABP (ng∙mL ) CON 0.843 [0.475, 1.215] 0.914 [0.527, 1.289] 1.088 [0.707, 1.454] 1.122 [0.751, 1.487]

HA 0.659 [0.291, 1.032] 1.118 [0.740, 1.474]† 0.687 [0.308, 1.045] 0.953 [0.590, 1.323] † indicates a statistical time difference at same condition and day (i.e., pre- to post-exercise).

Chapter 5: Study 3 165

Chapter 6: General Discussion

The studies collated in this thesis examined the effects of exercise in hot environmental conditions on intestinal damage, endotoxemia, inflammation, and alterations in neuromuscular function. Additionally, the influence of potentially prophylactic interventions (i.e., glutamine supplementation and heat acclimation training) on self- paced cycling performance in the heat were explored.

6.1 THESIS AIMS

The development of this body of work arose from early, field-based reports of endotoxemia following prolonged athletic events in hot and/or humid events

(Bosenberg et al., 1988; Brock-Utne et al., 1988; Camus et al., 1998; Camus et al.,

1997; Jeukendrup et al., 2000). Competing demand for cardiac output between the skin and the skeletal muscle, for thermoregulation and locomotion, respectively, was found to result in a redistribution of blood away from the gut during exercise in the heat (Lim

& Mackinnon, 2006; Qamar & Read, 1987). This lead to the death of intestinal epithelial cells and hyperthermic injury to tight junctions (Pires et al., 2016), that in turn permited the translocation of endotoxins into systemic circulation (Lambert,

2008). Accumulation of endotoxins in the blood was found to evoke a strong inflammatory response and the release of pro-inflammatory cytokines (Gill, Hankey, et al., 2015; Lambert, 2009). As these signalling molecules have been linked to feelings of fatigue (Dantzer, 2004), increased pro-inflammatory cytokine activity could, theoretically, downregulate the voluntary activation of skeletal muscle (Vargas

& Marino, 2014), potentially explaining the development of central fatigue and reduced exercise performance in the heat (Nybo et al., 2014). Accordingly, the studies in this thesis aimed to investigate:

166 Chapter 6: General Discussion

1) The effect of exercise in the heat on intestinal barrier dysfunction, endotoxin

and inflammatory cytokine release and the development of central fatigue;

2) The efficacy of glutamine supplementation on self-paced cycling

performance in the heat and how this relates to neuromuscular function and

inflammation; and

3) The effect of short-duration heat acclimation training on gastrointestinal

damage, inflammation and central fatigue during subsequent exercise

performance in hot environmental conditions.

6.2 PRINCIPAL FINDINGS

The initial study (Chapter 3: Study 1) was undertaken as a proof-of-concept research project, which aimed to explore the effect of exercise in the heat on intestinal damage and permeability, endotoxemia, inflammation and central fatigue. Sixty minutes of moderate-to-vigorous intensity cycling in a hot environment (34.5 ± 0.1 °C; 55 ± 3%

RH) saw a significant rise in markers of intestinal damage (+68%; 1.043 [0.786, 1.298] ng∙mL-1) when compared to a thermoneutral condition (0.619 [0.371, 0.872] ng∙mL-1).

Post-exercise voluntary activation was also found to be significantly lower for the hot conditions (87.9% [85.2, 90.8]) compared to the thermoneutral (90.5% [87.8, 93.2]).

No difference in gut permeability, level of endotoxins or cytokines were observed.

However, this could have been due to methodological issues, such as; venepuncture collection timepoints not aligning with inflammatory marker plasma expression, degradation of blood samples due to fluctuations in storage freezer temperatures or inhibition of endotoxins due to serum binding proteins. Therefore, these biomarkers may not have accurately reflected the true inflammatory response. It could be

Chapter 6: General Discussion 167

speculated that the observed GI damage and impaired voluntary activation may tenuously suggest a potential link between intestinal barrier dysfunction and central fatigue. However, additional research is required to verify or reject this possible association.

Study 2 (Chapter 4: Study 2) was designed to address an applied and performance- oriented aspect of this research topic; the effect of glutamine supplementation on a 20 km time trial performance in the heat. The use of an externally valid, self-paced exercise protocol (i.e., 20 km cycling time trial) allowed participants to alter their performance based on their perceived level of exertion and fatigue, and better replicated the demands of real-world competition. Contrary to the research hypothesis, glutamine supplementation did not significantly alter 20 km time trial performance in trained cyclists compared to that of a placebo, with a mean difference between groups of 11 s [-23, 44]. Nevertheless, glutamine was observed to attenuate a rise in markers of GI damage (i.e., I-FABP) and inflammation (i.e., TNF-α), with no significant increase over pre-exercise levels. In contrast, the placebo condition saw statistically significant post-exercise increases in both of these markers, when compared to glutamine (I-FABP: 83% increase; TNF-α: 47% increase). Despite the similar 20 km time trial performance, and therefore exercise workload, between conditions, post- exercise torque only decreased in the placebo trial (-9.3%), whereas glutamine was not statistically different from baseline values. Interestingly, this maintenance of muscular strength did not result in 20 km time trial improvements, possibly because the small torque difference (9 N∙m) between conditions was not physiologically meaningful.

These findings tentatively suggest that glutamine supplementation may reduce cumulative gut damage and maintain neuromuscular function. While not ergogenic in short duration exercise tasks, glutamine could possibly prove advantageous during

168 Chapter 6: General Discussion

longer-duration events in hot conditions (e.g., prolonged races and multi-day competitions).

The third and final study (Chapter 5: Study 3) examined the effect of short-duration heat acclimation training on 20 km time trial performance, endotoxemia, inflammation, gut damage and neuromuscular function. These data demonstrated that

5 days of cycling in the heat improved time trial performance by 2.9% [0.8-4.9] and enhanced knee extensor strength (+4.9% [4.4, 5.2]), but did not result in traditional heat acclimation adaptations (i.e., alterations in resting HR, Tre or Tsk). Enhanced cardiovascular stability, arising from heat acclimation-induced plasma volume expansion (+7% [3, 12]), may have preserved blood flow to the gut and minimised intestinal damage. Alternatively, the moderate core temperatures and short cycling duration of the 20TT may simply not have caused enough thermal strain to result in damage to the intestinal barrier. Further, the faster post-heat acclimation 20 km time trial was not observed to incur additional central fatigue, elevate inflammatory cytokines or increase circulating endotoxins. In contrast, 5 days of matched thermoneutral training did not improve 20 km time trial performance and resulted in cumulative peripheral fatigue of the knee extensors (-5.2% [-4.6, -6.0]). An interesting finding of this study was the divergence in neuromuscular function between conditions, with heat acclimation reporting increased, and CON decreased, knee extensor strength. This study supports the ergogenic potential of short-duration heat acclimation training before short-duration athletic competition in hot conditions; although the effect of such training on neuromuscular function and intestinal barrier integrity in an applied setting is less clear.

Chapter 6: General Discussion 169

6.3 INTEGRATIVE NEUROINFLAMMATORY MODEL OF FATIGUE – A CONCEPT REVISITED

Hot environmental conditions are well-known to impair athletic performance, as the thermoregulatory burden results in premature fatigue (Galloway & Maughan, 1997;

Nybo et al., 2014; Périard, Cramer, et al., 2011a; Schlader, Stannard, et al., 2011;

Tatterson et al., 2000). While numerous mechanisms have been suggested as potential driving factors behind the development of hyperthermic fatigue, considerable research has focused on a causal association between core temperature and centrally-mediated reductions in skeletal muscle activation (Nybo et al., 2014). However, as contradictory observations have diminished the credibility of a direct mechanistic link between these two factors, researchers have urged that alternative possibilities also be explored

(Nybo & González-Alonso, 2015). Thus, we proposed that inflammation, arising from exertional-endotoxemia, may elucidate the relationship between hyperthermic exercise and the development of central fatigue. Specifically, heat-induced damage to the intestinal barrier (Pires et al., 2016) could allow for translocation of endotoxins into the systemic circulation and the subsequent release of pro-inflammatory cytokines

(Lim & Mackinnon, 2006). These signalling molecules may potentially modulate the

CNS, both directly, by crossing the blood-brain barrier, or indirectly, through afferent feedback to the peripheral nervous system (Dantzer, 2004). The resultant cytokine- mediated fatigue could downregulate voluntary drive and result in impaired performance (Vargas & Marino, 2014).

However, evidence from Study 1 (Chapter 3: Study 1) did not support this proposed neuroinflammatory model, as voluntary activation reductions were observed to occur in the absence of endotoxemia. While sample storage issues may explain this finding,

Studies 2 and 3 (Chapter 4: Study 2; Chapter 5: Study 3) also observed no detectable changes in circulating endotoxin concentrations, despite elevated core temperatures

170 Chapter 6: General Discussion

and central fatigue. This outcome was particularly surprising when viewed beside previous literature which has repeatedly reported endotoxin increases following exercise-heat stress (Table 2.1), Notably, however, the majority of these studies utilised athletes competing in prolonged, field-based events where cumulative strain from additional factors may also affect endotoxin translocation (e.g., dehydration, event duration and radiant heat load). Further, the use of cycling as a modality is virtually unique to this current research project, which may have eliminated the additional mechanical stress and tearing of the intestinal barrier that is associated with running (Lim & Mackinnon, 2006; van Nieuwenhoven et al., 2004). Indeed, of the three studies that have investigated cycling and endotoxemia (Table 2.1), only one utilised a hot environment (Guy et al., 2016), and translocation was only reported by a single group (Ashton et al., 2003).

Despite the lack of detectable endotoxemia, all three studies (Chapter 3: Study 1;

Chapter 4: Study 2 and Chapter 5: Study 3) reported that strenuous cycling in the heat significantly increased GI damage, which aligns with previous research on this issue

(Lambert, 2004; Lim & Mackinnon, 2006). Study 1 (Chapter 3: Study 1) demonstrated that hot environments place additional strain on required cardiac output, theoretically resulting in gut hypoxia leading to additional intestinal damage compared to temperate conditions (I-FABP: +68%; Figure 3.2). As GI barrier dysfunction has been proposed as the initiatory factor for exertional-endotoxemia, a metaphorical ‘canary in the coal mine’ (Lambert, 2008), it is surprising that all three studies failed to detect endotoxins, despite an increase in biochemical markers of gut damage. Further, the rise in cytokines (i.e., TNF-α) seen in Study 2 and 3 (Chapter 4: Study 2; Chapter 5: Study 3) supports the possibility that increased translocation of endotoxins was driving this inflammatory cascade. Specifically, a concurrent rise in gut damage (i.e., I-FABP) and

Chapter 6: General Discussion 171

TNF-α was observed during Study 2 (Chapter 4: Figure 4.3) and on the first day of heat acclimation in Study 3 (Chapter 5: Table 5.9), potentially implying a slight translocation of endotoxins. One explanation may be the occurrence of endotoxic flux, with exercise-heat stress permitting endotoxins to pass through the weakened intestinal barrier, producing the observed inflammatory cytokine response, before being inactivated and neutralised by anti-LPS antibodies and low-density lipoproteins (Lim

& Mackinnon, 2006). Measurement of the quantity of these circulating anti-LPS antibodies could also have been utilised as a potential indictor of endotoxin release during exercise, as has been previously described (Bosenberg et al., 1988; Brock-Utne et al., 1988; Camus et al., 1997). In comparison to the shorter duration of the present studies and previous laboratory-based research (~20–140 min), the prolonged endurance events (~3-7 h) described in previous research (Table 2.1), may arguably overload the limited capacity of these clearance mechanisms, and therefore explain the detection of post-exercise endotoxemia.

Mechanistically, the proposed neuroinflammatory association between cytokine release and the development of central fatigue (Vargas & Marino, 2014) was not supported in present studies. For example, while voluntary activation was comparably diminished for both conditions in Study 2, with a corresponding increase in IL-6

(Figure 4.2; Figure 4.3), this was not observed in Study 3 (Table 5.4; Table 5.5).

Similarly, TNF-α was only significantly elevated in one condition for Study 2 (Figure

4.3), despite a parallel decrease in voluntary activation for all groups (Figure 4.2; Table

5.4). In contrast, the TNF-α concentration was moderately associated with central fatigue (i.e., VA) following heat acclimation training (ρ = 0.43). This inconsistent relationship between inflammation and central fatigue does not support the thesis hypothesis, although future research into this potential association utilising a

172 Chapter 6: General Discussion

mechanistic study design and different exercise durations should also be considered.

The use of isometric MVCs to determine voluntary activation, which is itself analogous to combined central and peripheral activation, may not accurately reflect the true neuromuscular demands faced by athletes during dynamic exercise in the heat.

The inclusion of additional measures of neural activity during physical activity, such as electroencephalogram, may have also provided additional insights into neural activity and the association between hyperthermic fatigue and cytokinemia (Périard,

De Pauw, Zanow, & Racinais, 2018; Roelands, De Pauw, & Meeusen, 2015).

Brief MVCs may have also masked the identification of central fatigue, as the short duration (~5 s) of the contractile requirements may have been met by a momentary upregulation of motor unit activation (Périard, Caillaud, et al., 2011). In contrast, longer-duration MVCs (e.g., 2 min) requiring sustained voluntary activation may have better revealed possible impairments in central drive (Nybo & Nielsen, 2001a).

However, as dynamic exercise involves multiple, brief sub-maximal muscular contractions, it could be argued that the external validity of sustained isometric MVC tests to cycling is questionable (Cheung, 2007). Therefore, care should be taken when attempted to extrapolate reported impairments in neuromuscular function from sustained, maximal isometric tasks to dynamic, sub-maximal exercise tasks. For example, the present thesis observed torque reductions during 5 s MVCs, which is in contrast to previous findings that force is reduced during sustained (~30 s), but not brief (3-5 s), isometric and isokinetic MVCs (Cheung & Sleivert, 2004a, 2004b; Nybo

& Nielsen, 2001b; Todd et al., 2005). However, as it is difficult to draw parallels between MVCs tests and the true neuromuscular stresses experienced during dynamic cycling, the relevance of these current findings to applied exercise science is debatable.

Chapter 6: General Discussion 173

A novel finding of the present thesis was the inverse relationship observed between

TNF-α concentration and knee extensor strength (i.e., MVC torque) following exercise in the heat (Figure 4.2; Figure 4.3; Table 5.8; Table 5.9). Interestingly, this neuro- inflammation link seems to mirror the initially-proposed model of fatigue (Chapter

2.4: Neuroinflammatory fatigue). However, the elevation in pro-inflammatory cytokines appears to induce peripheral, as opposed to central, fatigue. Mechanistically this may occur due to a TNF-α inhibition of myofilament contractile properties arising from stimulation of intracellular oxidant cascades (Hardin et al., 2008) and resulting in fatigue that is distal to the neuromuscular junction. Study 2 demonstrated that strength deficits (Figure 4.2) did not result in detrimental cycling performance over a short-duration activity (~33 min). However, the occurrence of peripheral fatigue throughout prolonged athletic events, such as an ultra-marathon (Gill, Hankey, et al.,

2015), may cumulatively and progressively impair exercise performance (Amann et al., 2013). Further, it could be postulated that prolonged periods of elevated TNF-α may increase afferent feedback via signal amplification or peripheral nerve innervation

(Szelényi, 2001), evoking the sensations of pain and discomfort (Dantzer, 2004).

Psychophysiological aspects which feed into the development of fatigue have been previously described in detail (Enoka, 1995), and include increased sensation of discomfort, pain and the perception of effort. As mentioned previously, systemic inflammatory cytokines may indirectly modulate neuronal activity through afferent feedback (Vargas & Marino, 2014), although this concept is not supported by the findings in the present thesis. Alternatively, endotoxemia may induce up-regulation of receptor-mediated transcytosis, allowing cytokines to cross the blood-brain barrier and act directly on the CNS (Varatharaj & Galea, 2017). Hyperthermia-mediated alterations in blood-brain barrier permeability could also conceivably permit cytokines

174 Chapter 6: General Discussion

to diffuse across the disrupted barrier and influence neurotransmitter function, and thus fatigue (Davis & Bailey, 1997; Sharma & Hoopes, 2003). Alterations in neurotransmitter function have been proposed to potentially modify pacing strategies, as well as the development, or attenuation, of central fatigue during exercise in the heat (Nybo et al., 2014; Roelands et al., 2013). While the majority of neurotransmitter research has involved CNS-manipulating drugs, it could be reasoned that a cytokine- mediated change in neurotransmitter concentrations may also affect motivation, perception and influence pacing. Unfortunately, this hypothesis was unable to be tested in the present thesis due to the study design and research methodology.

Acute glutamine supplementation (Chapter 4: Study 2) was found to protect the GI system from hyperthermia-induced damage. Glutamine may have attenuated the rise in circulating inflammatory cytokines (i.e., TNF-α), although this did not result in improvements in cycling performance. Prolonged elevation of circulating TNF-α may, as previously discussed, contribute to peripheral and central fatigue. Therefore, glutamine supplementation could prove ergogenic during longer-duration activities, where sustained inflammation may impair contractile function or central drive.

Reduced inflammation may also improve psychological status (e.g., motivation) and preserve immune status over multi-day competitions (Dantzer, 2004; Davis & Bailey,

1997; Vargas & Marino, 2014). However, we recognise that the between-conditions difference in TNF-α levels observed in Study 2 (Chapter 4: Study 2) may not be physiologically relevant, and additional research is required to determine the possible effect of glutamine on circulating levels of inflammatory cytokines. An additional, and unexpected, benefit of glutamine supplementation may also arise from its protective effect on the intestinal barrier (Figure 4.3), by reducing exercise-induced GI problems

Chapter 6: General Discussion 175

(e.g., diarrhoea, nausea and vomiting) (de Oliveira, Burini, & Jeukendrup, 2014) and improving nutritional digestion and absorption (van Wijck et al., 2012).

Short-term heat acclimation training protocols have become a popular intervention strategy aimed at efficiently improving performance in hot environments. The findings of Study 3 (Chapter 5: Study 3) indicate that 5 days of moderate intensity training in the heat improves subsequent self-paced athletic performance in the heat, without inducing additional inflammation or endotoxemia (Table 5.2; Table 5.5). These results add to the scarcity of literature around short-term heat acclimation and endotoxemia, and provide novel data due to the methodological inclusion of an externally-valid self- paced exercise protocol and control group (Barberio et al., 2015; Guy et al., 2016;

Kuennen et al., 2011). As previously theorised (Lambert, 2004), enhanced plasma volume (Table 5.1) may have contributed to maintaining gut blood flow, diminishing

GI damage and minimising inflammatory stress, despite the greater workload completed.

A particularly novel finding of Study 3 (Chapter 5: Study 3) was the divergence in neuromuscular strength between heat acclimation and thermoneutral training, despite similarities in voluntary activation and peripheral twitch properties. We hesitantly postulated that this might have occurred due to hyperthermia-induced hypertrophy or enhanced neural recruitment from the multiple days of cycling. Although speculative, another possibility highlights the dual nature of inflammatory cytokines, as harmful signalling molecules that perpetuate pathological pain (Zhang & An, 2007), but also important regulators of skeletal muscle regeneration and homeostasis (Yang & Hu,

2018). For example, acute elevations in pro-inflammatory cytokines (i.e., TNF-α) may disrupt contractile function (Reid et al., 2002). Conversely, repeated inflammation arising from multiple days of heat acclimation training (Study 3) may also assist with

176 Chapter 6: General Discussion

stimulating the repair and regeneration of exercise-induced skeletal muscle injury

(Yang & Hu, 2018). However, we recognise that the reported TNF- α values in Study

3 (Chapter 5: Study 3) could be deemed minimal, and stress that the aforementioned concepts should be considered theoretical and speculative. Future research should attempt to investigate the hypothesised association between inflammatory cytokines, heat acclimation and skeletal muscle injury and/or repair.

6.4 CONCLUSION

The findings presented in this thesis suggest that exercise in the heat may result in GI damage, pro-inflammatory cytokine release and impaired neuromuscular force due to the development of central, and peripheral, fatigue. Interestingly, no study observed a detectable change in endotoxin concentration following hyperthermic-exercise, although the increase in gut damage and inflammation may support a possible endotoxic flux. Additionally, this thesis provides evidence that short-duration heat acclimation training can be undertaken by an athlete before a competitive event to improve exercise performance without additional inflammation or endotoxemia.

Athletic performance could potentially be further enhanced through the acute ingestion of glutamine before completion in a hot environment; reducing gut damage. Future research should continue to elucidate the possible relationship between hyperthermia, intestinal damage, inflammation and neuromuscular function, and aim to refine, develop or refute the neuroinflammatory model of fatigue.

Chapter 6: General Discussion 177

6.5 LIMITATIONS AND DELIMITATIONS

x Well-trained male cyclists were used for Studies 1 and 2, while Study 3 used

recreationally-trained males. The use of exclusively young, male

participants may restrict the extrapolation of these results to other

populations (e.g., masters athletes and females).

x Running is a more common exercise modality in which exertional-

endotoxemia is seen, as the mechanical movement contributes to greater GI

damage. However, as a self-paced treadmill was not available, cycling was

utilised to ensure participants could freely alter their workload (Studies 2

and 3).

x Sample sizes of all three studies were small (8-12 participants) as the

invasive nature of the research and restrictive inclusion criteria made

recruitment challenging. This limitation was lessened by the use of Bayesian

statistics for all studies, where vague informative prior distributions are

combined with raw collected data (likelihood) to determine posterior

predicted values (Kruschke & Liddell, 2018; McNeish, 2016). Importantly,

Bayesian analysis of hierarchical models applies mutual constraints between

different level parameters, reducing spurious posterior outliers and

decreasing the chance of false-positive results (Kruschke & Liddell, 2018).

x The short-duration 20 km time trials of Study 2 and 3 do not reflect the

prolonged exercise tasks of field-based exertional-endotoxemia studies

(Brock-Utne et al., 1988; Gill, Hankey, et al., 2015; Gill, Teixeira, et al.,

2015).

178 Chapter 6: General Discussion

- However, the high core temperatures and rate of heat storage

during the 20 km time trials suggest that some participants may

have reached the ethical trial termination temperature (40 °C) if

the task had been longer (e.g., 40 km time trial).

x The addition of a fixed-load exercise task before the post-acclimation 20 km

time trial in Study 3 would have allowed identification of heat acclimation

adaptations (physiological and perceptual), that may have been masked by

the variable intensity of the self-paced 20 km time trial

x The fixed-intensity heat acclimation protocol used for Study 3 (60 min at

50% VO2max) may not result in optimal adaptations for all participants when

compared to a controlled-hyperthermia protocol.

- However, fixed-workload training does not require access to

invasive core temperature measurement devices.

x The environmental conditions of the heat acclimation training protocol for

Study 3 (35 °C; 50% relative humidity) are lower than other similar short-

term heat acclimation studies (Garrett et al., 2009; Guy et al., 2016), and

this reduced thermal load combined with the short-duration, fixed-intensity

workload may have led to sub-optimal adaption.

x Additional blood collection and neuromuscular assessment time points

following Study 2 (e.g., 1 h post-exercise) or Study 3 (e.g., training days 2,

3 and 4) would have provided additional information around the clearance

of inflammatory biomarkers and neuromuscular recovery.

x Intestinal permeability was assessed through blood analysis (i.e., CLDN-3)

and not the more common lactulose:rhamnose urinary excretion method.

Chapter 6: General Discussion 179

Furthermore, CLDN-3 levels were only measured for Study 1, and not for

Studies 2 and 3, due to funding limitations.

x Issues with the storage of Study 1 blood samples (-20 °C freezer) may have

altered the level of detectable biomarkers, and these values may not

accurately reflect the true concentrations.

x Dietary intake was not controlled during any study.

- Participants were asked to diarise their food intake for the 24 h before

each initial experimental trial and then replicate this before the

subsequent trials.

x Dietary protein intake, and therefore dietary glutamine intake, was not

controlled during Study 2.

- However, this is unlikely to have confounded the study as: 1)

participants were asked to replicate their diet for the 24 h prior to

each experiment trial, ensuring that dietary glutamine was similar

between trials; 2) the average daily dietary glutamine intake (~3-

6 g∙day-1; (Gleeson, 2008)) is considerably smaller than the

quantity consumed as a single dose during the study (~58 ± 6 g);

and 3) glutamine is a non-essential amino acid and is therefore

produced by the body.

x Neuromuscular function was assessed via isometric MVCs of the right knee

extensor muscles, immediately pre- and post-exercise. While the time delay

between the cycle tasks and neuromuscular testing was minimised, these

neuromuscular measurements may not accurately reflect the true neural

processes occurring during exercise. The neural recruitment patterns during

180 Chapter 6: General Discussion

dynamic exercise may also differ from those required during an isometric

MVC. Further, the demands of an isometric MVC (sustained maximal

contraction) are divergent to that of s dynamic exercise task such as cycling

(brief, repeated, submaximal contractions)

x The use of a single twitch, as opposed to a doublet or train stimulation, may

have impaired the sensitivity to detect changes in voluntary activation and

muscle contractile properties.

x The use of transcranial magnetic stimulation (TMS) would have allowed for

the identification of the spinal and supraspinal components of central

fatigue.

- Nerve stimulation and the twitch interpolation technique is

commonly used to differentiate between central and peripheral

fatigue, and was utilised in all studies in the thesis.

x Electroencephalogram during the exercise task would have provided

information about the neural activity of the brain during the dynamic

task, and identification of possible changes due to interventions or

environmental conditions.

x Due to funding, only a small number of cytokines were able to be

measured (i.e., IL-1β, IL-6, TNF-α). These cytokines were primarily pro-

inflammatory in nature and are expressed due to physiological stress.

Because of this, these selected cytokines have been commonly measured

in previous thermal physiology and exertional-endotoxemia research

(see Table 2.2).

Chapter 6: General Discussion 181

6.6 PRACTICAL APPLICATIONS

x Exercise in the heat may damage the GI tract. There is a possibility that

intestinal disturbance may lead to detrimental performance and

symptoms of GI distress.

x Acute glutamine supplementation may present a cheap and effective

intervention strategy to reduce GI damage.

x A single 0.9g∙kg-1 FFM dose ingested ~60 min before athletic

competition appears to be well-tolerated and without identified side-

effects.

x Short-term heat acclimation training (60 min at 50% Pmax) presents a

time-efficient method to improve self-paced, short duration performance

in the heat, without increased GI damage, endotoxemia or inflammation.

x A combination of these ergogenic interventions may provide

considerable benefit for athletes preparing to compete in hot

environmental conditions. Specifically, short-term heat acclimation

training could be implemented in the week before, and glutamine

ingested the day of competition.

182 Chapter 6: General Discussion

6.7 RECOMMENDATIONS FOR FUTURE RESEARCH

Further research should be conducted to verify and elucidate the proposed relationship between intestinal damage, inflammation and fatigue. The analysis of additional inflammatory molecules and stress markers over multiple time points would provide information around the pro- and anti-inflammatory cascade and clearance within the body. For example, measurement of endotoxin-clearance molecules, such as Ig(G) and

Ig(M), may provide information around the appearance of endotoxins in the blood, as well as the immune status of the participant.

While this research focused on inflammation and neuromuscular function, the relationship between cytokinemia and mental fatigue should also be explored. Clinical patients with chronic inflammation report lower levels of motivation, reduced performance in cognitive tasks and altered mood-states when compared to healthy controls. Therefore, it stands to reason that a transient increase in inflammatory cytokines during exercise in the heat may affect an athlete’s ability to make appropriate tactical decisions or maintain the required focus. This potential relationship between intestinal dysfunction, inflammation and poor mental acuity may also be of importance to military and similar occupations. For example, firefighters or bomb disposal technicians often undertake considerable physical activity in hot environments, while wearing restrictive protective clothing. These occupations also require workers to make high-consequence decisions and complete tasks which require advanced cognitive function.

Pre- and post-exercise cooling strategies, such as ice slurry ingestion or cold-water immersion, are common strategies used to delay, or reduce, elevations in core temperature. However, the effect that these interventions may have on preventing or modulating endotoxemia and inflammation, particularly over consecutive days,

Chapter 6: General Discussion 183

remains to be seen. Further, the use of passive heat acclimation (e.g., hot water bath) may also provide additional protection against subsequent endotoxemia due to sub- lethal leakage of LPS resulting in an upregulation in clearance mechanisms. If employed before surgery, this intervention may reduce the risk of developing sepsis and systemic shock in clinical patients.

Heat shock proteins (HSP) are well known to drive heat adaptation responses, while also providing a cross-tolerance to endotoxemia (Ryan et al., 1992). Interestingly, hypoxic stress, such as altitude, also appears to induce the production of HSP and provide enhanced tolerance to heat. To the best of our knowledge, the relationship and possible cross-adaptations between altitude, heat and endotoxemia have not yet been explored, particularly in an athletic population.

The observed preservation, and enhancement, of knee extensor torque in Study 2

(Chapter 4: Study 2) and Study 3 (Chapter 5: Table 5.9), respectively, are novel and unexpected findings. Future research should endeavour to explore this area, ideally with more sensitive equipment, such as TMS, electroencephalogram and wireless

EMG, which could better identify alterations in neural and muscular activation during dynamic activity, as well as supraspinal, spinal and peripheral fatigue.

184 Chapter 6: General Discussion

Chapter 7: Bibliography

Abbiss, C. R., Burnett, A., Nosaka, K., Green, J. P., Foster, J. K., & Laursen, P. B. (2010). Effect of hot versus cold climates on power output, muscle activation, and perceived fatigue during a dynamic 100-km cycling trial. Journal of Sports Sciences, 28(2), 117-125.

Abbiss, C. R., & Laursen, P. B. (2008). Describing and understanding pacing strategies during athletic competition. Sports Medicine, 38(3), 239-252.

Aggarwal, B. B., Gupta, S. C., & Kim, J. H. (2012). Historical perspectives on tumor necrosis factor and its superfamily: 25 years later, a golden journey. Blood, 119(3), 651-665.

Akagi, R., Ohno, M., Matsubara, K., Nakai, A., & Inouye, S. (2013). Glutamine protects intestinal barrier function of colon epithelial cells from ethanol by modulating HSP70 expression. Pharmacology, 91(1-2), 104-111.

Allan, J. R., & Wilson, C. G. (1971). Influence of acclimatization on sweat sodium concentration. Journal of Applied Physiology, 30(5), 708-712.

Allen, G. M., Gandevia, S. C., & Mckenzie, D. K. (1995). Reliability of measurements of muscle strength and voluntary activation using twitch interpolation. Muscle & Nerve, 18(6), 593-600.

Amann, M., Venturelli, M., Ives, S. J., Mcdaniel, J., Layec, G., Rossman, M. J., & Richardson, R. S. (2013). Peripheral fatigue limits endurance exercise via a sensory feedback-mediated reduction in spinal motoneuronal output. Journal of Applied Physiology, 115(3), 355-364.

Ament, W., & Verkerke, G. J. (2009). Exercise and Fatigue. Sports Medicine, 39(5), 389-422.

Andreasen, A., Krabbe, K., Krogh-Madsen, R., Taudorf, S., Pedersen, B., & Moller, K. (2008). Human endotoxemia as a model of systemic inflammation. Current Medicinal Chemistry, 15(17), 1697-1705.

Armstrong, L. E., Pumerantz, A. C., Fiala, K. A., Roti, M. W., Kavouras, S. A., Casa, D. J., & Maresh, C. M. (2010). Human hydration indices: acute and longitudinal reference values. International Journal of Sport Nutrition and Exercise Metabolism, 20(2), 145-153.

Ashton, T., Young, I. S., Davison, G. W., Rowlands, C. C., Mceneny, J., Van Blerk, C., . . . Jackson, S. K. (2003). Exercise-induced endotoxemia: the effect of ascorbic acid supplementation. Free Radical Biology and Medicine, 35(3), 284-291.

Chapter 7: Bibliography 185

Baker, B., Gaffin, S. L., Wells, M., Wessels, B. C., & Brock-Utne, J. G. (1988). Endotoxaemia in racehorses following exertion. Journal of the South African Veterinary Association, 59(2), 63-66.

Banks, W. A., & Erickson, M. A. (2010). The blood–brain barrier and immune function and dysfunction. Neurobiology of Disease, 37(1), 26-32.

Barberio, M. D., Elmer, D. J., Laird, R. H., Lee, K. A., Gladden, B., & Pascoe, D. D. (2015). Systemic LPS and inflammatory response during consecutive days of exercise in heat. International Journal of Sports Medicine, 36(3), 262-270.

Barnett, A., & Maughan, R. J. (1993). Response of unacclimatized males to repeated weekly bouts of exercise in the heat. British Journal of Sports Medicine, 27(1), 39-44.

Bigland-Ritchie, B., Jones, D. A., Hosking, G. P., & Edwards, R. H. T. (1978). Central and peripheral fatigue in sustained maximum voluntary contractions of human quadriceps muscle. [10.1042/cs0540609]. Clinical Science, 54(6), 609-614.

Borg, D. N., Osborne, J. O., Stewart, I. B., Costello, J. T., Sims, J. N. L., & Minett, G. M. (2018). The reproducibility of 10 and 20km time trial cycling performance in recreational cyclists, runners and team sport athletes. Journal of Science and Medicine in Sport, 21(8), 858-863.

Borg, G. (1970). Perceived exertion as an indicator of somatic stress. Scandinavian Journal of Rehabilitation Medicine, 2(2), 92-98.

Borg, G. (1998). Borg's perceived exertion and pain scales. Champaign, IL, US: Human Kinetics.

Bosenberg, A. T., Brock-Utne, J. G., Gaffin, S. L., Wells, M. T. B., & Blake, G. T. W. (1988). Strenuous exercise causes systemic endotoxemia. Journal of Applied Physiology, 65(1), 106-108.

Brazaitis, M., & Skurvydas, A. (2010). Heat acclimation does not reduce the impact of hyperthermia on central fatigue. European Journal of Applied Physiology, 109(4), 771-778.

Brock-Utne, J. G., Gaffin, S. L., Wells, M. T., Gathiram, P., Sohar, E., James, M. F., . . . Norman, R. J. (1988). Endotoxaemia in exhausted runners after a long- distance race. South African Medical Journal, 73(9), 533-536.

Burden, A., & Bartlett, R. (1999). Normalisation of EMG amplitude: an evaluation and comparison of old and new methods. Medical Engineering & Physics, 21(4), 247-257.

Byrne, C., Lee, J. K., Chew, S. A., Lim, C. L., & Tan, E. Y. (2006). Continuous thermoregulatory responses to mass-participation distance running in heat. Medicine & Science in Sports & Exercise, 38(5), 803-810.

186 Chapter 7: Bibliography

Calder, P. C., & Yaqoob, P. (1999). Glutamine and the immune system. Amino Acids, 17(3), 227-241.

Callahan, L. A., Nethery, D., Stofan, D., Dimarco, A., & Supinski, G. (2001). Free radical–induced contractile protein dysfunction in endotoxin-induced sepsis. American Journal of Respiratory Cell and Molecular Biology, 24(2), 210-217.

Camilleri, M., Madsen, K., Spiller, R., Van Meerveld, B. G., & Verne, G. N. (2012). Intestinal barrier function in health and gastrointestinal disease. Neurogastroenterology and Motility, 24(6), 503-512.

Camus, G., Nys, M., Poortmans, J. R., Venneman, I., Monfils, T., Deby-Dupont, G., . . . Duchateau, J. (1998). Endotoxaemia, production of tumour necrosis factor α and polymorphonuclear neutrophil activation following strenuous exercise in humans. European Journal of Applied Physiology and Occupational Physiology, 79(1), 62-68.

Camus, G., Poortmans, J., Nys, M., Deby-Dupont, G., Duchateau, J., Deby, C., & Lamy, M. (1997). Mild endotoxaemia and the inflammatory response induced by a marathon race. Clinical Science, 92(4), 415-422.

Cannon, J., Kay, D., Tarpenning, K. M., & Marino, F. E. (2007). Comparative effects of resistance training on peak isometric torque, muscle hypertrophy, voluntary activation and surface EMG between young and elderly women. Clinical Physiology and Functional Imaging, 27(2), 91-100.

Cannon, J., Kay, D., Tarpenning, K. M., & Marino, F. E. (2008). Reproducibility and changes in twitch properties associated with age and resistance training in young and elderly women. Scandinavian Journal of Medicine & Science in Sports, 18(5), 627-635.

Casa, D. J. (1999). Exercise in the Heat. I. Fundamentals of thermal physiology, performance implications, and dehydration. Journal of Athletic Training, 34(3), 246-252.

Castell, L. M. (2003). Glutamine Supplementation In Vitro and In Vivo, in Exercise and in Immunodepression. [journal article]. Sports Medicine, 33(5), 323-345.

Chalmers, S., Esterman, A., Eston, R., Bowering, K. J., & Norton, K. (2014). Short- term heat acclimation training improves physical performance: a systematic review, and exploration of physiological adaptations and application for team sports. Sports Medicine, 44(7), 971-988.

Chan, K. O., Wong, S. H., & Chen, Y. J. (2008). Effects of a hot environment on simulated cycling and running performance in triathletes. Journal of Sports Medicine and Physical Fitness, 48(2), 149-157.

Chen, H., Wu, Y., Zhang, Y., Jin, L., Luo, L., Xue, B., . . . Yin, Z. (2006). HSP70 inhibits lipopolysaccharide-induced NF-kappaB activation by interacting with TRAF6 and inhibiting its ubiquitination. FEBS Letters, 580(13), 3145-3152.

Chapter 7: Bibliography 187

Cheung, S. S. (2007). Hyperthermia and voluntary exhaustion: integrating models and future challenges. Applied Physiology, Nutrition and Metabolism, 32(4), 808- 817.

Cheung, S. S., & Mclellan, T. M. (1998). Heat acclimation, aerobic fitness, and hydration effects on tolerance during uncompensable heat stress. Journal of Applied Physiology, 84(5), 1731-1739.

Cheung, S. S., & Sleivert, G. G. (2004a). Lowering of skin temperature decreases isokinetic maximal force production independent of core temperature. European Journal of Applied Physiology, 91(5), 723-728.

Cheung, S. S., & Sleivert, G. G. (2004b). Multiple triggers for hyperthermic fatigue and exhaustion. Exercise and Sport Sciences Reviews, 32(3), 100-106.

Chinevere, T. D., Kenefick, R. W., Cheuvront, S. N., Lukaski, H. C., & Sawka, M. N. (2008). Effect of heat acclimation on sweat minerals. Medicine & Science in Sports & Exercise, 40(5), 886-891.

Chow, A., & Zhang, R. (1998). Glutamine reduces heat shock-induced cell death in rat intestinal epithelial cells. The Journal of Nutrition, 128(8), 1296-1301.

Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: L. Erlbaum Associates.

Corbett, J., Neal, R. A., Lunt, H. C., & Tipton, M. J. (2014). Adaptation to Heat and Exercise Performance Under Cooler Conditions: A New Hot Topic. Sports Medicine, 44(10), 1323-1331.

Crandall, C. G. (2008). Heat stress and baroreflex regulation of blood pressure. Medicine and Science in Sports and Exercise, 40(12), 2063-2070.

Crandall, C. G., & González-Alonso, J. (2010). Cardiovascular function in the heat- stressed human. Acta Physiologica, 199(4), 407-423.

Daanen, H. a. M., Racinais, S., & Périard, J. D. (2018). Heat Acclimation Decay and Re-Induction: A Systematic Review and Meta-Analysis. Sports Medicine, 48(2), 409-430.

Dantzer, R. (2001). Cytokine-Induced Sickness Behavior: Mechanisms and Implications. Annals of the New York Academy of Sciences, 933(1), 222-234.

Dantzer, R. (2004). Cytokine-induced sickness behaviour: a neuroimmune response to activation of innate immunity. European Journal of Pharmacology, 500(1), 399-411.

Dantzer, R., Heijnen, C. J., Kavelaars, A., Laye, S., & Capuron, L. (2014). The neuroimmune basis of fatigue. Trends in Neurosciences, 37(1), 39-46.

188 Chapter 7: Bibliography

Davis, J. M. (1995). Central and peripheral factors in fatigue. Journal of Sports Sciences, 13 Spec No, S49-S53.

Davis, J. M., & Bailey, S. P. (1997). Possible mechanisms of central nervous system fatigue during exercise. Medicine & Science in Sports & Exercise, 29(1), 45- 57.

Davison, G., Marchbank, T., March, D. S., Thatcher, R., & Playford, R. J. (2016). Zinc carnosine works with bovine colostrum in truncating heavy exercise–induced increase in gut permeability in healthy volunteers. The American Journal of Clinical Nutrition, 104(2), 526-536.

De-Souza, D. A., & Greene, L. J. (2005). Intestinal permeability and systemic infections in critically ill patients: Effect of glutamine. Critical Care Medicine, 33(5), 1125-1135.

De Maio, A. (1999). Heat shock proteins: facts, thoughts, and dreams. Shock, 11(1), 1-12.

De Oliveira, E. P., Burini, R. C., & Jeukendrup, A. (2014). Gastrointestinal Complaints During Exercise: Prevalence, Etiology, and Nutritional Recommendations. Sports Medicine, 44(1), 79-85.

De Pauw, K., Roelands, B., Cheung, S. S., De Geus, B., Rietjens, G., & Meeusen, R. (2013). Guidelines to classify subject groups in sport-science research. International Journal of Sports Physiology and Performance, 8(2), 111-122.

Dill, D. B., & Costill, D. L. (1974). Calculation of percentage changes in volumes of blood, plasma, and red cells in dehydration. Journal of Applied Physiology, 37(2), 247-248.

Dill, D. B., Edwards, H. T., Bauer, P. S., & Levenson, E. J. (1931). Physical performance in relation to external temperature. Arbeitsphysiologie, 4(6), 508- 518.

Dinarello, C. A. (1997). Interleukin-1. Cytokine & Growth Factor Reviews, 8(4), 253- 265.

Ding, X. Z., Fernandez-Prada, C. M., Bhattacharjee, A. K., & Hoover, D. L. (2001). Expression of HSP-70 inhibits bacterial lipopolysaccharide-induced production of cytokines in human monocyte-derived macrophages. Cytokine, 16(6), 210-219.

Dokladny, K., Lobb, R., Wharton, W., Ma, T. Y., & Moseley, P. L. (2010). LPS- induced cytokine levels are repressed by elevated expression of HSP70 in rats: possible role of NF-κB. Cell Stress and Chaperones, 15(2), 153-163.

Dokladny, K., Moseley, P. L., & Ma, T. Y. (2006). Physiologically relevant increase in temperature causes an increase in intestinal epithelial tight junction

Chapter 7: Bibliography 189

permeability. American Journal of Physiology - Gastrointestinal and Liver Physiology, 290(2), G204-G212.

Dokladny, K., Wharton, W., Lobb, R., Ma, T. Y., & Moseley, P. L. (2006). Induction of physiological thermotolerance in MDCK monolayers: contribution of heat shock protein 70. Cell Stress and Chaperones, 11(3), 268-275.

Dokladny, K., Ye, D., Kennedy, J. C., Moseley, P. L., & Ma, T. Y. (2008). Cellular and molecular mechanisms of heat stress-induced up-regulation of occludin protein expression : regulatory role of heat shock factor-1. The American Journal of Pathology, 172(3), 659-670.

Dokladny, K., Zuhl, M. N., & Moseley, P. L. (2016). Intestinal epithelial barrier function and tight junction proteins with heat and exercise. Journal of Applied Physiology, 120(6), 692-701.

Dugan, M. E. R., & Mcburney, M. I. (1995). Luminal glutamine perfusion alters endotoxin-related changes in ileal permeability of the piglet. Journal of Parenteral and Enteral Nutrition, 19(1), 83-87.

Ely, B. R., Ely, M. R., Cheuvront, S. N., Kenefick, R. W., Degroot, D. W., & Montain, S. J. (2009). Evidence against a 40 degrees C core temperature threshold for fatigue in humans. Journal of Applied Physiology, 107(5), 1519-1525.

Ely, B. R., Lovering, A. T., Horowitz, M., & Minson, C. T. (2014). Heat acclimation and cross tolerance to hypoxia. Temperature, 1(2), 107-114.

Enoka, R. M. (1995). Mechanisms of muscle fatigue: Central factors and task dependency. Journal of Electromyography and Kinesiology, 5(3), 141-149.

Enoka, R. M. (1997). Neural adaptations with chronic physical activity. Journal of Biomechanics, 30(5), 447-455.

Enoka, R. M., & Stuart, D. G. (1992). Neurobiology of muscle fatigue. Journal of Applied Physiology, 72(5), 1631-1648.

Epstein, Y., & Roberts, W. O. (2011). The pathopysiology of heat stroke: an integrative view of the final common pathway. Scandinavian Journal of Medicine & Science in Sports, 21(6), 742-748.

Febbraio, M. A., Snow, R. J., Hargreaves, M., Stathis, C. G., Martin, I. K., & Carey, M. F. (1994). Muscle metabolism during exercise and heat stress in trained men: effect of acclimation. Journal of Applied Physiology, 76(2), 589-597.

Fleming, S. E., Zambell, K. L., & Fitch, M. D. (1997). Glucose and glutamine provide similar proportions of energy to mucosal cells of rat small intestine. American Journal of Physiology - Gastrointestinal and Liver Physiology, 273(4), G968- G978.

190 Chapter 7: Bibliography

Flouris, A. D., & Schlader, Z. J. (2015). Human behavioral thermoregulation during exercise in the heat. Scandinavian Journal of Medicine & Science in Sports, 25(S1), 52-64.

Fox, R. H., Goldsmith, R., Kidd, D. J., & Lewis, H. E. (1963). Acclimatization to heat in man by controlled elevation of body temperature. The Journal of physiology, 166(3), 530-547.

Fritzsche, R. G., Switzer, T. W., Hodgkinson, B. J., & Coyle, E. F. (1999). Stroke volume decline during prolonged exercise is influenced by the increase in heart rate. Journal of Applied Physiology, 86(3), 799-805.

Fuller, A., Carter, R. N., & Mitchell, D. (1998). Brain and abdominal temperatures at fatigue in rats exercising in the heat. Journal of Applied Physiology, 84(3), 877- 883.

Furuse, M., Hirase, T., Itoh, M., Nagafuchi, A., Yonemura, S., Tsukita, S., & Tsukita, S. (1993). Occludin: a novel integral membrane protein localizing at tight junctions. The Journal of Cell Biology, 123(6), 1777-1788.

Gagge, A. P., Stolwijk, J. A., & Hardy, J. D. (1967). Comfort and thermal sensations and associated physiological responses at various ambient temperatures. Environmental Research, 1(1), 1-20.

Galloway, S. D. R., & Maughan, R. J. (1997). Effects of ambient temperature on the capacity to perform prolonged cycle exercise in man. Medicine & Science in Sports & Exercise, 29(9), 1240-1249.

Gandevia, S. C. (2001). Spinal and Supraspinal Factors in Human Muscle Fatigue. Physiological Reviews, 81(4), 1725-1789.

Garrett, A., Rehrer, N., & Patterson, M. (2011). Induction and decay of short-term heat acclimation in moderately and highly trained athletes. Sports Medicine, 41(9), 757-771.

Garrett, A. T., Creasy, R., Rehrer, N. J., Patterson, M. J., & Cotter, J. D. (2012). Effectiveness of short-term heat acclimation for highly trained athletes. European Journal of Applied Physiology, 112(5), 1827-1837.

Garrett, A. T., Goosens, N. G., Rehrer, N. G., Patterson, M. J., & Cotter, J. D. (2009). Induction and decay of short-term heat acclimation. European Journal of Applied Physiology, 107(6), 659.

Gibson, O. R., Mee, J. A., Tuttle, J. A., Taylor, L., Watt, P. W., & Maxwell, N. S. (2015). Isothermic and fixed intensity heat acclimation methods induce similar heat adaptation following short and long-term timescales. Journal of Thermal Biology, 49-50, 55-65.

Gill, N., & Sleivert, G. (2001). Effect of daily versus intermittent exposure on heat acclimation. Aviation, Space, and Environmental Medicine, 72(4), 385-390.

Chapter 7: Bibliography 191

Gill, S. K., Hankey, J., Wright, A., Marczak, S., Hemming, K., Allerton, D. M., . . . Costa, R. J. (2015). The impact of a 24-h ultra-marathon on circulatory endotoxin and cytokine profile. International Journal of Sports Medicine, 36(8), 688-695.

Gill, S. K., Teixeira, A., Rama, L., Prestes, J., Rosado, F., Hankey, J., . . . Costa, R. J. (2015). Circulatory endotoxin concentration and cytokine profile in response to exertional-heat stress during a multi-stage ultra-marathon competition. Exercise Immunology Review, 21, 114-128.

Gleeson, M. (2008). Dosing and Efficacy of Glutamine Supplementation in Human Exercise and Sport Training. The Journal of Nutrition, 138(10), 2045S-2049S.

Gnauck, A., Lentle, R. G., & Kruger, M. C. (2016). Chasing a ghost? – Issues with the determination of circulating levels of endotoxin in human blood. Critical Reviews in Clinical Laboratory Sciences, 53(3), 197-215.

Gong, J., & Jing, L. (2011). Glutamine induces heat shock protein 70 expression via O-GlcNAc modification and subsequent increased expression and transcriptional activity of heat shock factor-1. Minerva Anestesiologica, 77(5), 488-495.

González-Alonso, J., & Calbet, J. A. (2003). Reductions in systemic and skeletal muscle blood flow and oxygen delivery limit maximal aerobic capacity in humans. Circulation, 107(6), 824-830.

González-Alonso, J., Calbet, J. A., & Nielsen, B. (1999). Metabolic and thermodynamic responses to dehydration-induced reductions in muscle blood flow in exercising humans. The Journal of physiology, 520 Pt 2(Pt 2), 577-589.

González-Alonso, J., Crandall, C. G., & Johnson, J. M. (2008). The cardiovascular challenge of exercising in the heat. The Journal of physiology, 586(Pt 1), 45- 53.

González-Alonso, J., Mora-Rodríguez, R., Below, P. R., & Coyle, E. F. (1997). Dehydration markedly impairs cardiovascular function in hyperthermic endurance athletes during exercise. Journal of Applied Physiology, 82(4), 1229-1236.

González-Alonso, J., Teller, C., Andersen, S. L., Jensen, F. B., Hyldig, T., & Nielsen, B. (1999). Influence of body temperature on the development of fatigue during prolonged exercise in the heat. Journal of Applied Physiology, 86(3), 1032- 1039.

Goto, K., Oda, H., Kondo, H., Igaki, M., Suzuki, A., Tsuchiya, S., . . . Yoshioka, T. (2011). Responses of muscle mass, strength and gene transcripts to long-term heat stress in healthy human subjects. European Journal of Applied Physiology, 111(1), 17-27.

192 Chapter 7: Bibliography

Gray, S. R., Clifford, M., Lancaster, R., Leggate, M., Davies, M., & Nimmo, M. A. (2009). The response of circulating levels of the interleukin-6/interleukin-6 receptor complex to exercise in young men. Cytokine, 47(2), 98-102.

Grootjans, J., Thuijls, G., Verdam, F., Derikx, J. P., Lenaerts, K., & Buurman, W. A. (2010). Non-invasive assessment of barrier integrity and function of the human gut. World Journal of Gastrointestinal Surgery, 2(3), 61-69.

Grove, J. R., & Prapavessis, H. (1992). Preliminary evidence for the reliability and validity of an abbreviated profile of mood states. International Journal of Sport Psychology, 23(2), 93-109.

Guy, H. J., & Vincent, E. G. (2018). Nutrition and Supplementation Considerations to Limit Endotoxemia When Exercising in the Heat. Sports, 6(1), 12.

Guy, J. H., Deakin, G. B., Edwards, A. M., Miller, C. M., & Pyne, D. B. (2015). Adaptation to hot environmental conditions: an exploration of the performance basis, procedures and future directions to optimise opportunities for elite athletes. Sports Medicine, 45(3), 303-311.

Guy, J. H., Pyne, D. B., Deakin, G. B., Miller, C. M., & Edwards, A. M. (2016). Acclimation Training Improves Endurance Cycling Performance in the Heat without Inducing Endotoxemia. Frontiers in Physiology, 7, 318.

Habes, Q. L. M., Linssen, V., Nooijen, S., Kiers, D., Gerretsen, J., Pickkers, P., . . . Kox, M. (2017). Markers of Intestinal Damage and their Relation to Cytokine Levels in Cardiac Surgery Patients. Shock, 47(6), 709-714.

Haglind, E., Wang, D., & Klein, A. S. (1996). Hepatic reticuloendothelial system dysfunction after intestinal ischemia-reperfusion. Shock, 5(1), 72-75.

Hailes, W. S., Slivka, D., Cuddy, J., & Ruby, B. C. (2011). Human plasma inflammatory response during 5 days of exercise training in the heat. Journal of Thermal Biology, 36(5), 277-282.

Hall, D. M., Baumgardner, K. R., Oberley, T. D., & Gisolfi, C. V. (1999). Splanchnic tissues undergo hypoxic stress during whole body hyperthermia. American Journal of Physiology - Gastrointestinal and Liver Physiology, 276(5), G1195- G1203.

Hall, J. C., Heel, K., & Mccauley, R. (1996). Glutamine. British Journal of Surgery, 83(3), 305-312.

Hamiel, C. R., Pinto, S., Hau, A., & Wischmeyer, P. E. (2009). Glutamine enhances heat shock protein 70 expression via increased hexosamine biosynthetic pathway activity. American Journal of Physiology - Cell Physiology, 297(6), C1509-C1519.

Hardin, B. J., Campbell, K. S., Smith, J. D., Arbogast, S., Smith, J., Moylan, J. S., & Reid, M. B. (2008). TNF-α acts via TNFR1 and muscle-derived oxidants to

Chapter 7: Bibliography 193

depress myofibrillar force in murine skeletal muscle. Journal of Applied Physiology, 104(3), 694-699.

Hargreaves, M. (2008). Physiological limits to exercise performance in the heat. Journal of Science and Medicine in Sport, 11(1), 66-71.

Heled, Y., Fleischmann, C., & Epstein, Y. (Singer-songwriters). (2013). Cytokines and their role in hyperthermia and heat stroke. On Journal of Basic and Clinical Physiology and Pharmacology.

Hori, S. (1995). Adaptation to Heat. The Japanese Journal of Physiology, 45(6), 921- 946.

Houmard, J. A., Costill, D. L., Davis, J. A., Mitchell, J. B., Pascoe, D. D., & Robergs, R. A. (1990). The influence of exercise intensity on heat acclimation in trained subjects. Medicine & Science in Sports & Exercise, 22(5), 615-620.

Iso 9886. (2004). Evaluation of thermal strain by physiological measurements. Geneva: International Organization for Standardization.

James, K. E., Bloch, D. A., Lee, K. K., Kraemer, H. C., & Fuller, R. K. (1996). An index for assessing blindness in a multi-centre clinical trial: Disulfiram for alcohol cessation - a VA cooperative study. Statistics in Medicine, 15(13), 1421-1434.

Jeukendrup, A. E., Vet-Joop, K., Sturk, A., Stegen, J. H. J. C., Senden, J., Saris, W. H. M., & Wagenmakers, A. J. M. (2000). Relationship between gastro-intestinal complaints and endotoxaemia, cytokine release and the acute-phase reaction during and after a long-distance triathlon in highly trained men. Clinical Science, 98(1), 47-55.

Keiser, S., Flück, D., Hüppin, F., Stravs, A., Hilty, M. P., & Lundby, C. (2015). Heat training increases exercise capacity in hot but not in temperate conditions: a mechanistic counter-balanced cross-over study. American Journal of Physiology - Heart and Circulatory Physiology, 309(5), H750-H761.

Kent-Braun, J. A. (1999). Central and peripheral contributions to muscle fatigue in humans during sustained maximal effort. European Journal of Applied Physiology and Occupational Physiology, 80(1), 57-63.

King, D. S., Costill, D. L., Fink, W. J., Hargreaves, M., & Fielding, R. A. (1985). Muscle metabolism during exercise in the heat in unacclimatized and acclimatized humans. Journal of Applied Physiology, 59(5), 1350-1354.

Konsman, J. P., Parnet, P., & Dantzer, R. (2002). Cytokine-induced sickness behaviour: mechanisms and implications. Trends in Neurosciences, 25(3), 154- 159.

Kruschke, J. K., & Liddell, T. M. (2018). Bayesian data analysis for newcomers. Psychonomic Bulletin & Review, 25(1), 155-177.

194 Chapter 7: Bibliography

Kucuktulu, E., Guner, A., Kahraman, I., Topbas, M., & Kucuktulu, U. (2013). The protective effects of glutamine on radiation-induced diarrhea. Supportive Care in Cancer, 21(4), 1071-1075.

Kuennen, M., Gillum, T., Dokladny, K., Bedrick, E., Schneider, S., & Moseley, P. (2011). Thermotolerance and heat acclimation may share a common mechanism in humans. American Journal of Physiology - Regulatory, Integrative and Comparative Physiology, 301(2), 524-533.

Lambert, G. P. (2004). Role of gastrointestinal permeability in exertional heatstroke. Exercise and Sport Sciences Reviews, 32(4), 185-190.

Lambert, G. P. (2008). Intestinal barrier dysfunction, endotoxemia, and gastrointestinal symptoms: the 'canary in the coal mine' during exercise-heat stress? Medicine and Sport Science, 53, 61-73.

Lambert, G. P. (2009). Stress-induced gastrointestinal barrier dysfunction and its inflammatory effects. Journal of Animal Science, 87(14 Suppl.), E101-E108.

Lambert, G. P., Boylan, M., Laventure, J. P., Bull, A., & Lanspa, S. (2007). Effect of aspirin and ibuprofen on GI permeability during exercise. International Journal of Sports Medicine, 28(9), 722-726.

Lambert, G. P., Broussard, L. J., Mason, B. L., Mauermann, W. J., & Gisolfi, C. V. (2001). Gastrointestinal permeability during exercise: effects of aspirin and energy-containing beverages. Journal of Applied Physiology, 90(6), 2075- 2080.

Lambert, G. P., Lang, J., Bull, A., Pfeifer, P. C., Eckerson, J., Moore, G., . . . O'brien, J. (2008). Fluid Restriction during Running Increases GI Permeability. International Journal of Sports Medicine, 29(03), 194-198.

Lambert, G. P., Murray, R., Eddy, D., Scott, W., Laird, R., & Gisolfi, C. V. (1999). Intestinal permeability following the 1998 Ironman Triathlon. Medicine & Science in Sports & Exercise, 31(Supplement), S318.

Lee, F., & Scott, E. (1916). The action of temperature and humidity on the working power of muscles and on the sugar of the blood. American Journal of Physiology, 40, 486-501.

Li, N., Lewis, P., Samuelson, D., Liboni, K., & Neu, J. (2004). Glutamine regulates Caco-2 cell tight junction proteins. American Journal of Physiology - Gastrointestinal and Liver Physiology, 287(3), G726-G733.

Lim, C. L., & Mackinnon, L. T. (2006). The roles of exercise-induced immune system disturbances in the pathology of heat stroke : the dual pathway model of heat stroke. Sports Medicine, 36(1), 39-64.

Chapter 7: Bibliography 195

Lim, C. L., Pyne, D., Horn, P., Kalz, A., Saunders, P., Peake, J., . . . Mackinnon, L. T. (2009). The effects of increased endurance training load on biomarkers of heat intolerance during intense exercise in the heat. Applied Physiology, Nutrition and Metabolism, 34(4), 616-624.

Liu, S. F., & Malik, A. B. (2006). NF-κB activation as a pathological mechanism of septic shock and inflammation. American Journal of Physiology - Lung, Cellular and Molecular Physiology, 290(4), L622-L645.

Lorenzo, S., Halliwill, J. R., Sawka, M. N., & Minson, C. T. (2010). Heat acclimation improves exercise performance. [10.1152/japplphysiol.00495.2010]. Journal of Applied Physiology, 109(4), 1140-1147.

Lorenzo, S., & Minson, C. T. (2010). Heat acclimation improves cutaneous vascular function and sweating in trained cyclists. Journal of Applied Physiology, 109(6), 1736-1743.

March, D. S., Marchbank, T., Playford, R. J., Jones, A. W., Thatcher, R., & Davison, G. (2017). Intestinal fatty acid-binding protein and gut permeability responses to exercise. [journal article]. European Journal of Applied Physiology, 117(5), 931-941.

Marchbank, T., Davison, G., Oakes, J. R., Ghatei, M. A., Patterson, M., Moyer, M. P., & Playford, R. J. (2011). The nutriceutical bovine colostrum truncates the increase in gut permeability caused by heavy exercise in athletes. [10.1152/ajpgi.00281.2010]. American Journal of Physiology - Gastrointestinal and Liver Physiology, 300(3), G477-G484.

Marino, F. E. (2004a). Anticipatory regulation and avoidance of catastrophe during exercise-induced hyperthermia. Comparative Biochemistry and Physiology Part B: Biochemistry And Molecular Biology, 139(4), 561-569.

Marino, F. E. (2004b). The anticipatory regulation of fatigue in exercise. In F. E. Marino (Ed.), Regulation of Fatigue in Exercise (pp. 101-116). New York: Nova Science Publishers, Inc.

Marino, F. E., Mbambo, Z., Kortekaas, E., Wilson, G., Lambert, M. I., Noakes, T. D., & Dennis, S. C. (2000). Advantages of smaller body mass during distance running in warm, humid environments. Pflügers Archiv - European Journal of Physiology, 441(2-3), 359-367.

Maron, M. B., Wagner, J. A., & Horvath, S. M. (1977). Thermoregulatory responses during competitive marathon running. Journal of Applied Physiology, 42(6), 909-914.

Marshall, J. C. (1998). The gut as a potential trigger of exercise-induced inflammatory responses. Canadian Journal of Physiology and Pharmacology, 76(5), 479- 484.

196 Chapter 7: Bibliography

Maughan, R., Otani, H., & Watson, P. (2012). Influence of relative humidity on prolonged exercise capacity in a warm environment. European Journal of Applied Physiology, 112(6), 2313-2321.

Mcclung, J. P., Hasday, J. D., He, J. R., Montain, S. J., Cheuvront, S. N., Sawka, M. N., & Singh, I. S. (2008). Exercise-heat acclimation in humans alters baseline levels and ex vivo heat inducibility of HSP72 and HSP90 in peripheral blood mononuclear cells. American Journal of Physiology - Regulatory, Integrative and Comparative Physiology, 294(1), R185-R191.

Mclean, B. D., Coutts, A. J., Kelly, V., Mcguigan, M. R., & Cormack, S. J. (2010). Neuromuscular, endocrine, and perceptual fatigue responses during different length between-match microcycles in professional rugby league players. International Journal of Sports Physiology and Performance, 5(3), 367-383.

Mcneish, D. (2016). On Using Bayesian Methods to Address Small Sample Problems. Structural Equation Modeling: A Multidisciplinary Journal, 23(5), 750-773.

Mengersen, K. L., Drovandi, C. C., Robert, C. P., Pyne, D. B., & Gore, C. J. (2016). Bayesian Estimation of Small Effects in Exercise and Sports Science. PloS One, 11(4), e0147311.

Minett, G. M., Skein, M., Bieuzen, F., Stewart, I. B., Borg, D. N., Bach, A. J. E., & Costello, J. T. (2016). Heat acclimation for protection from exertional heat stress. Cochrane Database of Systematic Reviews(1).

Morris, G., Berk, M., Galecki, P., Walder, K., & Maes, M. (2016). The Neuro-Immune Pathophysiology of Central and Peripheral Fatigue in Systemic Immune- Inflammatory and Neuro-Immune Diseases. Molecular Neurobiology, 53(2), 1195-1219.

Morrison, A. L., Dinges, M., Singleton, K. D., Odoms, K., Wong, H. R., & Wischmeyer, P. E. (2006). Glutamine's protection against cellular injury is dependent on heat shock factor-1. [10.1152/ajpcell.00635.2005]. American Journal of Physiology - Cell Physiology, 290(6), C1625-C1632.

Morrison, S., Sleivert, G. G., & Cheung, S. S. (2004). Passive hyperthermia reduces voluntary activation and isometric force production. European Journal of Applied Physiology, 91(5-6), 729-736.

Morrison, S. A., Cheung, S. S., & Cotter, J. D. (2014). Bovine colostrum, training status, and gastrointestinal permeability during exercise in the heat: a placebo- controlled double-blind study. Applied Physiology, Nutrition, and Metabolism, 39(9), 1070-1082.

Moseley, P. L. (1997). Heat shock proteins and heat adaptation of the whole organism. Journal of Applied Physiology, 83(5), 1413-1417.

Chapter 7: Bibliography 197

Moseley, P. L., Gapen, C., Wallen, E. S., Walter, M. E., & Peterson, M. W. (1994). Thermal stress induces epithelial permeability. American Journal of Physiology - Cell Physiology, 267(2 Pt 1), C425-C434.

Mountjoy, M., Alonso, J.-M., Bergeron, M. F., Dvorak, J., Miller, S., Migliorini, S., & Singh, D. G. (2012). Hyperthermic-related challenges in aquatics, athletics, football, tennis and triathlon. British Journal of Sports Medicine, 46(11), 800- 804.

Musch, M. W., Hayden, D., Sugi, K., Straus, D., & Chang, E. B. (1997). Cell-specific induction of HSP72-mediated protection by glutamine against oxidant injury in IEC18 cells. Proceedings of the Association of American Physicians, 110(2), 136-139.

Nadel, E. R., Pandolf, K. B., Roberts, M. F., & Stolwijk, J. A. (1974). Mechanisms of thermal acclimation to exercise and heat. Journal of Applied Physiology, 37(4), 515-520.

Ng, Q. Y., Lee, K. W., Byrne, C., Ho, T. F., & Lim, C. L. (2008). Plasma endotoxin and immune responses during a 21-km road race under a warm and humid environment. Annals of the Academy of Medicine, Singapore, 37(4), 307-314.

Nichols, A. W. (2014). Heat-related illness in sports and exercise. Current Reviews in Musculoskeletal Medicine, 7(4), 355-365.

Nielsen, B. (1998). Heat acclimation - mechanisms of adaption to exercise in the heat. International Journal of Sports Medicine, 19(S 2), S154-S156.

Nielsen, B., Hales, J. R., Strange, S., Christensen, N. J., Warberg, J., & Saltin, B. (1993). Human circulatory and thermoregulatory adaptations with heat acclimation and exercise in a hot, dry environment. The Journal of physiology, 460, 467-485.

Nielsen, B., Strange, S., Christensen, N. J., Warberg, J., & Saltin, B. (1997). Acute and adaptive responses in humans to exercise in a warm, humid environment. Pflügers Archiv - European Journal of Physiology, 434(1), 49-56.

Nieman, D. C., Henson, D. A., Dumke, C. L., Oley, K., Mcanulty, S. R., Davis, J. M., . . . Morrow, J. D. (2006). Ibuprofen use, endotoxemia, inflammation, and plasma cytokines during ultramarathon competition. Brain, Behavior, and Immunity, 20(6), 578-584.

Nybo, L. (2008). Hyperthermia and fatigue. Journal of Applied Physiology, 104(3), 871-878.

Nybo, L., & González-Alonso, J. (2015). Critical core temperature: a hypothesis too simplistic to explain hyperthermia-induced fatigue. Scandinavian Journal of Medicine & Science in Sports, 25, 4-5.

198 Chapter 7: Bibliography

Nybo, L., & Nielsen, B. (2001a). Hyperthermia and central fatigue during prolonged exercise in humans. Journal of Applied Physiology, 91(3), 1055-1060.

Nybo, L., & Nielsen, B. (2001b). Perceived exertion is associated with an altered brain activity during exercise with progressive hyperthermia. Journal of Applied Physiology, 91(5), 2017-2023.

Nybo, L., Rasmussen, P., & Sawka, M. N. (2014). Performance in the heat- physiological factors of importance for hyperthermia-induced fatigue. Comprehensive Physiology, 4(2), 657-689.

Øktedalen, O., Lunde, O. C., Opstad, P. K., Aabakken, L., & Kvernebo, K. (1992). Changes in the gastrointestinal mucosa after long-distance running. Scandinavian Journal of Gastroenterology, 27(4), 270-274.

Oriordain, M. G., Debeaux, A., & Fearon, K. C. H. (1996). Effect of glutamine on immune function in the surgical patient. Nutrition, 12(11-12), S82-S84.

Osborne, J. O., Stewart, I. B., Beagley, K. W., & Minett, G. M. (2019). The effect of cycling in the heat on gastrointestinal-induced damage and neuromuscular fatigue. European Journal of Applied Physiology.

Otte, J. A., Oostveen, E., Geelkerken, R. H., Groeneveld, A. B. J., & Kolkman, J. J. (2001). Exercise induces gastric ischemia in healthy volunteers: a tonometry study. Journal of Applied Physiology, 91(2), 866-871.

Pals, K. L., Chang, R.-T., Ryan, A. J., & Gisolfi, C. V. (1997). Effect of running intensity on intestinal permeability. Journal of Applied Physiology, 82(2), 571- 576.

Pandolf, K. B. (1998). Time course of heat acclimation and its decay. International Journal of Sports Medicine, 19 Suppl 2, S157-S160.

Pandolf, K. B., Burse, R. L., & Goldman, R. F. (1977). Role of physical fitness in heat acclimatisation, decay and reinduction. Ergonomics, 20(4), 399-408.

Patterson, M. J., Stocks, J. M., & Taylor, N. a. S. (2004). Sustained and generalized extracellular fluid expansion following heat acclimation. The Journal of physiology, 559(1), 327-334.

Peake, J. M., Della Gatta, P., Suzuki, K., & Nieman, D. C. (2015). Cytokine expression and secretion by skeletal muscle cells: regulatory mechanisms and exercise effects. Exercise Immunology Review, 21, 8-25.

Pedersen, B. K., Steensberg, A., Fischer, C., Keller, C., Keller, P., Plomgaard, P., . . . Saltin, B. (2003). Searching for the exercise factor: is IL-6 a candidate? Journal of Muscle Research and Cell Motility, 24(2), 113.

Chapter 7: Bibliography 199

Peiffer, J. J., & Abbiss, C. R. (2011). Influence of environmental temperature on 40 km cycling time-trial performance. International Journal of Sports Physiology and Performance, 6(2), 208-220.

Périard, J. D. (2012). Influence of Hyperthermia on Aerobic Exercise Performance: Mechanisms of Fatigue. In X. C. Gao, HD (Ed.), Hyperthermia: Recognition, Prevention and Treatment. (pp. 85-113). New York: Nova Science Publishers Inc.

Périard, J. D., Caillaud, C., & Thompson, M. W. (2011). Central and peripheral fatigue during passive and exercise-induced hyperthermia. Medicine & Science in Sports & Exercise, 43(9), 1657-1665.

Périard, J. D., Christian, R. J., Knez, W. L., & Racinais, S. (2014). Voluntary muscle and motor cortical activation during progressive exercise and passively induced hyperthermia. Experimental Physiology, 99(1), 136-148.

Périard, J. D., Cramer, M. N., Chapman, P. G., Caillaud, C., & Thompson, M. W. (2011a). Cardiovascular strain impairs prolonged self-paced exercise in the heat. Experimental Physiology, 96(2), 134-144.

Périard, J. D., Cramer, M. N., Chapman, P. G., Caillaud, C., & Thompson, M. W. (2011b). Neuromuscular function following prolonged intense self-paced exercise in hot climatic conditions. European Journal of Applied Physiology, 111(8), 1561-1569.

Périard, J. D., De Pauw, K., Zanow, F., & Racinais, S. (2018). Cerebrocortical activity during self-paced exercise in temperate, hot and hypoxic conditions. Acta Physiologica, 222(1), e12916.

Périard, J. D., & Racinais, S. (2015). Self-paced exercise in hot and cool conditions is associated with the maintenance of %V̇ o2peak within a narrow range. Journal of Applied Physiology, 118(10), 1258-1265.

Périard, J. D., Racinais, S., & Sawka, M. N. (2015). Adaptations and mechanisms of human heat acclimation: applications for competitive athletes and sports. Scandinavian Journal of Medicine & Science in Sports, 25 Suppl 1, 20-38.

Petry, E. R., Cruzat, V. F., Heck, T. G., Homem De Bittencourt, P. I., Jr., & Tirapegui, J. (2015). L-glutamine supplementations enhance liver glutamine-glutathione axis and heat shock factor-1 expression in endurance-exercise trained rats. International Journal of Sport Nutrition and Exercise Metabolism, 25(2), 188- 197.

Pfeiffer, B., Cotterill, A., Grathwohl, D., Stellingwerff, T., & Jeukendrup, A. E. (2009). The effect of carbohydrate gels on gastrointestinal tolerance during a 16-km run. International Journal of Sport Nutrition and Exercise Metabolism, 19(5), 485-503.

Phillips, S. (2015). Fatigue in Sport and Exercise. New York, NY: Routledge.

200 Chapter 7: Bibliography

Pires, W., Veneroso, C. E., Wanner, S. P., Pacheco, D. a. S., Vaz, G. C., Amorim, F. T., . . . Coimbra, C. C. (2016). Association Between Exercise-Induced Hyperthermia and Intestinal Permeability: A Systematic Review. Sports Medicine, 1-15.

Preiser, J. C., & Wernerman, J. (2003). Glutamine, a life-saving nutrient, but why? Critical Care Medicine, 31(10), 2555-2556.

Pugh, J. N., Impey, S. G., Doran, D. A., Fleming, S. C., Morton, J. P., & Close, G. L. (2017). Acute high-intensity interval running increases markers of gastrointestinal damage and permeability but not gastrointestinal symptoms. Applied Physiology, Nutrition and Metabolism, 42(9), 941-947.

Pugh, J. N., Sage, S., Hutson, M., Doran, D. A., Fleming, S. C., Highton, J., . . . Close, G. L. (2017). Glutamine supplementation reduces markers of intestinal permeability during running in the heat in a dose-dependent manner. European Journal of Applied Physiology, 117(12), 2569-2577.

Pyne, D. B., Guy, J. H., & Edwards, A. M. (2014). Managing heat and immune stress in athletes with evidence-based strategies. International Journal of Sports Physiology and Performance, 9(5), 744-750.

Qamar, M. I., & Read, A. E. (1987). Effects of exercise on mesenteric blood flow in man. Gut, 28(5), 583-587.

Racinais, S., Alonso, J.-M., Coutts, A. J., Flouris, A. D., Girard, O., González-Alonso, J., . . . Périard, J. D. (2015). Consensus Recommendations on Training and Competing in the Heat. Sports Medicine, 45(7), 925-938.

Racinais, S., Buchheit, M., Bilsborough, J., Bourdon, P. C., Cordy, J., & Coutts, A. J. (2014). Physiological and performance responses to a training camp in the heat in professional Australian football players. International Journal of Sports Physiology and Performance, 9(4), 598-603.

Racinais, S., Mohr, M., Buchheit, M., Voss, S. C., Gaoua, N., Grantham, J., & Nybo, L. (2012). Individual responses to short-term heat acclimatisation as predictors of football performance in a hot, dry environment. British Journal of Sports Medicine, 46(11), 810-815.

Racinais, S., Périard, J. D., Karlsen, A., & Nybo, L. (2015). Effect of Heat and Heat Acclimatization on Cycling Time Trial Performance and Pacing. Medicine & Science in Sports & Exercise, 47(3), 601-606.

Racinais, S., Wilson, M. G., & Périard, J. D. (2017). Passive heat acclimation improves skeletal muscle contractility in humans. American Journal of Physiology - Regulatory, Integrative and Comparative Physiology, 312(1), R101-R107.

Chapter 7: Bibliography 201

Rao, R., & Samak, G. (2012). Role of Glutamine in Protection of Intestinal Epithelial Tight Junctions. Journal of Epithelial Biology & Pharmacology, 5(Suppl. 1- M7), 47-54.

Reid, M. B., Lännergren, J., & Westerblad, H. (2002). Respiratory and Limb Muscle Weakness Induced by Tumor Necrosis Factor-α. American Journal of Respiratory and Critical Care Medicine, 166(4), 479-484.

Robinson, S., Turrell, E. S., Belding, H. S., & Horvath, S. M. (1943). Rapid acclimitization to work in hot climates. American Journal of Physiology, 140(2), 168-176.

Robson-Ansley, P. J., De Milander, L., Collins, M., & Noakes, T. D. (2004). Acute interleukin-6 administration impairs athletic performance in healthy, trained male runners. Canadian Journal of Applied Physiology, 29, 411-418.

Robson-Ansley, P. J., Milander, L. D., Collins, M., & Noakes, T. D. (2004). Acute Interleukin-6 Administration Impairs Athletic Performance in Healthy, Trained Male Runners. Canadian Journal of Applied Physiology, 29(4), 411- 418.

Roelands, B., De Koning, J., Foster, C., Hettinga, F., & Meeusen, R. (2013). Neurophysiological Determinants of Theoretical Concepts and Mechanisms Involved in Pacing. Sports Medicine, 43(5), 301-311.

Roelands, B., De Pauw, K., & Meeusen, R. (2015). Neurophysiological effects of exercise in the heat. Scandinavian Journal of Medicine & Science in Sports, 25 Suppl 1, 65-78.

Roelands, B., & Meeusen, R. (2010). Alterations in Central Fatigue by Pharmacological Manipulations of Neurotransmitters in Normal and High Ambient Temperature. Sports Medicine, 40(3), 229-246.

Rohleder, N., Aringer, M., & Boentert, M. (2012). Role of interleukin-6 in stress, sleep, and fatigue. Annals of the New York Academy of Sciences, 1261(1), 88- 96.

Ropeleski, M. J., Riehm, J., Baer, K. A., Musch, M. W., & Chang, E. B. (2005). Anti- apoptotic Effects of L-Glutamine-Mediated Transcriptional Modulation of the Heat Shock Protein 72 During Heat Shock. Gastroenterology, 129(1), 170-184.

Rowell, L. B. (2011). Cardiovascular adjustments to thermal stress. In Comprehensive Physiology: John Wiley & Sons, Inc.

Rowell, L. B., Blackmon, J. R., Martin, R. H., Mazzarella, J. A., & Bruce, R. A. (1965). Hepatic clearance of indocyanine green in man under thermal and exercise stresses. Journal of Applied Physiology, 20(3), 384-394.

202 Chapter 7: Bibliography

Rowell, L. B., Brengelmann, G. L., Blackmon, J. R., Twiss, R. D., & Kusumi, F. (1968). Splanchnic blood flow and metabolism in heat-stressed man. Journal of Applied Physiology, 24(4), 475-484.

Rowell, L. B., Marx, H. J., Bruce, R. A., Conn, R. D., & Kusumi, F. (1966). Reductions in cardiac output, central blood volume, and stroke volume with thermal stress in normal men during exercise. The Journal of Clinical Investigation, 45(11), 1801-1816.

Ryan, A. J., Flanagan, S. W., Moseley, P. L., & Gisolfi, C. V. (1992). Acute heat stress protects rats against endotoxin shock. Journal of Applied Physiology, 73(4), 1517-1522.

Saat, M., Sirisinghe, R. G., Singh, R., & Tochihara, Y. (2005). Decay of heat acclimation during exercise in cold and exposure to cold environment. European Journal of Applied Physiology, 95(4), 313-320.

Saboisky, J., Marino, F. E., Kay, D., & Cannon, J. (2003). Exercise heat stress does not reduce central activation to non-exercised human skeletal muscle. Experimental Physiology, 88(6), 783-790.

Sainani, K. L. (2018). The Problem with "Magnitude-based Inference". Medicine & Science in Sports & Exercise, 50(10), 2166-2176.

Sakurada, S., & Hales, J. R. S. (1998). A role for gastrointestinal endotoxins in enhancement of heat tolerance by physical fitness. Journal of Applied Physiology, 84(1), 207-214.

Salgado, R. M., White, A. C., Schneider, S. M., & Mermier, C. M. (2014). A novel mechanism for cross-adaptation between heat and altitude acclimation: the role of heat shock protein 90. Physiology Journal, 2014, 12.

Sanders, M. M., & Kon, C. (1991). Glutamine is a powerful effector of heat shock protein expression in Drosophila Kc cells. Journal of Cellular Physiology, 146(1), 180-190.

Sawka, M. N., Leon, L. R., Montain, S. J., & Sonna, L. A. (2011). Integrated physiological mechanisms of exercise performance, adaptation, and maladaptation to heat stress. In Comprehensive Physiology: John Wiley & Sons, Inc.

Sawka, M. N., Pandolf, K. B., Avellini, B. A., & Shapiro, Y. (1983). Does heat acclimation lower the rate of metabolism elicited by muscular exercise? Aviation, Space, and Environmental Medicine, 54(1), 27-31.

Sawka, M. N., Wenger, C. B., & Pandolf, K. B. (2010). Thermoregulatory responses to acute exercise-heat stress and heat acclimation. In Comprehensive Physiology: John Wiley & Sons, Inc.

Chapter 7: Bibliography 203

Schlader, Z. J., Simmons, S. E., Stannard, S. R., & Mündel, T. (2011). Skin temperature as a thermal controller of exercise intensity. European Journal of Applied Physiology, 111(8), 1631-1639.

Schlader, Z. J., Stannard, S. R., & Mündel, T. (2011). Exercise and heat stress: performance, fatigue and exhaustion - a hot topic. British Journal of Sports Medicine, 45(1), 3-5.

Schlader, Z. J., & Vargas, N. T. (2019). Regulation of Body Temperature by Autonomic and Behavioral Thermoeffectors. Exercise and Sport Sciences Reviews, 47(2), 116-126.

Scoon, G. S., Hopkins, W. G., Mayhew, S., & Cotter, J. D. (2007). Effect of post- exercise sauna bathing on the endurance performance of competitive male runners. Journal of Science and Medicine in Sport, 10(4), 259-262.

Selkirk, G. A., Mclellan, T. M., Wright, H. E., & Rhind, S. G. (2008). Mild endotoxemia, NF-kappa B translocation, and cytokine increase during exertional heat stress in trained and untrained individuals. American Journal of Physiology - Regulatory, Integrative and Comparative Physiology, 295(2), R611-R623.

Senay, L. C., Mitchell, D., & Wyndham, C. H. (1976). Acclimatization in a hot, humid environment: body fluid adjustments. Journal of Applied Physiology, 40(5), 786-796.

Shapiro, Y., Moran, D., & Epstein, Y. (1998). Acclimatization Strategies - Preparing for Exercise in the Heat. International Journal of Sports Medicine, 19(S 2), S161-S163.

Sharma, H. S., & Hoopes, P. J. (2003). Hyperthermia induced pathophysiology of the central nervous system. International Journal of Hyperthermia, 19(3), 325- 354.

Shi, Y., Tu, Z., Tang, D., Zhang, H., Liu, M., Wang, K., . . . Xiao, X. (2006). The inhibition of LPS-induced production of inflammatory cytokines by HSP70 involves inactivation of the NF-kappaB pathway but not the MAPK pathways. Shock, 26(3), 277-284.

Shield, A., & Zhou, S. (2004). Assessing Voluntary Muscle Activation with the Twitch Interpolation Technique. Sports Medicine, 34(4), 253-267.

Shindoh, C., Dimarco, A., Nethery, D., & Supinski, G. (1992). Effect of PEG- Superoxide Dismutase on the Diaphragmatic Response to Endotoxin. American Review of Respiratory Disease, 145(6), 1350-1354.

Shing, C. M., Peake, J. M., Lim, C. L., Briskey, D., Walsh, N. P., Fortes, M. B., . . . Vitetta, L. (2014). Effects of probiotics supplementation on gastrointestinal permeability, inflammation and exercise performance in the heat. European Journal of Applied Physiology, 114(1), 93-103.

204 Chapter 7: Bibliography

Singleton, K. D., Serkova, N., Banerjee, A., Meng, X., Gamboni-Robertson, F., & Wischmeyer, P. E. (2005). Glutamine attenuates endotoxin-induced lung metabolic dysfunction: Potential role of enhanced heat shock protein 70. Nutrition, 21(2), 214-223.

Singleton, K. D., & Wischmeyer, P. E. (2006). Oral glutamine enhances heat shock protein expression and improves survival following hyperthermia. Shock, 25(3), 295-299.

Singleton, K. D., & Wischmeyer, P. E. (2007). Glutamine's protection against sepsis and lung injury is dependent on heat shock protein 70 expression. American Journal of Physiology - Regulatory, Integrative and Comparative Physiology, 292(5), R1839-R1845.

Skidmore, R., Gutierrez, J. A., Guerriero, V., & Kregel, K. C. (1995). HSP-70 induction during exercise and heat-stress in rats - Role of internal temperature. American Journal of Physiology - Regulatory Integrative and Comparative Physiology, 268(1), R92-R97.

Smetanka, R. D., Lambert, G. P., Murray, R., Eddy, D., Horn, M., & Gisolfi, C. V. (1999). Intestinal permeability in runners in the 1996 Chicago marathon. International Journal of Sport Nutrition, 9(4), 426-433.

Soares, A. D. N., Costa, K. A., Wanner, S. P., Santos, R. G. C., Fernandes, S. O. A., Martins, F. S., . . . Cardoso, V. N. (2014). Dietary glutamine prevents the loss of intestinal barrier function and attenuates the increase in core body temperature induced by acute heat exposure. British Journal of Nutrition, 112(10), 1601-1610.

Supinski, G., Nethery, D., Nosek, T. M., Callahan, L. A., Stofan, D., & Dimarco, A. (2000). Endotoxin administration alters the force vs. pCa relationship of skeletal muscle fibers. American Journal of Physiology-Regulatory, Integrative and Comparative Physiology, 278(4), R891-R896.

Supinski, G., Nethery, D., Stofan, D., & Dimarco, A. (1996). Comparison of the effects of endotoxin on limb, respiratory, and cardiac muscles. Journal of Applied Physiology, 81(3), 1370-1378.

Supinski, G. S., & Callahan, L. A. (2007). Free radical-mediated skeletal muscle dysfunction in inflammatory conditions. Journal of Applied Physiology, 102(5), 2056-2063.

Suzuki, K. (2018). Cytokine Response to Exercise and Its Modulation. Antioxidants, 7(1), 17.

Szelényi, J. (2001). Cytokines and the central nervous system. Brain Research Bulletin, 54(4), 329-338.

Chapter 7: Bibliography 205

Taga, T., & Kishimoto, T. (1997). gp130 and the interleukin-6 family of cytokines. Annual Review of Immunology, 15(1), 797-819.

Tanaka, T., Narazaki, M., & Kishimoto, T. IL-6 in inflammation, immunity, and disease. Cold Spring Harbor Perspectives in Biology, 6(10), a016295- a016295.

Tansey, E. A., & Johnson, C. D. (2015). Recent advances in thermoregulation. Advances in Physiology Education, 39(3), 139-148.

Tatterson, A. J., Hahn, A. G., Martini, D. T., & Febbraio, M. A. (2000). Effects of heat stress on physiological responses and exercise performance in elite cyclists. Journal of Science and Medicine in Sport, 3(2), 186-193.

Taylor, N. A. (2014). Human heat adaptation. Comprehensive Physiology, 4(1), 325- 365.

Thomas, M. M., Cheung, S. S., Elder, G. C., & Sleivert, G. G. (2006). Voluntary muscle activation is impaired by core temperature rather than local muscle temperature. Journal of Applied Physiology, 100(4), 1361-1369.

Thuijls, G., Derikx, J. P., De Haan, J. J., Grootjans, J., De Bruine, A., Masclee, A. A., . . . Buurman, W. A. (2010). Urine-based detection of intestinal tight junction loss. Journal of Clinical Gastroenterology, 44(1), e14-19.

Todd, G., Butler, J. E., Taylor, J. L., & Gandevia, S. C. (2005). Hyperthermia: a failure of the motor cortex and the muscle. The Journal of physiology, 563(2), 621- 631.

Tucker, R., Marle, T., Lambert, E. V., & Noakes, T. D. (2006). The rate of heat storage mediates an anticipatory reduction in exercise intensity during cycling at a fixed rating of perceived exertion. The Journal of physiology, 574(Pt 3), 905- 915.

Tucker, R., Rauch, L., Harley, Y. X., & Noakes, T. D. (2004a). Impaired exercise performance in the heat is associated with an anticipatory reduction in skeletal muscle recruitment. Pflugers Arch - Eur J Physiol, 448(4), 422-430.

Tucker, R., Rauch, L., Harley, Y. X. R., & Noakes, T. D. (2004b). Impaired exercise performance in the heat is associated with an anticipatory reduction in skeletal muscle recruitment. Pflugers Arch - Eur J Physiol, 448(4), 422-430.

Van Deventer, S. J., Buller, H. R., Ten Cate, J. W., Aarden, L. A., Hack, C. E., & Sturk, A. (1990). Experimental endotoxemia in humans: analysis of cytokine release and coagulation, fibrinolytic, and complement pathways. Blood, 76(12), 2520-2526.

Van Nieuwenhoven, M. A., Brouns, F., & Brummer, R.-J. M. (2004). Gastrointestinal profile of symptomatic athletes at rest and during physical exercise. European Journal of Applied Physiology, 91(4), 429-434.

206 Chapter 7: Bibliography

Van Wijck, K., Lenaerts, K., Grootjans, J., Wijnands, K. A., Poeze, M., Van Loon, L. J., . . . Buurman, W. A. (2012). Physiology and pathophysiology of splanchnic hypoperfusion and intestinal injury during exercise: strategies for evaluation and prevention. American Journal of Physiology - Gastrointestinal and Liver Physiology, 303(2), G155-G168.

Van Wijck, K., Lenaerts, K., Van Loon, L. J., Peters, W. H., Buurman, W. A., & Dejong, C. H. (2011). Exercise-induced splanchnic hypoperfusion results in gut dysfunction in healthy men. PloS One, 6(7), e22366.

Varatharaj, A., & Galea, I. (2017). The blood-brain barrier in systemic inflammation. Brain, Behavior, and Immunity, 60, 1-12.

Vargas, N., & Marino, F. (2016). Heat stress, gastrointestinal permeability and interleukin-6 signaling - Implications for exercise performance and fatigue. Temperature, 3(2), 240-251.

Vargas, N., & Marino, F. (2017a). Neuroinflammation, cortical activity, and fatiguing behaviour during self-paced exercise. Pflugers Arch - Eur J Physiol, 470(2), 413-426.

Vargas, N., & Marino, F. (2017b). Neuroinflammation, cortical activity, and fatiguing behaviour during self-paced exercise. Pflügers Archiv - European Journal of Physiology, 470(2), 413-426.

Vargas, N. T., & Marino, F. (2014). A neuroinflammatory model for acute fatigue during exercise. Sports Medicine, 44(11), 1479-1487.

Vezzani, A., & Viviani, B. (2015). Neuromodulatory properties of inflammatory cytokines and their impact on neuronal excitability. Neuropharmacology, 96, 70-82.

Vitkovic, L., Bockaert, J., & Jacque, C. (2000). “Inflammatory” cytokines: neuromodulators in normal brain? Journal of Neurochemistry, 74(2), 457-471.

Walters, T. J., Ryan, K. L., Tate, L. M., & Mason, P. A. (2000). Exercise in the heat is limited by a critical internal temperature. Journal of Applied Physiology, 89(2), 799-806.

Weller, A. S., Linnane, D. M., Jonkman, A. G., & Daanen, H. A. (2007). Quantification of the decay and re-induction of heat acclimation in dry-heat following 12 and 26 days without exposure to heat stress. European Journal of Applied Physiology, 102(1), 57-66.

Wendt, D., Van Loon, L. J. C., & Marken Lichtenbelt, W. D. (2007a). Thermoregulation during Exercise in the Heat. Sports Medicine, 37(8), 669- 682.

Chapter 7: Bibliography 207

Wendt, D., Van Loon, L. J. C., & Marken Lichtenbelt, W. D. (2007b). Thermoregulation during exercise in the heat: strategies for maintaining health and performance. Sports Medicine, 37(8), 669-682.

Williams, C. G., Wyndham, C. H., & Morrison, J. F. (1967). Rate of loss of acclimatization in summer and winter. Journal of Applied Physiology, 22(1), 21-26.

Willmott, A. G. B., Hayes, M., Waldock, K. a. M., Relf, R. L., Watkins, E. R., James, C. A., . . . Maxwell, N. S. (2017). Short-term heat acclimation prior to a multi- day desert ultra-marathon improves physiological and psychological responses without compromising immune status. Journal of Sports Sciences, 35(22), 2249-2256.

Windmueller, H. G., & Spaeth, A. E. (1990). Uptake and metabolism of plasma glutamine by the small intestine. Nutrition Reviews, 48(8), 310-312.

Wingfield, G. L., Gale, R., Minett, G. M., Marino, F. E., & Skein, M. (2016). The effect of high versus low intensity heat acclimation on performance and neuromuscular responses. Journal of Thermal Biology, 58, 50-59.

Wischmeyer, P. E. (2006). The glutamine story: Where are we now? Current Opinion in Critical Care, 12(2), 142-148.

Wischmeyer, P. E. (2008). Glutamine: role in critical illness and ongoing clinical trials. Current Opinion in Gastroenterology, 24(2), 190-197.

Wischmeyer, P. E., Kahana, M., Wolfson, R., Ren, H., Musch, M. M., & Chang, E. B. (2001). Glutamine induces heat shock protein and protects against endotoxin shock in the rat. Journal of Applied Physiology, 90(6), 2403-2410.

Wischmeyer, P. E., Musch, M. W., Madonna, M. B., Thisted, R., & Chang, E. B. (1997). Glutamine protects intestinal epithelial cells: role of inducible HSP70. American Journal of Physiology - Gastrointestinal and Liver Physiology, 272(4), G879-G884.

Wischmeyer, P. E., Riehm, J., Singleton, K. D., Ren, H., Musch, M. W., Kahana, M., & Chang, E. B. (2003). Glutamine attenuates tumor necrosis factor-α release and enhances heat shock protein 72 in human peripheral blood mononuclear cells. Nutrition, 19(1), 1-6.

Wyndham, C. H., Rogers, G. G., Senay, L. C., & Mitchell, D. (1976). Acclimization in a hot, humid environment: cardiovascular adjustments. Journal of Applied Physiology, 40(5), 779-785.

Xue, H., Slavov, D., & Wischmeyer, P. E. (2012). Glutamine-mediated dual regulation of heat shock transcription factor-1 activation and expression. The Journal of Biological Chemistry, 287(48), 40400-40413.

208 Chapter 7: Bibliography

Yang, W., & Hu, P. (2018). Skeletal muscle regeneration is modulated by inflammation. Journal of Orthopaedic Translation, 13, 25-32.

Yeh, Y. J., Law, L. Y. L., & Lim, C. L. (2013). Gastrointestinal response and endotoxemia during intense exercise in hot and cool environments. European Journal of Applied Physiology, 113(6), 1575-1583.

Young, A. J., Sawka, M. N., Epstein, Y., Decristofano, B., & Pandolf, K. B. (1987). Cooling different body surfaces during upper and lower body exercise. Journal of Applied Physiology, 63(3), 1218-1223.

Zhang, J.-M., & An, J. (2007). Cytokines, inflammation, and pain. International Anesthesiology Clinics, 45(2), 27-37.

Zhou, Y.-P., Jiang, Z.-M., Sun, Y.-H., Wang, X.-R., Ma, E.-L., & Wilmore, D. (2003). The effect of supplemental enteral glutamine on plasma levels, gut function, and outcome in severe burns: a randomized, double-blind, controlled clinical trial. Journal of Parenteral and Enteral Nutrition, 27(4), 241-245.

Ziegler, T. R., Benfell, K., Smith, R. J., Young, L. S., Brown, E., Ferrari-Baliviera, E., . . . Wilmore, D. W. (1990). Safety and Metabolic Effects of L-Glutamine Administration in Humans. Journal of Parenteral and Enteral Nutrition, 14(4_suppl), 137S-146S.

Zuhl, M., Dokladny, K., Mermier, C., Schneider, S., Salgado, R., & Moseley, P. (2015). The effects of acute oral glutamine supplementation on exercise- induced gastrointestinal permeability and heat shock protein expression in peripheral blood mononuclear cells. Cell Stress and Chaperones, 20(1), 85-93.

Zuhl, M. N., Lanphere, K. R., Kravitz, L., Mermier, C. M., Schneider, S., Dokladny, K., & Moseley, P. L. (2014). Effects of oral glutamine supplementation on exercise-induced gastrointestinal permeability and tight junction protein expression. Journal of Applied Physiology, 116(2), 183-191.

Zurawlew, M. J., Walsh, N. P., Fortes, M. B., & Potter, C. (2016). Post-exercise hot water immersion induces heat acclimation and improves endurance exercise performance in the heat. Scandinavian Journal of Medicine & Science in Sports, 26(7), 745-754.

Chapter 7: Bibliography 209

Chapter 8: Appendices

8.1 APPENDIX A - TIMELINE

PHD Timeline 2015 2016 2017 2018 Time elapsed (months) 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46

Research Milestones Stage 2 Document Confirmation Annual Progress Report Final Seminar (30/11/2018)

Lodgement (Early 2019) Coursework AIRS (IFN001) Research Procedure Review literature

210 Chapter 8: Appendices

Design studies Source materials and equipment Conduct studies: Study 1 Study 2 Study 3 Data Collection Statistical Analysis

Applications

Ethics Intellectual Property Health and Safety Scholarships and grants Write-up grant Thesis

Title and introduction Literature review

Chapter 8: Appendices 211

Methodology Results and Discussion Conclusion Publication

Present at conferences

Submit to journals

212 Chapter 8: Appendices

8.2 APPENDIX B – NEUROMUSCULAR VARIABLES AND DEFINITIONS

8.2.1 Evoked twitch contractile properties

Peak twitch torque (Pt): peak torque during the evoked twitch

Time to peak torque (TPt): time from the first rise in torque above baseline to peak torque

Half relaxation time (½ RT): time taken for torque to reduce by half of the peak torque value

Contraction duration (CD): time to peak torque plus half relaxation time

Rate of torque development (RTD): slope of twitch-torque curve from onset to peak torque

Rate of relaxation (RR): slope of twitch-torque curve peak torque to half relaxation time.

a) PT; b) TPt; c) RTD; d) ½ RT; e) RR; f) CD

Figure reprinted from “Reproducibility and changes in twitch properties associated with age and resistance training in young and elderly women,” by J. Cannon, D. Kay, K.M. Tarpenning & F.E. Marino, 2008, Scandinavian Journal of Medicine and Science in Sports.

Chapter 8: Appendices 213

8.3 APPENDIX C – PUBLICATIONS AND CONFERENCE ABSTRACTS

2019 Osborne, J.O., Stewart, I.B., Beagley, K.W., Borg, D.N., & Minett, G.M. 2019 ASICS Sports Medicine Australia (SMA) Conference Sunshine Coast, Australia Short-term heat acclimation training improves cycling performance in the heat and enhances knee extensor strength. (presentation)

2019 Osborne, J.O., Stewart, I.B., Beagley, K.W., & Minett, G.M. The effect of cycling in the heat on gastrointestinal-induced damage and neuromuscular fatigue. European Journal of Applied Physiology, 119(8), 1829-1840.

2018 Borg, D. N., Osborne, J. O., Stewart, I. B., Costello, J. T., Sims, J. N., & Minett, G. M. The reproducibility of 10 and 20 km time trial cycling performance in recreational cyclists, runners and team sport athletes. Journal of Science & Medicine in Sport, 21(8), 858-863.

2018 Osborne, J.O., Stewart, I.B., Beagley, K.W., & Minett, G.M. Institute of Health and Biomedical Innovation (IHBI) Inspires 2018 Brisbane, Australia Can glutamine supplementation improve 20-km cycling time trial performance in the heat? (poster)

2018 Osborne, J.O., Stewart, I.B., Beagley, K.W., & Minett, G.M. Exercise & Sports Science Australia (ESSA) – Research to Practice Brisbane, Australia Exercise heat stress induces gastrointestinal damage and impairs neuromuscular performance. (poster)

2017 Osborne, J.O., Stewart, I.B., Beagley, K.W., & Minett, G.M. 17th International Conference on Environmental Ergonomics (ICEE) Kobe, Japan

214 Chapter 8: Appendices

The effects of cycling in the heat on gastrointestinal inflammation and neuromuscular performance. (presentation)

2017 Osborne, J.O., Stewart, I.B., Beagley, K.W., & Minett, G.M. Institute of Health and Biomedical Innovation (IHBI) Inspires 2017 Brisbane, Australia Gastrointestinal permeability, inflammation and central fatigue during exercise in the heat. (presentation)

Chapter 8: Appendices 215

8.4 APPENDIX D – SCALES & DATASHEETS (STUDY 1)

8.4.1 Borg’s Rating of Perceived Exertion (RPE) scale (Borg, 1970) BORG RATING OF PERCIVED EXERTION

6 No exertion 7 Extremely light 8

9 Very light 10 11 Light 12 13 Somewhat hard 14 15 Hard 16 17 Very hard 18 19 Extremely hard 20 Maximal exertion

216 Chapter 8: Appendices

8.4.2 Thermal comfort scale (Gagge et al., 1967) THERMAL COMFORT SCALE

1 Comfortable

2 Slightly uncomfortable

3 Uncomfortable

4 Very uncomfortable

Chapter 8: Appendices 217

8.4.3 Thermal sensation scale (Young et al., 1987) THERMAL SENSATION SCALE

0.0 Unbearably cold 0.5 1.0 Very cold 1.5 2.0 Cold 2.5 3.0 Cool 3.5 4.0 Comfortable 4.5 5.0 Warm 5.5 6.0 Hot 6.5 7.0 Very hot 7.5 8.0 Unbearably hot

218 Chapter 8: Appendices

8.4.4 Adult Pre-exercise Screening Tool (ESSA)

Chapter 8: Appendices 219

8.4.5 Study 1 Setup Sheet (Dyno and Lode)

220 Chapter 8: Appendices

8.4.6 24 h Food & 48 h Physical Activity Diaries (Study 1)

Chapter 8: Appendices 221

8.4.7 GI and Medical Screening Form (Study 1)

222 Chapter 8: Appendices

8.5 APPENDIX E – SCALES & DATASHEETS (STUDY 2)

8.5.1 Borg’s Rating of Perceived Exertion (RPE) scale (Borg, 1970) BORG RATING OF PERCIVED EXERTION

6 7 Very, very light 8

9 Very light 10 11 Light 12 13 Somewhat hard 14 15 Hard 16 17 Very hard 18 19 Very, very hard 20

Chapter 8: Appendices 223

8.5.2 Thermal comfort scale - modified (Gagge et al., 1967)

THERMAL COMFORT SCALE

1.0 Comfortable

1.5

2.0 Slightly uncomfortable

2.5

3.0 Uncomfortable

3.5

4.0 Very uncomfortable

4.5

5.0 Extremely uncomfortable

224 Chapter 8: Appendices

8.5.3 Thermal sensation scale (Young et al., 1987) THERMAL SENSATION SCALE

0.0 Unbearably cold 0.5 1.0 Very cold 1.5 2.0 Cold 2.5 3.0 Cool 3.5 4.0 Comfortable 4.5 5.0 Warm 5.5 6.0 Hot 6.5 7.0 Very hot 7.5

8.0 Unbearably hot

Chapter 8: Appendices 225

8.5.4 Borg’s Sessional Rating of Perceived Exertion (sRPE) scale (Borg, 1998) Borg’s Sessional Rating of Perceived Exertion (sRPE) Scale

0 Nothing at all

0.5 Very, very light

1 Very light

2 Fairly light

3 Moderate

4 Somewhat strong

5 Strong

6

7 Very strong

8

9

10 Extremely strong (almost maximal)

226 Chapter 8: Appendices

8.5.5 Abbreviated POMS (Grove & Prapavessis, 1992)

Chapter 8: Appendices 227

8.5.5 Abbreviated POMS - continued (Grove & Prapavessis, 1992)

228 Chapter 8: Appendices

8.5.6 24 h Food & 48 h Physical Activity Diaries (Study 2)

Chapter 8: Appendices 229

8.5.7 Post-trial sheet for GI distress, sRPE and intervention guess (Study 2)

230 Chapter 8: Appendices

8.5.8 Adult Pre-exercise Screening Tool (ESSA)

Chapter 8: Appendices 231

8.5.9 Study 2 Setup Sheet (Dyno and Velotron)

232 Chapter 8: Appendices

8.5.10 GI and Medical Screening Form (Study 2)

Chapter 8: Appendices 233

8.6 APPENDIX F – SCALES & DATASHEETS (STUDY 3)

8.6.1 Borg’s Rating of Perceived Exertion (RPE) scale (Borg, 1998) BORG RATING OF PERCIVED EXERTION

6 7 Very, very light 8

9 Very light 10 11 Light 12 13 Somewhat hard 14 15 Hard 16 17 Very hard 18 19 Very, very hard 20

234 Chapter 8: Appendices

8.6.2 Thermal comfort scale - modified (Gagge et al., 1967)

THERMAL COMFORT SCALE

1.0 Comfortable

1.5

2.0 Slightly uncomfortable

2.5

3.0 Uncomfortable

3.5

4.0 Very uncomfortable

4.5

5.0 Extremely uncomfortable

Chapter 8: Appendices 235

8.6.3 Thermal sensation scale (Young et al., 1987) THERMAL SENSATION SCALE

0.0 Unbearably cold 0.5 1.0 Very cold 1.5 2.0 Cold 2.5 3.0 Cool 3.5 4.0 Comfortable 4.5 5.0 Warm 5.5 6.0 Hot 6.5 7.0 Very hot 7.5

8.0 Unbearably hot

236 Chapter 8: Appendices

8.6.4 Borg’s Sessional Rating of Perceived Exertion (sRPE) scale (Borg, 1998) Borg’s Sessional Rating of Perceived Exertion (sRPE) Scale

0 Nothing at all

0.5 Very, very light

1 Very light

2 Fairly light

3 Moderate

4 Somewhat strong

5 Strong

6

7 Very strong

8

9

10 Extremely strong (almost maximal)

Chapter 8: Appendices 237

8.6.5 Psychological wellbeing sheet – Study 3 (McLean et al., 2010)

238 Chapter 8: Appendices

8.6.6 24 h Food & 48 h Physical Activity Diaries (Study 3)

Chapter 8: Appendices 239

8.6.7 GI and Medical Screening Form (Study 3)

GASTROINTESTINAL AND MEDICAL HISTORY SCREENING FORM YES NO COMMENTS 1. Have you ever been diagnosed with, or suspected of having, an obstructive disease of the gastrointestinal tract? Including diverticulitis and/or inflammatory bowel disease? 2. Have you ever had any previous medical issues arising from blood collection?

3. Do you have a diagnosed blood clotting

disorder or severe phobia of needles?

4. Do you have a history of

gastrointestinal surgery? 5. Do you have a heart conduction defect(s) treated with an implantable pacemaker, defibrillator or any other active implantable device such as a nerve stimulator and/or medication? 6. Do you smoke cigarettes regularly? Or have quit smoking in the last 6 months?

Or are regularly exposed to environmental tobacco smoke? 7. Do you currently have any injury or illness that could be exacerbated by exercise?

TRAINING STATUS SCREENING FORM YES NO COMMENTS

1. Do you train and/or compete, on

average, at least twice a week?

2. Are you aged between 18 and 44?

3. Would you consider yourself a

moderate- to well-trained athlete?

240 Chapter 8: Appendices

8.6.8 Adult Pre-exercise Screening Tool (ESSA)

Chapter 8: Appendices 241

8.6.9 Study 3 Setup Sheet (Dyno, Velotron and Wattbike)

242 Chapter 8: Appendices

8.6.10 Post-trial sheet for GI distress and sRPE (Study 3)

Chapter 8: Appendices 243

8.7 APPENDIX G – CONSENT FORMS

8.7.1 Study 1 Consent Form

244 Chapter 8: Appendices

8.7.2 Study 2 Consent Form

Chapter 8: Appendices 245

8.7.3 Study 3 Consent Form

246 Chapter 8: Appendices