THE EFFECTS OF ENVIRONMENTAL AND PHYSICAL

STRESS ON ENERGY EXPENDITURE, ENERGY INTAKE,

AND APPETITE

by

Iva Mandic

A thesis submitted in conformity with the requirements for the degree of Doctor of Philosophy Graduate Department of Exercise Science University of Toronto

© Copyright by Iva Mandic (2018)

The effects of environmental and physical stress on energy expenditure, energy intake, and appetite

Iva Mandic

Doctor of Philosophy

The Graduate Department of Exercise Sciences University of Toronto

2018

Abstract

Body weight loss occurs frequently in military personnel engaged in field operations.

When this weight loss is rapid, or extensive it is associated with health and performance decrements. While the nature of military work does not allow for energy expenditure (EE) to be freely altered, energy intake (EI) can be increased to match EE and prevent weight loss.

Therefore a primary objective of the current dissertation was to develop a physiological and empirical basis to facilitate informed estimates of the EI that would be required to offset the

EE demand of military tasks during field operations. Three different approaches were undertaken: 1) The energy costs of 46 infantry tasks were measured; the results should reduce the dependency on less accurate predictions. 2) The impact of ambient temperature on EE of the tasks was minimal (~3%) when the ambient temperature was between -10°C and 30°C. This indicates that caloric supplementation of field rations on account of temperature is likely unnecessary during short-term operations occurring within this temperature range; and, lastly 3)

A simple algorithm based on accelerometry and heart rate was developed to assess EE in the field. The application of this algorithm should improve EE/EI matching.

ii

Unfortunately military personnel engaged in arduous field operations usually experience an energy deficit, even when food availability is adequate. Voluntary anorexia can ultimately thwart nutrition optimization in the field, therefore the role of appetite was also explored. While hormonal responses pointed towards appetite suppression with increased physical activity levels

(with a partial blunting of that response in the cold) and subjective appetite was the lowest in the heat and highest in the cold, actual EI was unchanged regardless of ambient temperature or whether the participant was sedentary or active.

This research demonstrated that even in the most favourable scenarios military personnel engaged in typical infantry tasks may under eat and plunge into a negative energy balance.

These results suggest that factors other than food availability and palatability, such as policy or procedural changes should be considered in addressing voluntary anorexia in the field in military personnel.

iii

Acknowledgements

First and foremost, I would like to thank my supervisor Dr. Ira Jacobs, for giving me the opportunity to work on such an interesting and involved research contract. Ira, thank you for being confident in my abilities even when I wasn’t, your continued trust and support have truly shaped the scientist I am today. Thank you for always challenging me and for knowing exactly how much guidance to provide. I have learned so much from you professionally and personally that I could not imagine having a better mentor throughout my PhD than you.

I am also grateful for my co-supervisor Dr. Catherine Amara, who was incredibly understanding and supportive when the opportunity to work with the Canadian Armed Forces came. Cathy, I truly enjoyed working with you at the beginning of my PhD and really appreciate the opportunities you provided me with. I always felt like your door was open to me, and I’m very thankful for all of your support and guidance. I’m also grateful for your many thought- provoking questions that forced me to consider a different angle and ultimately broadened my field of knowledge.

I would like to give special thanks to Dr. Mary L’Abbé and Dr. Greg Wells for serving on my committee and for providing me with useful feedback every step of the way. Mary, thank you for reviewing numerous papers, proposals, and ethics submissions, I really value your perspective and feel that your input, and attention to detail focused my work. Greg, thank you for asking me difficult, theoretical questions and pushing me outside my comfort zone. I was very fortunate to have a committee that challenged me and was so dedicated to my success.

To my examiners, Dr. Harris Lieberman and Dr. Karl Zabjek, thank you for your insightful comments and for your genuine interest in my dissertation. Dr. Lieberman, I have read much of your work, and it’s incredibly humbling to receive such high praise from someone like you.

I would also like to thank Dr. Paul Corey for helping me gain a deeper understanding of different statistical methods which ultimately allowed me to develop the energy expenditure algorithm. Paul, I’m thankful for the time and energy you spent helping me with my rather complicated dataset.

iv

To my wonderful collaborator Mavra Ahmed, I feel so fortunate to have had you by my side throughout our long and laborious data collection process. Whether we were working >12-h days, or waddling around in enormous snowsuits in Meaford, I always felt better knowing that you’d be there with me. Thank you for being a great colleague, and an even better friend.

I would like to acknowledge the Canadian Armed Forces, and in particular the many individuals who were always on hand to explain military jargon, assist with data collection, or participate as research subjects. I sincerely appreciate all of the time and effort you put into this project. In addition, I would also like to acknowledge the valuable contribution of Defence Research and Development Canada staff. Thank you for collaborating with me, for supporting this project, and for seeing it through. Specifically, I’d like to acknowledge Dr. Len Goodman and Dr. Shawn Rhind for their continued support. Len, thank you for jumping in whenever we were short staffed, I’ve enjoyed our many conversations, and appreciate your commitment to this research. Dr. Shawn Rhind, thank you for your selfless dedication to my success, you were under no obligation to help me, yet you did, tremendously. Thank you for your insightful suggestions, and for providing me with whatever I needed to analyze my blood samples.

To my friends and family, thank you for your love and understanding when I would disappear for months at a time to work on this dissertation. I also truly appreciate your patience and your avoidance of that joyless question “when will you be done?”

Finally, I would like to express my deepest gratitude to my husband Amerigo. Thank you for providing me with the means to analyze my millions of rows of data, I don’t know how I would’ve done it without you. Thank you for taking care of me when I was too busy to take care of myself, for building me up when I felt defeated, for teaching me how to relax, and for your constant love and support throughout this whole journey. No words can express how grateful I am to have you by my side.

Funding Acknowledgement

This research was funded by a research contract W7719-125107/001/TOR awarded to Dr. Ira Jacobs at the University of Toronto by the Canadian Department of National Defence. The research conducted was done in collaboration with Defence Research and Development Canada, and Mavra Ahmed, and Mary L’Abbé from the Department of Nutritional Sciences at the University of Toronto.

v

Table of Contents

Abstract ...... ii

Acknowledgements ...... iv

List of Abbreviations ...... xii

List of Tables ...... xiii

List of Secondary Tables ...... xv

List of Figures ...... xvi

List of Appendices ...... xviii

Chapter One: Introduction ...... 1

1.1 Overview of Dissertation ...... 1

Chapter Two: Literature Review ...... 3

2.1 Energy Balance ...... 3

2.1.1Physiological and Psychological Effects of Weight Loss...... 4

2.1.2 Typical Weight Loss in Military Personnel ...... 6

2.2 The Canadian Combat Ration Program ...... 7

2.2.1Energy content of supplied food ...... 9

2.2.2 Incremental Allowances (IA) ...... 11

2.3 Factors Affecting Energy Expenditure ...... 11

2.3.1 Energy Requirements of Military Personnel...... 14

2.3.2 Energy Cost of Environmental Temperature ...... 17

2.4 Methods of Estimating Energy Expenditure ...... 24

2.4.1 Combined Methods to Estimate Energy Expenditure in the Military Population ...... 29

2.5 Are military personnel eating what is provided to them? ...... 30

2.5.1 Why does Voluntary Anorexia occur? ...... 32

2.5.2 Appetite ...... 33

2.5.3 Hormonal Appetite Regulation ...... 37

vi

2.5.4 Effects of Exercise on Appetite ...... 42

2.5.5 Effects of Ambient Temperature on Appetite ...... 47

2.5.6 Effects of Ambient Temperature and Exercise on Appetite ...... 47

2.6 Summary of Literature Review ...... 54

Chapter Three: Dissertation Objectives and Hypotheses...... 58

3.1 Energy Cost of Infantry Tasks ...... 58

3.2 Ambient Temperature and Energy Expenditure ...... 58

3.3 Military–Specific Algorithm ...... 59

3.4 Appetite ...... 59

3.4.1 Energy Intake ...... 60

3.4.2 Subjective Appetite ...... 60

3.4.3 Appetite Regulating Hormones...... 60

Chapter Four: General Methods...... 61

4.1 Research Ethics Board approvals...... 61

4.2 Sample size calculations ...... 61

4.2.1 Study 1 2 and 3 Energy Cost of Activities and Differences Between Conditions...... 61

4.2.2 Study 4 - Appetite ...... 61

4.3 Participants ...... 62

4.4 Study Design ...... 63

4.4.1 Study Overview ...... 63

4.4.2 Visit 1: Questionnaire Completion ...... 64

4.4.3 Visit 2: Maximal Aerobic Power (V̇ O2max) ...... 64

4.4.4 Visit 3: Body Composition and Baseline Data Collection: ...... 65

4.4.5 Visits 4-11: Experimental Trials ...... 65

4.5 Measurements: ...... 70

4.5.1 Basic Measurements ...... 70

vii

4.5.2 Energy Expenditure and Activity Measurements ...... 72

4.5.3 Appetite Measurements ...... 73

4.6 Blood Sampling: ...... 74

4.6.1 Blood sample timing ...... 74

4.6.2 Collection procedures ...... 74

4.6.3 Blood Analysis ...... 75

4.7 Statistical Analysis ...... 76

Chapter Five: Study #1 - The Energy Cost of Various Infantry Tasks ...... 79

5.1 Abstract ...... 80

5.2 Introduction ...... 81

5.3 Methods...... 83

5.3.1 Participants ...... 83

5.3.2 Experimental Design ...... 83

5.3.3 Indirect Calorimetry ...... 85

5.3.4 Data Reduction...... 86

5.3.5 Statistical Analysis ...... 87

5.4 Results ...... 88

5.4.1 Treadmill Activities ...... 88

5.4.2 Non-Treadmill Activities ...... 92

5.5 Discussion ...... 94

5.6 Acknowledgements…………………………………………………………………………………………….……. 94

5.7 Conflict of Interest…………………………………………………………………………………………………... 94

5.8 References ...... 98

Chapter Six: Study #2 - The Effect of Ambient Temperature on the Energy Cost of Infantry Activities ...... 99

6.1 Abstract ...... 100

viii

6.2 Introduction ...... 101

6.3 Methods...... 103

6.3.1 Participants ...... 103

6.3.2 Experimental Design ...... 103

6.3.3 Statistical Analysis ...... 107

6.4 Results ...... 108

6.4.1 Missing Data ...... 108

6.4.2 Chamber Temperature...... 108

6.4.3 Clothing Weight ...... 108

6.4.4 Core Temperature ...... 109

6.4.5 Thermal Comfort ...... 110

6.4.6 Heart Rate ...... 112

6.4.7 Rate of Perceived Exertion ...... 113

6.4.8 Speed ...... 114

6.4.9 Repetitions ...... 114

6.4.10 Energy Expenditure by Time ...... 114

6.4.11 Total Energy Expenditure ...... 115

6.4.12 Energy Expenditure by Activity ...... 116

6.5 Discussion ...... 119

6.6 References ...... 124

Chapter Seven: Study #3 - Estimating Energy Expenditure in Military Personnel Using Accelerometry and Heart Rate ...... 126

7.1 Abstract ...... 127

7.2 Introduction ...... 128

7.3 Methods...... 130

7.3.1 Participants ...... 130

ix

7.3.2 Experimental Design ...... 130

7.3.3 Analysis...... 133

7.4 Results ...... 137

7.4.1 Participants ...... 137

7.4.2 Correlations ...... 137

7.4.3 Prediction Models ...... 138

7.4.4 Validation ...... 141

7.5 Discussion ...... 148

7.6 References ...... 153

Chapter Eight: Study #4 - The Effects of Exercise and Ambient Temperature on Appetite and Energy Intake ...... 156

8.1 Abstract ...... 158

8.2 Introduction ...... 159

8.3 Methods...... 161

8.3.1 Participants ...... 161

8.3.2 Experimental Design ...... 161

8.3.3 Measurements ...... 164

8.3.4 Blood Sampling ...... 166

8.3.5 Statistical Analyses ...... 167

8.4 Results ...... 168

8.4.1 Energy Expenditure (EE) ...... 168

8.4.2 Energy Intake (EI) and Relative Energy Intake (REI) ...... 168

8.4.3 Dietary Consumption ...... 169

8.4.4 Visual Analogue Scale for Appetite (VASA) ...... 171

8.4.5 Appetite Hormones ...... 174

8.4.6 Correlations between appetite variables ...... 177

x

8.5 Discussion ...... 179

8.6 References ...... 185

Chapter Nine: General Discussion ...... 188

9.1 Summary of Findings ...... 188

9.2 A Physiological Guide to Ensure the Provision of Sufficient Caloric Content during Field Operations ...... 189

9.2.1 A Catalogue of Energy Costs of Infantry Tasks ...... 190

9.2.2 Ambient Temperature ...... 191

9.2.3 A Military-Specific Algorithm ...... 194

9.3 An Investigation into the Under Consumption of Military Rations ...... 195

9.3.1 The Impact of Physical Activity and Ambient Temperature on Energy Intake ...... 195

9.3.2 The Impact of Physical Activity and Ambient Temperature on Subjective Appetite ...... 196

9.3.3 The Impact of Physical Activity and Ambient Temperature on Appetite Regulating Hormones ...... 197

9.4 Limitations ...... 200

9.5 Recommendations for Future Research ...... 203

9.6 Significance and Practical Applications ...... 205

9.7 Conclusions ...... 207

Appendices ...... 221

xi

List of Abbreviations AgRP Agouti gene-related peptide METs Metabolic equivalents ARC Arcuate nucleus MRE Meal, Ready-to-Eat (USA) AEBSF 4-(2-Aminoethyl) MSD Meso Scale Diagnostics benzenesulfonyl fluoride NPY Neuropeptide Y hydrochloride NTS Nucleus tractus solitarius ANOVA Analysis of variance OXM Oxyntomodulin AUC Area under the curve PAL Physical activity level BMR Basal metabolic rate PAR-Q+ Physical Activity Readiness CAF Canadian Armed Forces Questionnaire-Plus CART Cocaine- and amphetamine- POMC Pro-opiomelanocortin regulated transcript POMS Profile of Mood States CCK Cholecystokinin PP Pancreatic polypeptide CHO Carbohydrate PRO Protein DLW Doubly labelled water PSQI Pittsburgh Sleep Quality DPP-IV Dipeptidyl peptidase IV Index DRDC Defence Research and PVN Paraventricular nucleus Development Canada PYY Peptide tyrosine tyrosine or DRI Dietary Reference Intakes peptide YY EE Energy expenditure RDA Recommended Dietary EER Estimated Energy Allowance Requirements REB University of Toronto EI Energy intake Research Ethics Board EPOC Excess post-exercise oxygen REE Resting energy expenditure consumption REI Relative energy intake GABA Gamma-aminobutyric acid RH Relative humidity GI Gastrointestinal SEE Standard error of the estimate GLP-1 Glucagon-like peptide-1 SNAQ Simplified Nutritional GSQS Groningen Sleep Quality Appetite Questionnaire Scale TEE Total energy expenditure HR Heart rate TEF Thermic effect of food HREC DRDC Toronto Human UGR Unitized Group Ration Research Ethics Committee (USA) IA Incremental Allowances VASA Visual analogue scale for IMP Individual Meal Package appetite IOM Institute of Medicine V̇ CO2 Carbon dioxide production Kcal Kilocalorie V̇ O2 Oxygen consumption LMC Light Meal Combat V̇ O2max Maximal oxygen MDRI Military Dietary Reference consumption Intake VM Vector magnitude

xii

List of Tables Table 1: Average anthropometric data for men and women in the CAF ...... 10 Table 2: Energy expenditure in temperate environments as measured by the doubly labelled water technique ...... 15 Table 3: Energy expenditure in hot environments as measured by the doubly labelled water technique ...... 22 Table 4: Energy expenditure in cold environments as measured by the doubly labelled water technique ...... 23 Table 5: Determining energy expenditure - strengths and weaknesses of current methods ...... 27 Table 6: Appetite hormone summary table...... 42 Table 7: Effects of exercise and ambient temperature on appetite regulating hormones...... 50 Table 8: Participant characteristics ...... 62 Table 9: Battery of standardized infantry tasks that were completed in the environmental chamber...... 68 Table 10: Maximal aerobic power treadmill protocol adapted from [237]...... 70 Table 11: Participant characteristics...... 83 Table 12: Battery of standardized infantry tasks that were performed by all participants in the environmental chamber and the approximate times that these activities were done...... 85 Table 13: ‘Calculated METs’ from the treadmill activities compared to ‘Standard METs’ from similar activities found in the Compendium of Physical Activities...... 89 Table 14: ‘Calculated METs’ from the treadmill activities compared to ‘Calculated METs’ from similar activities found in military studies...... 91 Table 15:‘Calculated METs’ from non-treadmill activities compared to ‘Standard METs’ from similar activities found in the Compendium of Physical Activities...... 93 Table 16: Participant characteristics ...... 103 Table 17: Battery of standardized infantry tasks that were performed by all participants in the environmental chamber and the approximate times that these activities were done...... 106 Table 18: Temperature and humidity attained during the trial conditions...... 108 Table 19: The average thermal comfort score reported for each condition...... 110 Table 20: Battery of standardized infantry tasks that were performed by all participants in the environmental chamber and the approximate times that these activities were done...... 133 Table 21: Participant characteristics...... 137 Table 22: Correlations between variables of interest ...... 138 Table 23: Measured 4 h energy expenditure (kcal) vs. predicted 4 h energy expenditure (kcal) per participant per trial...... 147 Table 24: Comparison between the current algorithm and Wyss & Mader 2011...... 149 Table 25: Participant characteristics...... 161 Table 26. Energy expended (EE) during the four hours that the oxygen uptake measurement system was worn and estimates of daily EE on the test day...... 168 Table 27: Energy intake and relative energy intake by condition...... 169 Table 28: Percentage of participants whose energy intake varied by more than 10% of their median energy intake in each condition...... 170 xiii

Table 29: The area under the curve for the 4 indices of appetite over the trial day for each condition...... 173 Table 30:Pearson product-moment correlations between appetite hormones and dietary consumption 30 minutes following sample collection ...... 177 Table 31:Pearson product-moment correlations between appetite hormones and appetite sensation ...... 178 Table 32: Pearson product-moment correlations between appetite sensation and dietary consumption 30 minutes following visual analogue scale completion ...... 178 Table 33: Measured vs. Estimated 4-h energy expenditure during the temperate trial ...... 190 Table 34 :Measured vs. Estimated 4-h energy expenditure corrected for environmental condition ...... 194

xiv

List of Secondary Tables Table I: Typical weight loss in military personnel when adequate rations are provided (short term) ...... 221 Table II: Typical weight loss in military personnel when adequate rations are provided (long term) ...... 224 Table III: Typical weight loss in military personnel when restricted rations are provided ...... 225 Table IV: Effects of exercise on circulating leptin and appetite ...... 226 Table V: Effects of exercise on circulating ghrelin and appetite ...... 227 Table VI: Effects of exercise on circulating GLP-1 and appetite ...... 233 Table VII: Effects of exercise on circulating PYY and appetite ...... 236 Table VIII: Effects of exercise on circulating PP and appetite ...... 240 Table IX: Percentage of participants who thought they ate more, less, or the same amount they normally would during each trial ...... 287 Table X: Why participants thought they ate more in each condition ...... 288 Table XI: Why participants thought they ate less in each condition ...... 288 Table XII: Energy intakes in different scenarios ...... 290 Table XIII: 24-h urinary free cortisol before and after each trial ...... 294 Table XIV: POMS fatigue scores before and after each trial ...... 296 Table XV: Substrate oxidation (kcal∙h-1) during each trial ...... 297

xv

List of Figures Figure 1: Energy balance schematic...... 3 Figure 2: Graphic depiction of weight loss risk in otherwise healthy normal weight individuals. 4 Figure 3: Reported weight loss during military field training/operations...... 7 Figure 4: Factors that impact energy expenditure (EE) and energy balance...... 13 Figure 5: Average energy expenditure (EE) of each military group relative to the average ambient temperature...... 20 Figure 6: The amount of energy expended vs. the amount of energy provided during each military operation or training course...... 31 Figure 7: The amount of energy expended vs. the amount of energy consumed during each military operation or training course...... 32 Figure 8: Hormonal regulation of appetite...... 37 Figure 9: Average absolute energy intakes during both exercise and control trials as reported in 19 published repeated measures studies...... 43 Figure 10: Average relative energy intakes (absolute energy intake - energy expenditure) during both exercise and control trials as reported in 19 published repeated measures studies...... 44 Figure 11: A graphical depiction of the study design...... 63 Figure 12: Schematic of each experimental condition...... 69 Figure 13: The principal of electrochemiluminescent immunoassays...... 76 Figure 14: Graphical depiction of the study design...... 104 Figure 15: Average core temperature (°C) for each condition ...... 109 Figure 16: Average thermal comfort ratings reported for each condition ...... 111 Figure 17: The 5-min average core temperature (°C) measured at each hour vs. the thermal comfort rating reported at the same time for each participant...... 112 Figure 18: Average heart rate during each condition ...... 113 Figure 19: Average Borg scale rating during each active condition ...... 113 Figure 20: Total energy expended during four hours of activity (Hot, Temperate, Cold), or rest (Sedentary)...... 116 Figure 21: Average energy expended (kcal·h-1) during all activities and activity subtypes ...... 117 Figure 22: Average energy expended (kcal∙kg-1∙h-1) relative to nude or clothed weight...... 118 Figure 23: Graphical depiction of the study design...... 131 Figure 24: Accelerometry and metabolic data that was used to develop the algorithm...... 135 Figure 25: Regression plots ...... 140 Figure 26: Regression and Bland-Altman plots...... 142 Figure 27: Measured METs vs. Predicted METs for the cross-validation participants in all active conditions ...... 144 Figure 28: Measured METs vs. Predicted METs for the cross-validation participants in all active conditions (Hot, Temperate, and Cold) for each non-treadmill activity...... 145 Figure 29: Graphical depiction of the study design...... 162 Figure 30: Schematic of each experimental condition...... 164 Figure 31. The 24-h energy balance for each participant ...... 169 Figure 32. Average dietary intake over time ...... 171 xvi

Figure 33: Average appetite scores for all participants ...... 172 Figure 34: Average appetite scores during each condition ...... 173 Figure 35. Average plasma volume change relative to fasting (time 0) during each condition 174 Figure 36. Average appetite hormone concentrations collected at each time point ...... 175 Figure 37: Expected hormonal effects in Study 4...... 199 Figure 38: Average ratings of the military food that was provided...... 289 Figure 39: Average energy intake when consuming personal food vs. IMPs at home...... 291 Figure 40: Average scores for each index of subjective appetite during each condition ...... 292 Figure 41: Appetite regulating hormone concentrations during each condition...... 293

xvii

List of Appendices Appendix A – Weight Loss in Military Personnel...... 221 Appendix B – Effects of Exercise on Appetite-Regulating Hormones ...... 226 Appendix C – Recruitment Poster ...... 241 Appendix D - Recruitment Email ...... 242 Appendix E – Consent Form...... 243 Appendix F – Invasive Procedures Consent Form ...... 256 Appendix G – PAR-Q+ ...... 257 Appendix H – Participant Demographics ...... 261 Appendix I – Pittsburgh Sleep Quality Index ...... 262 Appendix J – Physical Activity Questionnaire ...... 268 Appendix K – Food Record Instructions ...... 269 Appendix L – Visual Analogue Scales for Appetite (VASA) ...... 273 Appendix M - GSQS ...... 275 Appendix N – Thermal Comfort Scale ...... 276 Appendix O – Borg Scale ...... 277 Appendix P – End of Study Questionnaire ...... 278 Appendix Q – Food Satisfaction Survey ...... 281 Appendix R – University of Toronto Ethics Approval ...... 284 Appendix S – DRDC Ethics Approval ...... 285 Appendix T – Perceptions of Food Intake ...... 287 Appendix U - Food Satisfaction ...... 289 Appendix V- Energy Intake at Home vs. in the Environmental Chamber ...... 290 Appendix W – Subjective Appetite ...... 292 Appendix X - Appetite Hormone Concentration ...... 293 Appendix Y – Cortisol Results ...... 294 Appendix Z – Profile of Mood States (POMS) Results ...... 295 Appendix AA – Substrate Oxidization ...... 297

xviii

Chapter One: Introduction 1.1 Overview of Dissertation There is a rich body of scientific literature on nutrition and body weight management, the vast majority of which focuses on obesity prevention and treatment, with the US National Institutes of Health spending over $965 million on obesity research in 2016 alone [1]. Most intervention strategies for obesity prevention are directed at decreasing food consumption and/or suppressing appetite. It is important to recognize that there are non-obese, healthy, and very active populations for whom the maintenance of body weight through adequate nutritional sufficiency can be an important component of their ability to meet their occupational or vocational performance objectives and to maintain their health. For such populations the requirement to maintain body weight can be just as important as weight loss is in the obese population. One such population are military personnel, who are engaged in sustained intense operations as a function of actual deployment or training for military operations.

The energy cost of military field work can be very high [2, 3], and often military personnel do not eat enough to offset the related energy expenditure (EE) [4]. Consequently, weight loss is prevalent in military populations engaged in field operations [5-8]. Rapid, or extensive weight loss is of particular concern, as such weight loss can be detrimental to physical performance [5, 9] and health [10]. In order to combat this, it is important to provide sufficient food and assess other factors (such as appetite) that may contribute to under consumption of available food.

The general civilian population guidelines for caloric consumption should not be considered as applicable to a population that is engaged in intense elevations of EE for days or weeks, which is a common characteristic of many military field operations. A logical starting point to appropriately assess the quantity of food that should be provided to sustain such operations would be knowledge of the EE expected during operations. In the case of the Canadian Armed Forces (CAF), such knowledge of EE is complicated by the fact that the physical demands (and the associated EE) of military operations can vary greatly depending on the mission at hand and potentially also as a result of the thermal stresses associated with the environments in which the CAF are expected to be able to operate [4].

As a result, three different approaches were considered in this dissertation. First, the energy cost of individual infantry tasks was measured and compiled in order to better estimate

1

EE when the activities executed by military personnel during operations/training are known (Study 1). Second, the effect of ambient temperature on EE in military personnel was investigated in order to determine whether a caloric supplement is necessary in harsher thermal climates (Study 2), and third a military specific algorithm was developed using inexpensive devices to allow for the ongoing assessment and collection of EE data during military operations to further improve EE/EI matching over time (Study 3).

In order to ensure that CAF members are optimally nourished and as a result performance ready, it is important that the supplied food is in fact consumed. Unfortunately military personnel engaged in physically arduous field operations or training under eat even when sufficient food is provided to them [3, 11, 12]. Considering that voluntary anorexia can confound and ultimately thwart attempts to optimize nutritional adequacy, the role of appetite in this observed phenomenon was also investigated (Study 4). Currently voluntary anorexia is often explained by the palatability of military food, insufficient time to eat, and inconvenient or lengthy food preparation [13-15], suggesting that the solution to this phonmenon can be remedied by improving several food related factors (taste, packaging, and ease of military food preparation). This further highlights the importance of examining the role appetite suppression plays in voluntary anorexia, as many of these efforts would be ineffective in restoring appetite.

In effect, this research lays the groundwork for how much food to provide to CAF members in varying field-feeding situations, and investigates the role subjective appetite, and various appetite regulating hormones play on food intake in military personnel.

For the purposes of this dissertation harsh environments were defined as the following: hot refers to ambient temperatures ≥ 30°C, cold refers to ambient temperatures ≤ 0°C, and temperate climates were defined as those environments where the temperature on the majority of days fell well within the range of 1°C and 29°C.

2

Chapter Two: Literature Review

2.1 Energy Balance

Maintaining a normal body weight is associated with improved health [16] and performance [17, 18] outcomes while decrements in both health and performance are found with both increased weight (obese), and decreased weight (underweight) categories. Although maintaining a normal body weight alone does not necessarily preclude health risks, a J-shaped curve is still typically found between weight and various health factors [19]. In general, increased energy intake (EI) leads to weight gain [20], while decreased EI leads to weight loss [21, 22], similarly, increased energy expenditure (EE) leads to weight loss [23, 24], and decreased EE leads to weight gain [25](Figure 1).

Figure 1: Energy balance schematic. Optimal health and performance are associated with the maintenance of a normal body weight, while decrements are found at both ends of the weight spectrum. Increased energy intake generally leads to weight gain, while decreased energy intake leads to weight loss, similarly, increased energy expenditure leads to weight loss, and vice versa. For military personnel the implications of EE/EI imbalance can be more complex because any negative health and performance effects can impact the ability to sustain military operations or training. In military personnel engaged in field operations, health and performance are more likely to be affected by weight loss rather than weight gain due to the physical demands of the work. Weight loss in these scenarios is often due to a combination of increased EE and decreased (or a lack of compensatory increase in) EI. Both the absolute magnitude of

3 weight loss and the rate of weight loss have been reported to impact health and performance in this population [5, 10].

2.1.1Physiological and Psychological Effects of Weight Loss

In healthy individuals of average weight, short term undernutrition resulting in weight losses of <3% body weight over 7 to 30 days have negligible effects on measurable indicators of health and working capacity [11, 26, 27] (Figure 2). Nevertheless a 3% decrease in body weight may not be trivial considering that behavioural changes such as irritability and lethargy have previously been reported [28], suggesting that such changes can ultimately have negative impacts on group dynamics and concentration [29]. In addition when such weight losses occur in a short timeframe dehydration is the likely culprit, in which case there may be a negative impact on performance [15]. Weight losses between 1%-3% have previously been shown to result in a 15% impairment in physical performance when such losses occur over a short period of time (<1 week) [5] (Figure 2).

Figure 2: Graphic depiction of weight loss risk in otherwise healthy normal weight individuals. The green area depicts low risk weight loss: weight loss of this magnitude and at this rate will likely lead to few detrimental effects. The yellow area depicts moderate risk weight loss: weight loss of this magnitude and at this rate will likely lead to some detrimental effects which will likely affect physical performance more than health. The red area depicts high risk weight loss: weight loss of this magnitude and at this rate will likely lead to detrimental health and performance effects.

When bodyweight losses are greater than 10% of initial weight, there are serious consequences (Figure 2). This magnitude of weight loss has been reported during military field exercises and can result in suppressed immune function [10], reduced physical performance [5, 10, 30-33], and reduced cognitive function [31]. Additionally, the greater the discrepancy

4 between EE and EI, the more pronounced the health or performance decrement is at every absolute weight.

In a study conducted on 55 U.S. soldiers from the Ranger Training Program (an arduous 62-day course where trainees are physically and mentally tested under highly stressful conditions) who received restricted rations (military food supplying only 1350 kcal per day), weight losses of 15.6% of initial body weight over 62 days resulted in the suppression of the immune system [10]. The immune suppression peaked after the first two weeks when the energy discrepancy between EE (high) and EI (low) was the largest. Furthermore, it was closely followed by greater incidences of infection during the phase with the largest energy imbalance [10]. When given an additional 400 kcal per day infection rates decreased from 25% (19 out of 75 of participants) to 8% (7 out of 85 of participants) [31]. Increased EI decreased the incidence of infection, however caloric intake was still inadequate as soldiers lost an average of 12.6% of body weight and immune function, although improved compared with the lower EI trial, was still clinically suppressed [31]. The soldiers in these two studies did have short periods of time between phases during which more food was provided. Regardless, these periods between phases were not long enough to prevent the negative health consequences of undernutrition.

Regarding physical performance, with >10% of weight loss, no effect on grip strength was noted, however a 16%-20% reduction in maximal lift capacity, and a 21% decline in peak power as determined from two maximal jump tests (one with, and one without counter- movements) were previously reported [9, 10, 31]. Strength decrements are similar regardless of whether subjects lose 16% or 13% of initial body weight illustrating that weight losses greater than 10% of initial weight have detrimental effects on performance [10, 31].

Food restriction also has unfavourable effects on cognitive performance. During the first phase of the Rangers Training Program described above (16 days in length), soldiers lost an average of 6% of their initial bodyweight; at this level of weight loss, no cognitive impairments were found [31]. As training continued however, and soldiers lost more weight it became evident by the end of training that information processing capacity, basic memory, reasoning and pattern recognition were impaired [31]. Soldiers became slower on these tests, presumably in an attempt to uphold baseline accuracy levels. The decrements amount to a 10-34% decline in cognitive performance, but it is important to note that the Ranger Training Program involves

5 several stressors beyond food restriction such as sleep deprivation and danger which can also hinder cognitive function [31].

2.1.2 Typical Weight Loss in Military Personnel

During field exercises and deployment, it is not uncommon for military personnel to lose weight (Figure 3). In typical military scenarios lasting between 4 and 12 days, with the provision of adequate rations (military food composed of pre-cooked, ready-to eat foods), military personnel were reported to consume approximately 67% of what they are given and lose approximately 1.5% of their initial bodyweight (Appendix A: Table I). When sufficient rations are supplied to soldiers performing support and administrative activities, body weight is maintained as was reported for the Zimbabwean support soldiers [2]. However more strenuous missions [5], or those completed at altitude [11, 12] resulted in weight loss. Although the greatest extent of weight loss (for operations/training lasting <12 days) did not exceed 3% of initial body weight, where such weight loss occurred rapidly (within 5 days) physical performance can be expected to be impaired [5].

With longer use, rations are less likely to be effective at sustaining body weight. When rations are used for 60 consecutive days, weight losses above 3% have been reported to occur (Appendix A: Table II). The increased weight loss has been attributed to the decreasing palatability of rations over time [8].

Although, military personnel tend to consume enough food to support health and performance when provided with calorie-dense rations (rations that contain considerably more calories than are physiologically required), the same has not been the case for personnel on missions involving restricted rations (Appendix A: Table III). When restricted rations are provided, weight losses above 3% are very common, and very substantial losses, averaging 5.7% of initial weight have been reported within 16 days (Appendix A: Table III). Military personnel provided with these severely restricted rations were often dehydrated, and displayed the expected impairments: in immune function, capacity for muscle power generation, and cognitive function associated with undernutrition [10].

Although certain training programs provide restricted rations in an attempt to habituate soldiers to function effectively when food is scarce, the remainder of this dissertation will focus on military situations where the amount of food that was provided to military personnel was

6 intended to be adequate for the conditions endured. In addition, even though military personnel are at times required to operate at altitude, this is also outside of the scope of the current dissertation. As a result, when considering military training/operations where adequate food was provided, and where soldiers were not at altitude, 37% (7 of 19) of military groups lost weight more rapidly or to a greater extent than would be deemed safe (<3% body weight over 7 to 30 days) (Figure 3).

Figure 3: Reported weight loss during military field training/operations when the provided food was deemed to be adequate or in surplus of assumed need. Each data point represents the percentage of initial weight lost throughout a military exercise of a particular military group. These data were collected from publicly available military reports and journal articles. Red data points signify data conducted in the Canadian Armed Forces (CAF). The green area depicts the low risk weight loss zone; military groups (data points) that fall within this area will likely experience few detrimental health and performance effects, whereas the military groups that fall within the yellow area have a higher likelihood of enduring health and performance decrements due to the rate and/or extent of weight loss [2, 5-8, 26, 27, 34-37].

Limited data are available regarding weight loss in Canadian Armed Forces (CAF) members engaged in field operations. When assessing three relevant reports [5, 27, 36], weight loss fell outside of the low risk weight loss zone in 3 of the 5 reported groups (Figure 3). CAF members lost weight very rapidly suggesting that likely both food and fluid intake were insufficient. Therefore, it is important to clarify whether the energy content of rations provided is adequate.

2.2 The Canadian Combat Ration Program

Frequently when CAF are in the field (on a military exercise, training, or mission), having a kitchen on site is unfeasible. Under these conditions military personnel are provided with combat rations. Combat rations are composed of pre-cooked, ready-to eat foods that have a shelf life of 36 months when stored under appropriate conditions (7°C-24°C) [38]. According to 7

CAF guidelines and policies, these rations are not meant to be exclusively used for more than 30 consecutive days. Even though every effort is made to make fresh foods available (especially after individuals subsist exclusively on rations for two weeks), this is not always possible and consequently, fresh foods are not relied upon for nutritional adequacy.

The activities and tasks that may be expected of military personnel vary greatly from mission to mission, as do the environmental conditions they are exposed to; resultantly, the nutritional requirements might vary greatly as well. For this reason, the Canadian Combat Ration Program has several different types of rations in an attempt to match the nutritional needs of soldiers when taking environmental conditions, activities at hand, and the supply chain into account [38, 39].

The five components of the Canadian Combat Ration Program are: the Individual Meal Pack (IMP), the Light Meal Combat (Emergency Ration) (LMC), the Survival Food Packets, the Alternative Meal Pack which includes vegetarian, Halal, and Kosher options, and the Food Supplements which include the Tropical Supplement, the Arctic Supplement, and the High Protein Drink [40].

The IMP is the most frequently used as it is intended for standard military operations. It was created to be nutritionally sufficient and quick to prepare. It can be eaten cold, or heated via body heat or in boiling water [39]. Each IMP contains an entrée, dessert, sport drink, bread, jam/peanut butter/honey, 2 hot beverages, and condiments. Depending on the meal (breakfast, lunch or dinner), additional products are also provided [38]. Each IMP contains approximately 1200 kcal, of which 50% come from carbohydrate, 15% from protein and 35% from fat [40]. IMPs reflect Canadian food preferences and also include several meatless options [40].

When compared to the combat rations provided by 11 other nations (Australia, Belgium, Czech Republic, France, Germany, Italy, Netherlands, Norway, Slovenia, United Kingdom and the United States of America), Canada’s IMP is the most calorically dense, supplies the greatest amount of carbohydrate and protein and is moderate in its fat content [38]. Nonetheless, the IMP does not contain enough calcium and folic acid to meet the Nutrition Recommendations for Canadians; however, the amount supplied is acceptable as long as the IMP is not exclusively consumed for more than 30 consecutive days [40]. If IMPs are consumed for more than 30 consecutive days, medical officers are required to assess whether vitamin and mineral

8 supplements are required [40]. Each IMP weighs about 0.73 kg, as a result food for one day increases a soldier’s pack by 2.2 kg which is relatively high when compared to rations from other countries [38].

In particularly demanding circumstances when three daily IMPs are insufficient, LMCs can be provided as a supplement. Furthermore in extreme heat or cold when it is deemed that additional nutrients and fluids are necessary, the Arctic or Tropical rations are supplied [40].

2.2.1Energy content of supplied food

The energy content in the three IMPs (totaling ~3600kcal assuming everything in each ration pack is consumed) is based on the Dietary Reference Intakes (DRI). The DRIs represent the current state of knowledge with regards to the nutrient requirements of healthy populations. In terms of energy requirements, the DRIs take into consideration the Estimated Energy Requirement (EER) – which is the amount of energy that a healthy, normal-weight, individual should consume in order to maintain energy balance [41]. Since the EER is devised for healthy adults within the desirable weight range, the goal of the EER is to match total energy expenditure (TEE).

In order to determine energy requirements for North Americans, the Institute of Medicine (IOM) collected doubly labeled water (DLW) studies displaying TEE in a wide range of people and compiled them all in a database. The normative database for adults included 407 people of whom 169 were men and 238 were women. All adults that were included in the database had their heights and weights measured and were: healthy, not gaining or losing body weight, and had a BMI between 18.5 and 25 kg/m2 [41]. Subjects were excluded if they were taking part in underfeeding or overfeeding studies, if they were exceptionally active (military personnel, elite athletes, etc.), or if they had any acute or chronic illnesses [41]. From this database, TEE prediction equations were developed based on age, sex, height, weight, and physical activity level (PAL).

The equations for adult (age 19+) men and women are as follows: “Adult man: EER = 662 – 9.53 × age (y) + PA × (15.91 × wt[kg] + 539.6 × ht[m])

Adult woman: EER = 354 – 6.91 × age (y) + PA × (9.36 × wt[kg] + 726 × ht[m])”

PA (physical activity) is given a score of 1.0, 1.11 (1.12 for women), 1.25 (1.27 for women) and 1.48 (1.45 for women) for PALs designated as sedentary, low active, active, and 9 very active respectively [42]. The four PA designations represent activities up to a PAL of 2.5 (very active), as all PALs above 2.5 were omitted from the database used to derive these equations [42]. A PAL of 1.9-2.5 is the equivalent to completing typical activities of daily living (household tasks, office work etc.) plus about 180 min of moderate physical activity (slow swimming, leisurely cycling etc.), or 60 min of moderate and 60 min of vigorous physical activity (jogging, tennis, climbing hills etc.). Not surprisingly, PALs above 2.5 include elite athletes and military personnel, accordingly, these equations are likely not valid under all circumstances in the military population.

There is a special designation within the DRI for military personnel (Military Dietary Reference Intakes (MDRI)) that the US military uses to guide energy provisions. The MDRI for energy are similar to the EER equations but generally exclude height [43]. Considering that military personnel have largely different activity levels depending on their job (i.e. support vs. combat), it is reasonable that both the CAF and the US Armed Forces standard military rations are moderate in EI and can be supplemented when operations are more strenuous. The daily energy content that is currently provided in both countries is ~3600kcal and is based on a moderate, not vigorous activity level. When applying the most recent anthropometric data of the CAF (Table 1) and assuming a PAL of 1.6-1.89 (active - typical activities of daily living and ≥60 min of moderate activity) the EER for men and women is 3288 kcal/day and 2447 kcal/day respectively which is similar to 3250 kcal/day and 2300 kcal/day recommended for men and women by the MDRI for energy [44].

Table 1: Average anthropometric data for men and women in the CAF Males Females Age (y) 33.4 34.7 Weight (kg) 88 69 Height (m) 1.77 1.64 Data from [45]

The MDRI do acknowledge that these estimates will be inadequate for military members who are: 1) larger than the reference military soldiers (heavier than 79kg and taller than 175cm); 2) involved in heavy or prolonged physical activity; or 3) exposed to extreme environmental conditions [44]. Although these acknowledgments are explicitly made, the procedures regarding energy provision increases are not clearly defined. Similarly, the CAF are able to supplement the standard daily rations, however current procedures governing increased food provisions are poor. 10

2.2.2 Incremental Allowances (IA)

Although the energy content of IMPs is based on data that likely underestimates the caloric need of most field operations (especially combat operations), the standard IMPs can be supplemented with LMCs and other rations. Furthermore, on top of the 5 different types of rations, additional food is also available to supplement military rations in the form of incremental allowances (IA). IA are expressed in financial terms, as the amount of money available to be used to give military personnel additional food. It is only authorized under certain circumstances i.e. when it is subjectively deemed that the usual amount of food that is given for the day is insufficient, or when military rations are exclusively consumed for an extended period of time. IA often fund fresh food, and snacks that are not necessarily part of the rations listed above. Although it is clear that there are some general procedures in place to provide more food to military personnel in different scenarios, these procedures are fundamentally based on financial factors with the duration of ration-only sustenance, morale, perceived mission intensity, and climate harshness taking a secondary role which is often determined subjectively. Considering the potential health and performance implications of under-supplying and the financial waste of over-supplying food, the current dissertation aims to provide a physiological foundation from which food provision procedures can be determined for CAF field operations.

2.3 Factors Affecting Energy Expenditure

CAF members engaged in field operations are required to meet the minimum operational standards, and as such it is reasonable to assume that such individuals would fall within a healthy and normal weight. As a result, the amount of energy that military personnel should consume in order to maximize health and performance should (similarly to the process outlining the EER) match total energy expenditure (TEE).

There are four components of TEE – basal metabolic rate (BMR), thermic effect of food (TEF), thermoregulation, and physical activity/exercise [46]. The relative divisions of these three components in thermoneutral environments in average individuals are as such: BMR makes up 60-75% of TEE, TEF contributes 10%, and physical activity contributes the rest. BMR and TEF are determined involuntarily, while the amount of energy expended through physical activity is directly proportional to the voluntarily chosen intensity of physical exertion

11 plus whatever burden is added by the environment. In exceptionally active individuals (like military personnel) physical activity can constitute the biggest fraction of TEE [46].

BMR is highly influenced by body weight [47], with increases in mass leading to increases in BMR mostly as a result of increased fat free mass (the more metabolically active tissue) [48] and vice versa. The TEF is affected by EI, with the amount of energy expended being directly related to the size of the meal consumed: larger meals (especially those higher in protein or alcohol content) require more energy for digestion than smaller ones [49]. Thermoregulation is typically not a large component of TEE in a thermoneutral environment, but can become substantial as the ambient temperature becomes more extreme [50, 51]. Exercise is the most easily affected component of TEE. Body weight also typically affects the amount of energy expended through exercise, as more energy is required to move a heavier body than a lighter one [52, 53], similarly the more muscle mass that is recruited during exercise the more energy is expended [54] (Figure 4) .

12

Figure 4: Factors that impact energy expenditure (EE) and energy balance. Optimal health and performance are associated with the maintenance of a normal body weight, while decrements are found at both ends of the weight spectrum. Increased energy intake, and/or decreased energy expenditure generally lead to weight gain and vice versa. The four components of EE are: – basal metabolic rate (BMR), thermic effect of food (TEF), thermoregulation, and exercise. BMR is influenced by body weight, with increases in mass leading to increases in BMR mostly as a result of increased fat free mass. TEF is affected by energy intake, with the amount of energy expended being directly related to the size of the meal consumed: larger meals require more energy for digestion than smaller ones. Thermoregulation is typically not a large component of EE in a thermoneutral environment but can become substantial as the ambient temperature becomes more extreme. Body mass also typically affects the amount of energy expended through exercise, as more energy is required to move a heavier body than a lighter one, similarly the more muscle mass that is recruited during exercise the more energy is expended.

Energy requirements of military personnel are greatly dependent on the activities performed, and the environmental conditions endured. While BMR and TEF are relatively invariable, physical activity and thermoregulation can vary greatly within military personnel as military operations are at times very physically demanding and can occur all around the world in largely different thermal environments. As a result, determining the energy cost of physical activities, and the added cost of thermal environment becomes paramount in this population.

13

2.3.1 Energy Requirements of Military Personnel

Military personnel generally expend considerably more energy than civilians. On average military men expend 38% more energy per day than male civilians of the same age while military women expend 17% more energy, than civilian women of the same age [4]. EE tends to be higher for men because men tend to have larger body sizes, more lean body mass and historically more physically demanding jobs than women [4]. Currently there is more equivalence in job assignment than in previous years, and consequently energy requirements in both sexes are mostly related to body weight [55, 56]. Energy requirements for male and female military personnel engaged in a variety of activities in temperate climates are approximately 4008 kcal/day and 3200 kcal/day respectively (Table 2).

Energy requirements are mainly driven by the activities that are required of military personnel, combat training activities for example tend to be more taxing than support activities [4]. Considering that the diversity of duties range from administrative jobs to construction to combat training, it is important to recognize that even in a temperate climate, energy requirements can range from 2332 ± 373 kcal/day for female administrative workers, to 6353± 478 kcal/day for male Norwegian soldiers in the midst of field training as measured via DLW (Table 2).

14

Table 2: Energy Expenditure in Temperate Environments as Measured by the Doubly Labelled Water Technique Temperature Location Duration of Total Energy Expenditure Population (Relative Activities Reference (Climate) Trial (Days) (kcal/day) Humidity) 3709 Administrative duties during field training (Men) 5 (1 man, 4 women) 2332 ± 373 (Women) Medical duties (nurses, operating room specialists) 3880 ± 872 (Men) 3959 ±159 27 Combat during field training 12 Camp Mackall area 2872 ± 229 (Men) Support Hospital (6 men, 6 women) of Temperate (Women) (CSH), [56] Fort Bragg, NC Environment 2745 personnel (10 4261 (Men) Medical specialists and Lab specialists during field ±122 men. 17 women) 5 training (1 man, 4 women) 2940 ± 268 (Women (Women) ) Support duties (Equipment repair, Radio operators, 4174 ±431 (Men) Laundry specialists) during field training (2 men, 3 5 women) 2781 ±320 (Women) 16 (10 men and Forested military 15°C -30°C 6353± 478 6 women) from Field Exercise Course - long-distance marches, training areas (Men) 5923 ± the Norwegian combat patrols, obstacle courses, and 7 [57] northwest (Elevation: 740 Military marksmanship training. of Oslo, Norway ~ 500 m) 5231 ± 478 Academy (Women) 3473 + 807 28 Naval Training mission at sea aboard a U.S. Navy Ship In-port: 8 days 3141 + San Diego, (Men) personnel (10 NA -High physical demands (5 men. 9 women) At-sea: 10 618 [56] CA men, 18 women) -Low physical demands (5 men. 9 women) days 2808 + 429 (Women)

6 US Marines Administrative and support work 3109 ± 543 Great Inagua Island, 3328 ± Bahamas 26°C (70%) 21 [8] 637 10 US Marines- (Warm and Humid) construction Construction work 3460 ± 732 engineers 16 Special Camp Ethan Allen Field Training: Reconnaissance, surveillance, and 28 operations Training Center, -1°C to 16°C 3400 ± 260 [35] electronic warfare. soldiers Jerico, VT

15

Table 2: Energy Expenditure in Temperate Environments as Measured by the Doubly Labelled Water Technique Temperature Location Duration of Total Energy Expenditure Population (Relative Activities Reference (Climate) Trial (Days) (kcal/day) Humidity) Phase 4 - Desert Phase at Dugway 5 US Ranger 19°C Ranger Training- desert combat operations, live- Proving Ground, 14 ~3800 [10] students (6°C-33°C) fire raids and ambushes, desert survival skills, etc. Utah OR Fort Bliss, Texas 9 Special Forces Combat Training in Garrison - foreign language A Team Soldiers practice, mountain hiking, rock climbing, urban 4099 ± 740 in Garrison warfare training, and small weapons handling. Ft Carson, CO NA 9 [58] 9 Support Routine physical training, equipment assembly, Personnel in 3136 ± 652 driving, and administrative support functions. Garrison Chocolate Mountain Gunnery Range, 7°C -31°C Artillery training exercise. Different crews likely 19 US Marines 12 4115±724 [34] California (10% - 55 %) had different tasks: i.e. loader vs. radio operator (Warm-Hot and Dry)

15 Male US

Special Forces Pre-Mission training - Physical training, weapons 3904±521 soldiers in handling, airborne, urban, and convoy operations. Garrison NA NA 7 [59] 16 Male US Combat Diver Qualification Course - 5-mile runs, Special Forces callisthenic workouts, high intensity pool 4569±351 soldiers in workouts, drills carrying 30kg of diving gear etc. Garrison

Phase 2 - Mountain 22°C Ranger Training- mountain 5 US Ranger Phase Camp Frank (12°C -32°C) rappel, rock climb, conduct ambushes and raids, 18 ~4500 [10] students D. Merrill near (Elevation: environmental Dahlonega, Georgia ~ 500 m) and survival training etc. All: 3838kcal/day WEIGHTED Men: 4008kcal/day AVERAGE Women: 3200kcal/day

16

2.3.2 Energy Cost of Environmental Temperature

While physical activity has the largest impact on EE, the thermal environment can also affect EE. Since military operations often include physically demanding activities of varying durations in thermally stressful environments, the impact the environment has on EE needs to be considered.

2.3.2.1 Physiological Effects of Exercise in Hot Environments

During exercise, the energy released in association with each litre of oxygen consumed to fuel muscular work amounts to about 5 kcal. Typical mechanical efficiencies of human exercise mean that only about 1 kcal of that energy will be transduced into mechanical work while the rest is released as heat [60]. This rise in heat ultimately needs to be dispersed into the atmosphere to maintain the body temperature within safe limits. Humans can transfer heat out of the body via radiation, convection, evaporation, and to a lesser extent conduction. Exercise in the heat, as opposed to a temperate environment, adds an additional stressor on an already taxed system, because heat loss through radiation and convection are greatly reduced in hot environments significantly increasing our reliance on evaporation to regulate temperature [61]. The capability of evaporation to adequately maintain body temperature is dependent on several factors: humidity, body surface area, clothing, hydration state, physical fitness, and heat acclimatization. While sweat evaporation may suffice for thermoregulation on a hot, dry day, high humidity will substantially diminish evaporation, increasing the necessity of heat loss via radiation and convection. On a hot and humid day, radiation, convection and evaporation are all largely reduced. Studies that reported increases in EE in the heat, tended to expose their participants to greater heat stress, for example Consolazio et al. found that energy requirements for work done at 100° F (38°C) were roughly 8-12% greater than the same work done at 85°F (29°C) or 70°F (21.2°C). [62].

There are three considerations that greatly determine an individual’s ability to effectively thermoregulate during exercise in a hot environment and potentially indirectly affect EE in the heat: the individual’s hydration level, acclimatization state and aerobic fitness level [63]. Fluid balance is of utmost concern in the heat because severe dehydration can result in decreased blood volume which can impair heat dissipation through the evaporation of sweat [64]. Such decreased blood volume would also significantly increase the cardiovascular burden associated with the shunting of blood to the periphery during heat stress which is important for more

17 effective conductive and convective heat loss. Heat acclimation on the other hand increases sweat production, and resultantly is associated with decreases in both core and skin temperature at rest and during exercise (although dehydration blunts this effect) [65]. As well, acclimation can significantly decrease the rate of metabolism during exercise in the heat (by 3%) and thereby decrease energy requirements for exercise in such weather [64]. Lastly the more aerobically fit an individual is the greater their heat tolerance also tends to be [65] and perhaps the increase in EE during exercise in the heat is partially blunted in fit individuals.

2.3.2.2 The Energy Cost of Exercise in Hot Environments

Previous studies assessing the effect of high ambient temperature on EE have been inconsistent in their findings with reported decreases [66, 67], no effect [68], and increases [51, 69-72] of EE in high ambient temperatures. Inconsistencies seem to arise as a result of methodological differences between studies (exercise employed, duration of heat exposure, participant acclimatization status, etc.) as well as discrepancies in the definitions used; with "temperate" referring to 15°C in some studies[66] and 25°C [72]in others and "hot" referring to temperatures ranging from 30°C [67] to 46.2°C [71].

Studies that report decreased EE in the heat tend to suggest that the EE reductions are due to enhanced muscle efficiency [68] or increased metabolic efficiency in the heat [73], that may result from a pre-emptive mechanism by which the cutaneous thermoreceptors signal the preoptic anterior hypothalamus to decrease metabolism [67]. It is important to note however that these studies also tend to subject their participants to a lower heat stress, such that heart rate responses between temperate and hot conditions are not different [67], or the heat exposure is too short in duration, resulting in limited increases in both core temperature, and thermoregulatory processes such as sweating and skin blood flow [74].

Others report that total EE is unchanged in hot environments, but the aerobic component (as determined by oxygen consumption) decreases, resulting in an increased reliance on the anaerobic component of EE [75]. This commonly reported increase in the anaerobic component as a result of exercise in the heat has been postulated to be due to reduced muscle blood flow (and accordingly reduced oxygen delivery to muscle) [70, 75], increased plasma epinephrine concentrations [76], and increased enzymatic activity or Q10 effect[68].

18

When heat stress conditions are more severe due to higher temperature, humidity, and/or exposure duration, increased EE is often reported [51, 69, 72]. Elevated EE that is seen during exercise in the heat is often attributed to the energy cost of heat dissipation – increased blood circulation, increased sweat gland activity, increased ventilation, and elevated body temperature [51].

Ultimately the amount EE is increased in hot vs. temperate environments may be due to the severity of the heat stress endured.

2.3.2.3 Physiological Effects of Exercise in Cold Environments

In cold environments, if an individual feels cold, a ~7% increase in EE occurs as a result of non-shivering thermogenesis [77], once shivering is initiated, EE can increase as much as five times above the basal metabolic rate [50]. Since 75% of the energy that is produced by muscular contraction is released as heat, the energy cost of exercise in the cold will depend on whether the exercise intensity is high enough to create enough heat to match or surpass the rate of heat loss [78]. If the exercise intensity is too low thereby being unable to maintain the skin and core temperatures above the shivering threshold, shivering thermogenesis will be initiated and ventilation, oxygen consumption and EE will be elevated compared to a thermoneutral environment [79]. If the exercise intensity is sufficient to maintain body temperature, shivering will be prevented (although skin temperature will still be low) and ventilation, oxygen consumption and EE will mimic values more commonly seen in temperate environments [79, 80]. When dressed appropriately for the ambient temperature under sedentary conditions, there does not seem to be any difference in EE. During exercise, EE has also been reported to be similar between cold and temperate conditions [67, 81]. Increases in EE reported in the cold vs. during temperate conditions are interpreted as being attributed to the weight [82] and hobbling effect of cold weather clothing [66]. Similarly, cold weather terrain (e.g. deep snow, ice) can also impact movement [83]; as a result increases in EE in appropriately dressed individuals are associated with cold weather clothing and the terrain traversed rather than physiological thermoregulatory responses.

2.3.2.4 Energy Requirements of Military Personnel in Different Ambient Temperatures

Although there is a theoretical effect of environmental temperature on EE in military personnel, it is not clear whether that translates to higher overall EE in military operations completed in harsher climates. When plotting daily EE vs. the ambient temperature as reported 19 in 42 military groups (found in19 different studies and military reports), no clear trend emerges (Figure 5).

Figure 5: Average energy expenditure (EE) of each military group relative to the average ambient temperature. Data were gathered from peer reviewed studies and military reports where EE was determined using the doubly labelled water technique. Each data point represents the mean EE for each military group. Data were grouped by task and sex when such data were available. Lavender data points depict female military personnel and grey depict male personnel, data points with a white star depict Canadian data. The dashed horizontal line depicts the amount of energy provided by the standard CAF military rations without any form of supplementation (3 IMPs =3600kcal). Where ambient temperature was not given, average ambient temperature for the region was used as the ambient temperature. [2, 3, 7, 8, 10, 27, 34-36, 56-58, 84-90].

It is important to note that the military groups plotted in Figure 5 represent groups from different countries, who completed different tasks, and who had different objectives. As such, the role of ambient temperature on EE is likely overshadowed by the role physical activity has on EE. Nevertheless, the impact of ambient temperature on EE does not seem to be large enough to unquestionably dictate caloric need in the absence of military activity data. In addition, Figure 5 demonstrates how important it is to supplement the 3 IMPs provided during most field operations and training, as most military groups were found to expend more than 3600kcal/day in these published studies and military reports.

In the seven studies that measured EE in military personnel in hot environments using the DLW technique, EE ranged from 3344 ± 239kcal per day (for Zimbabwean support soldiers in the African bush) to 5494 ± 1239 kcal per day (for Zimbabwean combat soldiers engaged in combat training) [2]. The average EE in this population under these conditions was found to be 20 around 4283 kcal/day (Table 3). In the eight studies (10 military groups) that measured EE of military personnel in cold environments using the DLW technique, average EE was found to be

4910 kcal/day (Table 4).

21

Table 3: Energy Expenditure in Hot Environments as Measured by the Doubly Labelled Water Technique Temperature Duration of Total Energy Population Location (Climate) (Relative Activities Trial Expenditure Reference Humidity) (Days) (kcal/day) 4 Zimbabwean Kariba, Zimbabwe Food preparation, cleaning and marinating 40°C (29%) 12 3344±239 [2] soldiers (Hot and Dry) camp Phase 3 - Florida Phase at Ranger Training- small boat operations, 6 US Ranger Camp James E. Rudder at 27°C (74%) stream crossing, jungle/swamp survival 16 ~3550 [10] students Eglin AFB, Florida (22°C -34°C) techniques, live-fire training, air assault, (coastal swamp/jungle) etc., 8 Australian RAAF Base Scherger, 24°C-33°C (71%- Airfield Defence Australia Routine ground defense training exercise. 12 3702±1051 [84] 96%) Guards (Hot and Wet) Multifunctional military activities: 10 Israeli Infantry Northern Israel Trekking over difficult terrain, enduring 21°C-32°C 12 3937±503 [85] Soldiers (Hot and Humid) severe weather conditions in unsheltered areas etc. Ranger Training - 3-5 mile runs, hand-to- Phase 1 Temperate Forest 5 US Ranger 27°C (65%) hand combat, road marches, battle drills, -Ft. Benning/Camp 17 ~4150 [10] students (21°C -34°C) map reading, airborne operations, land Darby, Georgia navigation, survival-type training etc. Jungle warfare training - Lectures, Short Northern Queensland, 4 Australian Army periods of intense activity (bayonet Australia NA 7 4750±531 [86] Platoon fighting, obstacle courses), and 10-18km (Hot and Wet) walks. 6 Special Forces Camp Mackall area of 26°C (65%) Special Forces Assessment and Selection Assessment and Fort Bragg, NC Course – Orienteering, obstacle courses, 20 5182 ±333 [3] Selection (Warm-Hot) (14°C-36°C) marches, swim tests, etc. Candidates 8 Zimbabwean Kariba, Zimbabwe Combat training: Callisthenics, combat 40°C (29%) 12 5494±1239 [2] soldiers (Hot and Dry) drills, sports etc. WEIGHTED 4283±711 AVERAGE

22

Table 4: Energy Expenditure in Cold Environments as Measured by the Doubly Labelled Water Technique Temperature Duration of Total Energy Population Location (Relative Activities Trial Expenditure Reference Humidity) (Days) (kcal/day) Quantico, -5°C to 15°C Combat training: live fire exercises, 10 10 US Marines 5378±678 [87] VA (20%-90%) patrolling, full gear marches etc. -15°C to 13°C Bridgeport, Field Training: Ski and snowshoe training, 23 US Marines 11 4919±911 [88] California. [2,200- to 2,550 setting up bivouac areas, 10k biathlon m elevation] Baffin 10 Canadian -40°C to 5°C Skiing, compass training, snowmobiling, Island, 10 4317±927 [27] Soldiers hunting, ice fishing, igloo building etc. Canada Multifunctional military activities: 14 Israeli Infantry Northern 0°C to 13°C Trekking over difficult terrain, enduring 12 4281±636 [85] Soldiers Israel severe weather conditions in unsheltered areas etc. -39°C to -8°C 10 US Field Fort Greely, (59%-73%) Arctic Warrior Field Training Exercise 10 4253±478 [89] Artillery Battalion AK

-15°C Pulling a heavy (150 kg) 15 6750-8260 [2000 m sleds on deep/soft snow

elevation] 2928 km ski- 2 Norwegian Navy, Greenland -15°C trek in 86 [90] SEALS (flat, with some days. Travel on flat terrain 12 3500-3850 5335± weather delays) 1826 Rugged terrain, during a -15°C 7 4710-4940 concerted effort to finish 6118 ± 977 25 Marine Recruits Parris Island, (Men) 5423 ± (15 men, 10 South Cold Gruelling 54.4-h Crucible event 5 [56] 4727 ± 651 814 women) Carolina. (Women) Winter training course including: Males: 18 Canadian Meaford, constructing trenches, pulling sleighs, 5099 ± 676 4917 ± Soldiers (9 males, -21°C to -2°C 5 [36] Ontario reconnaissance, snow shoeing, loading Females: 693 9 females) trucks, digging, and defence routines. 4643 ± 559 WEIGHTED 4910 ± 773 AVERAGE

23

2.4 Methods of Estimating Energy Expenditure

In order to make appropriate recommendations about the quantity of food provided to military personnel, it is important to first accurately estimate EE. There are numerous ways to estimate EE including: questionnaires, intake-balance method, factorial method, use of accelerometers, heart-rate recording, direct calorimetry, indirect calorimetry and the DLW technique (Table 5).

Questionnaires rely heavily on subjects accurately reporting and recalling their physical activity levels over set periods of time. Unfortunately it is very easy to over or underestimate activity levels especially as the time between activity engagement and reporting increases [91]. Another common problem with questionnaires is that many only inquire about purposeful exercise which leaves out normal everyday activity [91]. In some cases EE during every day random activity can be quite large, further increasing the likelihood of estimation error [92]. Although it would be reasonable to assume that this kind of questionnaire design would lead to underestimations of TEE, about half of the studies reviewed by Neilson et al. underestimated EE when compared to DLW, while EE was overestimated in the other half [91]. The most accurate questionnaires included non-purposeful physical activity and depended on subjects recording activities performed every 15 min of the day [91, 93]. Although more laborious for the subject, the results are more valid. For a review of questionnaire validity as compared to DLW, refer to [91].

The intake-balance method is based on the difference between the energy ingested and the body composition changes that result, for example, when weight loss occurs, EE is greater than EI. This method greatly relies on detailed and accurate measurements of intake and body composition. In controlled settings where such measurements can be made with ease, the intake- balance method is valid [88]. Outside of such an environment, accurate food intake measurements are difficult to do, and as a result reasonable EE estimates cannot be made [86, 94].

The factorial method involves accurately recording all daily activities and finding the energy costs that match these activities in published works. Similarly, to questionnaires or the intake-balance method, estimation accuracy is greatly dependent on precise recording. Small errors in intensity or duration documentation can result in large estimation errors. Currently we do not have measurements of energy requirements that correspond to every possible activity, as 24 a result, generalizations of activities that are considered “similar” increase the error in EE estimation. Furthermore activities such as fidgeting cannot be accounted for, yet such activities can increase EE substantially [92]. When well researched activities are performed and precisely documented this method is relatively valid [86].

Motion sensors such as pedometers and accelerometers quantify daily movement by counting steps or making uni-axis or multi-axis acceleration measurements. In certain instances, motion sensors are relatively reliable in estimating TEE. For military personnel engaged in walking-type activities for example, foot-ground contact pedometers were accurate in their estimations of TEE when compared to estimations with the DLW technique [95]. In a more recent military study, wrist accelerometers were also found to have relatively accurate results with an estimation error around 13% [96]. Not all studies however, have such favourable results, the Caltrac uniaxial accelerometer underestimated EE by 50%-60% in older persons [97]. Considering the wide array of available accelerometers and the way they measure movement, it does not appear that any one accelerometer is appropriately accurate in all settings as none are likely to capture all of the numerous unconventional ways the human body can move. For particular repetitive activities (marching, running, etc.), accelerometers seem to be suitably accurate when taking height, body weight, and load carried into account [95, 96].

Another method that has previously been used to estimate TEE is heart-rate recording. Although this is a relatively inexpensive and non-invasive technique, it will never be as accurate as direct calorimetry, indirect calorimetry and DLW [98]. On average the heart rate recording method has an estimation error between 5% and 20%, and tends to be more accurate in the elderly population [99, 100]. There is a relationship between minute-by-minute heart-rate and EE; however this relationship is different for each individual and is also different under various conditions [98]. As a result, for this method to be useful, initial measurements describing the individual relationship of heart-rate and EE under a variety of conditions need to be done for each person. Furthermore, non-exercise induced increases in heart rate (fear, stress, etc.) cannot be distinguished from exercise induced increases in heart rate with heart rate monitors alone; resultantly heart rate monitors are prone to overestimations in these instances.

25

Direct calorimetry measures the heat produced by an individual in a whole room calorimeter. The measurement is direct and valid, but is expensive and does not provide information on the type of fuel that is used [46]. It is also important to consider that the measurement is largely dependent on the sensitivity of the calorimeter used; as well, values need to be corrected for any equipment inside the chamber that also produce heat (e.g. treadmill).

Indirect calorimetry measures gas exchange through the use of either the closed-circuit or open-circuit methods. The closed circuit method entails breathing from a respirometer with a known volume of oxygen, while the more common open-circuit method involves breathing in atmospheric air and exhaling into a Douglas bag, Tissot tank or other collection device [93]. Indirect calorimetry is based on the assumption that the difference between the volume of oxygen inhaled and the volume of oxygen exhaled is equal to the volume of oxygen consumed. Indirect calorimetry is particularly useful for determining the substrate utilization at rest and during steady state exercise. Indirect calorimetry is accurate, but expensive, and volunteers need to be attached to a breathing apparatus. Unlike direct calorimetry however, portable systems allow for EE measurements to be done outside of a laboratory.

Doubly labelled water (DLW) is the gold standard for TEE over longer durations; it has been used in numerous populations and validated against several methods including: indirect calorimetry, the food intake-balance method, and the factorial method [86, 88, 101]. The DLW method requires the subject to ingest known amounts of deuterium and oxygen-18, which are then closely monitored in the subject’s urine, saliva or blood throughout the evaluation period. As energy is utilized, carbon dioxide and water are produced and lost via, respiration, sweating etc. Since oxygen-18 is found in carbon dioxide and water while deuterium is only found in water, oxygen-18 is lost more expediently than deuterium. The difference between the rates that deuterium and oxygen-18 are lost corresponds to the rate that carbon dioxide is produced. Although EE can be estimated from these data alone, a ratio of carbon dioxide production to oxygen consumption is made either directly or as an estimation from macronutrient intake to increase the validity of the final EE estimation. The DLW technique is ideal for studies lasting 4-21 days where activities performed are not conducive to direct or indirect calorimetry. Although DLW is easy to administer and does not interfere with daily activities, it is one of the most expensive methods.

26

Table 5: Determining Energy Expenditure - Strengths and Weaknesses of Current Methods Method Procedure Estimation Error Advantages Disadvantages - Non-invasive - Subjects self-report physical activity - Usually not very accurate - Easy to administer -The amount of time spent undertaking each From ≤10% to - Depending on the questionnaire used, Questionnaires - Inexpensive activity is then multiplied by the metabolic 113% [91] subjects at times need to accurately recall -Great for large sample equivalent corresponding to that activity to get EE activities performed days-weeks prior. sizes - Requires a long evaluation period. - Inaccurate over short evaluation periods. - Record caloric intake From 0% to 35% -Useful in “free living” -Small errors in food intake and body Intake-balance - Record body composition changes vs. DLW [86, 88, situations composition lead to larger estimation method 94] -Inexpensive errors. energy intake - change in body energy -Highly dependent on accuracy of measurements - Small errors in intensity or duration result - Record all physical activities (include intensity in large estimation errors and duration). - Non-invasive Factorial From 5% to 26% - cannot account for activities where EE - Estimate energy costs of the recorded activities -Useful in “free living” method vs. DLW [86, 99, measurements have not previously been by matching the activities to those found in situations 102] made available publications where EE was previously -Inexpensive - does not account for arbitrary movements measured. like fidgeting. - Many are not sensitive enough to -Worn motion device respond to accelerations in 1- total amount, the Motion From 6% to 60% accurately quantify EE. 3 directions. Total daily movement is later frequency, the intensity, Sensors vs. DLW [95, 97] -May not be useful for all types of physical converted to EE. and the duration of PA activity. - Individual VO and HR relationships are -Not very accurate 2 From 5% to 20% established beforehand. (Measure HR and VO at -HR is affected by numerous factors Minute-by- 2 vs. indirect - Does not interfere with rest and during exercise) outside of exercise (i.e. fear) minute heart- calorimetry [99, daily activities -A heart rate recorder is worn, HR is measured - Estimation does not account for the fact rate recording 100] - Inexpensive minute by minute and the data are analyzed based that EE returns to resting levels more

on the VO2 + HR relationship. quickly than HR [100]. Gold Standard for short-term -Subject is placed in a thermally isolated chamber Valid and Reliable for - Expensive Direct controlled and heat emission is measured. short-term controlled - Not practical for long evaluation periods Calorimetry experiments. It is a experiments. - Only one person can be studied at a time. direct and accurate method.

27

Table 5: Determining Energy Expenditure - Strengths and Weaknesses of Current Methods Method Procedure Estimation Error Advantages Disadvantages Indirect -Respiratory-gas exchange is measured Valid and Reliable for Calorimetry is - Requires volunteer to be attached to a -Assuming that all of the oxygen consumed is used short-term controlled Indirect equivalent to breathing apparatus to oxidize fuels, and all of the carbon dioxide that experiments Calorimetry Direct Calorimetry - Not practical for long evaluation periods is produced is recovered, it is possible to calculate - Capable of assessing for rest and steady energy production. fuel utilization. state activity.[103] Gold Standard for - Subjects ingest known amounts of deuterium and long term “free oxygen-18 - Valid and Reliable living” - Expensive - Samples of body water (e.g. urine) are collected - Does not interfere with Doubly experiments. - Cannot determine EE for individual throughout the study period. daily activities Labelled - depending on the activities throughout the day. -The difference between the rates that deuterium - Relatively easy to Water (DLW) calculation used -Meant for longer evaluation periods (4 and oxygen-18 is lost corresponds to the rate that administer validation studies days to 3 weeks) carbon dioxide is produced from which EE can be suggest between calculated. 2% and 9% [104] EE: energy expenditure; VO2: oxygen consumption; HR: heart rate

28

2.4.1 Combined Methods to Estimate Energy Expenditure in the Military Population

Attaining accurate measurements of EE in military personnel engaged in various activities allows for better recommendations to be made about nutrient requirements. Ideally the DLW method would be employed whenever accurate measurements of EE are necessary. Measuring the entire sample or a subset of the group is incredibly valuable as the method’s reliability has been validated numerous times [86, 88, 101]. Understandably using the DLW method is not always possible especially when cost is a concern. In these situations, other methods can be used in combinations to increase the accuracy of estimation.

In long term studies where nutrient intake can be controlled and easily recorded, the intake-balance method is a valid option [88]. It is important that skilled personnel conduct all of the body composition measurements as slight inaccuracies can significantly decrease precision. When possible, it would be valuable to employ the factorial method to support the findings. With both methods, information is gathered about food intake, estimated EE and body composition changes, allowing for reasonable estimations to be made even if one area is lacking usable data [86]. In order to adequately employ the factorial method alongside the intake- balance method the energy costs of military tasks need to be known. While the Compendium of Physical Activities [105] (a compilation of various activities and their objectively measured energy costs) is an effective instrument for most populations, military relevant data are lacking with few applicable activities. In addition other sources are limited in their usefulness due to vaguely described activities and/or poorly defined results [106, 107] (i.e. data depicting only absolute energy costs without providing relative values that are more easily applied) thereby limiting the applicability of the data. In order to further improve the factorial method (and combination methods employing the factorial method) a compilation of military tasks and their respective energy costs needs to be created.

The use of heart rate monitors in combination with accelerometers may also be a great option. The accelerometers allow verification of movement with increases in heart rate, thereby decreasing the likelihood of EE overestimations as a result of non-movement increases in heart rate. Until now, there have only been a few studies that have employed this combination method. Thus far one study has shown matching results when compared to direct calorimetry during cycling [108]. Another study assessing EE of walking and running found group estimations to be very accurate (within 0.54%) when compared to direct calorimetry, individual

29

estimations were less exact (within 14%) [109]. In free living situations in sub-Saharan Africa, the combined method was less ideal in predicting physical activity EE by accounting for only 16% of the total variance measured by DLW [110]. The estimations were more accurate for urban populations than rural populations, likely due to the load-bearing tasks that are necessary in rural environments [110]. Lastly in a study measuring EE in young European men, estimations using the combination method were fairly comparable to DLW when the relationship between heart rate and exercise was individually calibrated [111] The estimations for all of these studies are more accurate on a group level than the individual level, making this combination method useful for making larger scale EE inferences.

In general combination methods need further validation before recommendation is warranted. In particular military specific combination methods should be devised to better estimate EE in the military population.

2.5 Are military personnel eating what is provided to them?

When taking into account all of the publicly available military studies and reports that described the amount of energy expended and the amount of food provided to military personnel during the documented missions, the majority of military groups were provided with military rations that supplied adequate caloric content assuming that all of the provided food was in fact consumed (11 groups were provided more than enough energy, 5 provided just enough, and 8 groups were insufficiently supplied) (Figure 6).

30

Figure 6: The amount of energy expended vs. the amount of energy provided during each military operation or training course. Each data point represents a particular military group as collected from publicly available military reports and journal articles. Red data points signify data conducted in Canadian Armed Forces members. The diagonal black line depicts energy balance with studies above the line portraying military groups who were provided food in excess of energy requirements, and those below the line portraying military groups who were provided with insufficient food [5-8, 11, 27, 34-37, 39, 59, 112].

Interestingly, even though many studies provide calorically sufficient military rations (Figure 6), military personnel consistently consumed an inadequate number of calories during field operations and exercises (Figure 7). This was also true for all available military research conducted in CAF members.

31

Figure 7: The amount of energy expended vs. the amount of energy consumed during each military operation or training course. Each data point represents a particular military group as collected from publicly available military reports and journal articles. Red data points signify data conducted in CAF. Data only includes studies where food restriction was not imposed. The diagonal black line depicts energy balance with studies below the line portraying military groups who consumed insufficient energy to match energy requirements [5-8, 11, 27, 34-37, 39, 59, 112].

In a recent Canadian study, it was found that 66% of the studied CAF consumed personal foods in addition to military rations. Although personal foods increased consumption by 800 kcal per day, soldiers were still not eating enough (~1000 kcal/day deficit)[39]. During military operations EE remains high even when military personnel grossly under eat because the mission directs the amount of energy expended regardless of intake. In a 28 day field study, EE between groups was virtually the same regardless of whether they received the ready-to-eat meal ration (4020 kcal/day) or the lightweight ration (1980 kcal/day) [35]. Considering that many of the military groups in Figure 7 were provided with sufficient (or excess) energy, it is important to consider why military personnel are not consuming what is provided to them.

2.5.1 Why does Voluntary Anorexia occur?

The under consumption of military rations is a complex problem which is dependent on several factors. The factors that are generally considered include: continuous or high intensity missions, palatability of food, inconvenient or lengthy food preparation, stereotypically negative 32

views of rations and intentional dieting [15, 113, 114]. During continuous operations, there is not always enough time to eat and prepare meals, furthermore exhaustion from such operations will at times result in soldiers choosing sleep, or the completion of a mission over food [4]. The palatability of combat rations is also of great importance because unappetizing food is less likely to be eaten. Even with relatively appealing rations, monotony is common and can arise in as little as a few days especially when the same ration is served repeatedly [114].Constant insinuations that combat rations are bland and unappetizing further lowers ration consumption regardless of the actual palatability of the ration, while others purposely under-eat, in an attempt to lose weight [15]. Although these factors certainly add to the under consumption of military rations, appetite, or rather the lack of appetite may also be a prevalent factor.

2.5.2 Appetite

Appetite typically defined as "a natural desire to satisfy a bodily need, especially for food" [115], is often assessed in a variety of ways: as a measure of food consumption (behavioural appetite), as a subjective perception of various aspects of appetite such as a feeling of fullness or hunger (psychological appetite), and more recently by the concentration of appetite regulating hormones (physiological appetite), largely in an effort to explain the behavioural and psychological actions and perceptions. Most appetite research will assess at least two of these factors as these factors are sometimes [116], but not always [117], highly correlated with one another.

2.5.2.1 Food Consumption

Arguably the most relevant measure of appetite is food consumption, as most interventions aim to ultimately target this aspect of appetite. The conscious and instinctive decisions surrounding food intake are not well established and are known to be affected by multiple factors that interact with one another [118]. Previous experience and knowledge regarding the availability and palatability of food as well as internal (sensory cues including perceptions of hunger, fullness etc.) and external (visual cues regarding how much was eaten, social and cultural contexts, etc…) cues can influence eating behaviour [118]. Appetite when described as voluntary food consumption, is often determined in one of two ways: 1) providing a single food item (pizza, or soup etc...) in excess of expected consumption and measuring the quantity the participant eats [119, 120], or 2) by providing participants an assortment of foods

33

(buffet style) in excess of expected consumption and documenting what foods in what quantities are consumed [121-123].

Although these methods accurately measure the amount of food consumed of the food that is provided, these simulated meals can themselves influence the amount consumed. Meal times for example are often scheduled which can artificially delay or advance meal timing and ultimately impact the amount of food consumed [124]. Similarly presenting only one food item or several has also been found to impact EI, with more variety leading to higher intake sometimes by as much as 40% [125]. In addition the palatability, appearance, texture, even the food labelling can impact which foods are selected and largely how much is eaten [118]. Resultantly as a measure of appetite, food intake measures alone can be fallible and are often used in combination with other appetite measures.

2.5.2.2 Subjective Assessment of Appetite

Subjective assessments of appetite often consider various indices of appetite (hunger or satiety sensations, pleasantness of food etc...), and are most often measured using questionnaires or visual analog scales.

Various categorical (multiple choice) questionnaires have been developed and used effectively in a variety of populations [126, 127]. One of the most frequently used appetite questionnaires (especially in clinical populations and the elderly) is the Simplified Nutritional Appetite Questionnaire (SNAQ) which asks questions about the typical frequency of food intake (< one meal a day, one meal a day, two meals a day, three meals a day, > three meals a day), the general taste of food (very bad, bad, average, good, very good), when during a meal the participant feels full (after a few mouthfuls, after about a third of a meal, after eating over half a meal, after eating most of the meal, I hardly ever feel full), and general appetite with questions such as "my appetite is" (very poor, poor, average, good, very good) [126]. Questionnaires are typically used to identify individuals with poor general appetite; they are not frequently used to assess instantaneous appetite sensations. Most appetite research tends to focus on transitory appetite sensations which are generally measured various times throughout the day and evaluated using visual analogue scales.

Visual analogue scales assess different sensations (hunger, satisfaction, fullness, prospective eating etc.), that together contribute to the assessment of appetite. Each sensation

34

scale is a 10 cm line anchored at each end with the extremes of that sensation (e.g. ‘I am completely empty’ on the left end of the “satisfaction” scale and ‘I can’t eat another bite’ at the right end). Paper and pencil as well as electronic versions of these scales can easily be administered [128]. Although variations between participants can be large, visual analogue scales have been found to be reliable, and relatively accurate and sensitive especially when used in repeated measures designs [129].

It is not clear whether there is an agreement between subjective appetite sensation and EI. While some studies report a correlation between visual analogue scale scores and EI [130], others report no correlation [131, 132]. In a recent systematic review assessing 462 papers it was reported that subjective appetite does not predict subsequent EI [133], however the vast majority of the papers (94%) included in this review did not assess the relationship between subjective appetite and food intake. As a result, the authors inferred whether a link between subjective appetite and intake was demonstrated based on whether or not the changes in subjective appetite and EI were in the expected direction. In addition it is unclear how far apart the appetite measurement was from food consumption. It has also been suggested that the strength of the subjective appetite and EI relationship is dependent on the circumstance employed; with stronger correlations found during structured experiments (especially fasting studies and those that incorporate pharmacological interventions), and weak correlations found under free-living conditions [134].

2.5.2.3 Physiological Regulation of Energy Balance

As obesity rates continue to increase, more and more research has focused on the physiological regulation of appetite in an effort to understand and manipulate the mechanisms involved and ultimately create weight loss aids. Although this area of research is still not well understood, a general summary of the current state of knowledge is provided.

The brainstem and hypothalamus receive and coordinate neural and hormonal signals in order to regulate body weight. Within the arcuate nucleus (ARC) of the hypothalamus, one group of neurons co-expresses the appetite stimulating neuropeptides: neuropeptide Y (NPY) and agouti gene-related peptide (AgRP) [135]. In addition to stimulating appetite, NPY and AgRP decrease EE, further promoting a positive energy balance, by decreasing the amount of energy brown adipose tissue and other thermogenic tissues expend [135]. The second group of neurons in the ARC co-expresses the appetite suppressing precursor protein pro- 35

opiomelanocortin (POMC) [136] and the neuropeptide cocaine and amphetamine-regulated transcript (CART)[135]. POMC and CART suppress appetite and stimulate EE. The ARC communicates with the paraventricular nucleus (PVN) and other centers in the brain which have roles in altering EE [135].

The median eminence, an area of the brain near the ARC of the hypothalamus has an incomplete blood-brain barrier; as a result, plasma hormones originating from the gut, pancreas, and adipose tissue are able to inhibit or stimulate the neurons of the ARC directly [137]. Alternatively, these plasma hormones can pass the blood-brain barrier at the area postrema of the medulla oblongata or be sensed by the vagus nerve. The signals from these gut hormones(along with signals from gastric stretch receptors and nutrient chemoreceptors) come together in the nucleus of the tractus solitarius (NTS) with information emanating from the higher brain centres, to ultimately regulate appetite through the ARC [137].

The appetite hormones that have an influence on this system are: leptin, insulin, ghrelin, glucagon-like peptide-1 (GLP-1), oxyntomodulin (OXM), peptide YY (PYY), pancreatic polypeptide (PP) and cholecystokinin (CCK) (Figure 8).

36

Figure 8: Hormonal regulation of appetite. Circulating hormones act on the Arcuate nucleus (ARC) of the hypothalamus denoted as the blue oval. Within the ARC are two important clusters of neurons: one that stimulates appetite and co-expresses Neuropeptide Y (NPY) and Agouti-related protein (AgRP) signified as the brown cluster; and one that suppresses appetite and co-expresses Pro-opiomelanocortin (POMC), and Cocaine- and amphetamine-related transcript (CART) signified as the orange cluster. These clusters interact with each other and are directly affected by circulating levels of ghrelin, peptide YY (PYY), Glucagon-like peptide-1 (GLP-1), insulin, leptin, and potentially oxyntomodulin (OXM). Cholecystokinin (CCK) and Pancreatic Polypeptide (PP) on the other hand affect the ARC indirectly via the vagus nerve and brainstem respectively. GHS= Growth hormone secretagogue receptor; Y2R=PYY Y2 receptor; IR=Insulin receptor; LEPR=Leptin receptor; GLP1R= GLP-1 receptor

2.5.3 Hormonal Appetite Regulation

Appetite is thought to be regulated by both: 1) long-term signals concerned with adiposity and health, and 2) short-term signals that play a larger role in meal initiation and cessation [138].

37

2.5.3.1 Long-term regulation

Leptin

Plasma leptin levels are proportional to total fat mass, decrease with increasing age, and are generally lower in men than in women [139]. Leptin is produced by white adipose tissue, it inhibits NPY/AgRP expressing neurons and stimulates POMC/CART expressing neurons leading to a decrease in appetite [138]. Decreased leptin levels on the other hand, are associated with increased appetite and individuals with leptin deficiency are obese [140]. Following these findings, several studies attempted to exploit leptin’s apparent effect on appetite by administering exogenous leptin to overweight and obese individuals in an attempt to aid weight loss. In some studies, weight loss was significantly greater with regular exogenous leptin administration [141, 142], however this was not always the case [143].

Furthermore, obese individuals tend to be leptin resistant with higher plasma leptin levels, but a weakened appetite-suppressing effect [144] suggesting that exogenous leptin administration for weight loss in this population is likely ineffective.

Although leptin typically plays a larger role in long-term regulation, large changes in energy balance also lead to changes in leptin levels and appetite. With fasting, dieting, and weight loss, plasma leptin levels decrease more than would be anticipated from fat loss alone, and appetite correspondingly increases [145]. It is suggested that this drop in plasma leptin plays a role in reinstating energy balance in the short-term as well.

Insulin

Insulin, is in many ways similar to leptin; blood levels of insulin are proportional to total fat mass [146], insulin inhibits NPY/AgRP expressing neurons and stimulates POMC/CART expressing neurons leading to a decrease in appetite [138], and increased body fatness is associated with decreased insulin sensitivity [146]. Insulin and leptin seem to be intimately connected, insulin regulates leptin production, and leptin suppresses insulin release [147].

In terms of appetite, the role of insulin seems more convoluted than that of leptin. In some studies, increased insulin levels are correlated with increased hunger, increased pleasantness of sweet taste, and increased food consumption [148]. In addition, in some

38

individuals insulin levels increase just with the thought or sight of food [148]. In others, insulin seemed to decrease appetite [149] or have no effect [150].

2.4.3.2 Short-term regulation

Ghrelin

Ghrelin is primarily synthesized in the stomach, and stimulates NPY/AgRP expressing neurons leading to increased appetite and food consumption [138]. Ghrelin may also act on the hypothalamus through the vagus nerve and the brain stem [151].

Due to its powerful appetite augmenting effects in humans [122, 152], ghrelin is thought to regulate day to day appetite as an effective meal initiator. Ghrelin, unlike leptin and insulin, is just as potent in obese individuals as it is in lean [152], however in obese individuals the typical decrease in ghrelin following a meal is not always apparent [153].

Ghrelin's role in appetite and food intake appears to be very robust. Ghrelin deficiency reduces EI, and ghrelin infusion increases EI. Ghrelin deficient mice fed a high fat diet were protected from weight gain, and instead, they displayed lower body fatness, and increased EE [154]. Similarly in humans, ghrelin administration has been found to increase food intake and subjective appetite scores without exception [122].

Ghrelin can be found in two forms, acylated and desacylated. The acyl form seems to be responsible for the appetite stimulating effects [155, 156], while the impact of the desacyl form on appetite and food intake has been equivocal. Some studies report no effect on appetite or food intake [155] while others suggest that desacyl ghrelin blunts acyl ghrelin's orexigenic effect [156], still others suggest that desacyl ghrelin independently inhibits food intake [157].

Glucagon-like peptide-1 (GLP-1)

GLP-1 is an appetite suppressing peptide that is released from the L-cells of the intestine. The amount secreted is in proportion to the amount of calories consumed, although GLP-1 can also be secreted as a result of meal anticipation [158]. GLP-1 directly stimulates POMC/CART expressing neurons, and indirectly (potentially through an inhibitory gamma-aminobutyric acid (GABA) neuron) suppresses NPY/AgRP expressing neurons [159].

39

Administering exogenous GLP-1 to healthy males resulted in decreased hunger, decreased food intake, and an earlier onset of fullness [160]. In obese males, exogenous GLP-1 resulted in decreased hunger, and delayed gastric emptying [161].

Oxyntomodulin (OXM)

OXM is an appetite suppressing peptide that is co-secreted with GLP-1 and is released from the L-cells of the intestine. Similarly to GLP-1, the amount secreted is in proportion to the caloric content of the food consumed [162]. The mode of action is not fully known, but it is suggested that OXM acts through the GLP-1 receptor considering that the effect of OXM on appetite is eradicated in GLP1 receptor knockout mice [163].

Both central and peripheral administration of exogenous OXM decreases EI and augments EE in rats [138]. Additionally, exogenous administration of OXM to healthy participants results in a decrease in EI and hunger ratings [162]. However, 24 h EI was no longer different from control, suggesting that OXM may not be a very potent appetite suppressor when administered in the physiological range.

Peptide YY (PYY)

PYY is an appetite suppressing gut hormone mainly released by the L cells of the distal part of the gastrointestinal (GI) tract. PYY inhibits the NPY/AgRP expressing neurons leading to decreased appetite and food consumption [138]. It is released within 15 min of food intake, and plasma concentrations peak within one to two h [164]. It is suggested that neural or other hormonal signals are responsible for activating the release of PYY, as the meal-triggered release occurs prior to food or nutrients reaching the L cells of the GI tract [165]. Plasma concentrations of PYY are also proportional to the caloric content of the ingested meal and remain elevated for several hours after the cessation of the meal [164].

PYY’s role in appetite suppression is convincing. PYY knockout mice eat voraciously, rapidly becoming obese; however this can be reversed with the exogenous administration of PYY [166]. Similarly administering exogenous PYY to humans decreases appetite and food intake, even in a fasted state [166-168]. Interestingly fasting and postprandial PYY levels are lower in obese vs. leans adults [166], but a PYY resistance does not seem to develop with obesity.

40

Pancreatic Polypeptide (PP)

PP is an appetite suppressing peptide mainly produced by the F-cells of the pancreas, and the colon. Unlike many of the other peptide hormones, PP cannot penetrate the blood-brain barrier at the ARC, but is able to enter near the brainstem at the area postrema [138]. PP has the greatest affinity for the Y4 receptor located in the brainstem and hypothalamus, although it can bind with all G-protein coupled Y receptors (Y1, Y2, Y4, Y5, Y6)[169]. PP affects some gastrointestinal actions (gallbladder and pancreatic functions, motility etc.) through the vagus nerve [170], and it is also suggested that PP can bind to Y4 receptors in the ARC and stimulate of POMC expressing neurons [171].

Similarly to GLP-1, OXM, and PYY, PP is released in proportion to the caloric content of the food consumed, and plasma PP levels can be elevated for six h following meal cessation [172]. Exogenous PP administration results in a reduction in EI. In mice, exogenous administration of PP not only decreased food intake, but also decreased gastric emptying and increased EE further encouraging a negative energy balance [173]. In humans, infusion of PP resulted in decreased appetite and EI [174]. Even more interestingly, this suppression of appetite and EI was maintained for 24 h [174].

Cholecystokinin (CCK)

CCK is produced predominantly in the duodenum and jejunum, and can decrease EI in lean and obese humans [175]. Like most of the other gut hormones, CCK plasma levels rise quickly, within 15 min of meal onset. CCK has two receptor subtypes: CCKA (in the gastrointestinal tract) and CCKB(in the brain), of which CCKA seems to be play a larger role in appetite and food intake [138]. Although exogenous administration of CCK decreases appetite and EI, these effects are short-lived, and compensatory behaviours soon follow, resulting in frequent meals [176]. CCK degrades quickly, and so to do the appetite suppressing effects. For a summary of all of the hormones and their effects on appetite see Table 6 below.

41

Table 6: Appetite hormone summary table. Hormone Site of Synthesis Effect on Appetite Leptin Adipose Tissue Decreases Appetite Insulin Pancreas Decreases Appetite Ghrelin Stomach Increases Appetite PYY Intestinal L-cells Decreases Appetite GLP-1 Intestinal L-cells Decreases Appetite PP Pancreas/Colon Decreases Appetite OXM Intestinal L-cells Decreases Appetite CCK Duodenum and jejunum Decreases Appetite

2.5.4 Effects of Exercise on Appetite

When assessing the impact of exercise on weight loss two recent meta analyses found that exercise alone has a limited effect on body weight [177, 178]. It was reported that moderate intensity aerobic training over 6-12 months resulted in very modest weight loss (<2kg) in previously sedentary, overweight, and obese participants [177]. Considering that exercise increases overall EE, it would be reasonable to expect that overall weight loss over 6-12 months would be greater that 2kg in this population, unless of course exercise also stimulates a compensatory increase in EI. These kinds of observations and an earlier study showing an increase in EI (to a degree that reasonably matched the energy expended) following 2 h of exercise [124] led to the assumption that exercise (and the energy deficit resulting from it) led to a compensatory drive to eat (and replenish the energy that was used).

Contrary to this sentiment there does not seem to be an increase in subjective appetite or food intake, within 24 h of an exercise bout [179]. When aggregating all of the studies that assessed the effect of acute exercise on subsequent EI (within 24 h of exercise) using a repeated measures design, it becomes clear that absolute EI does not generally vary between the exercise and control days (Figure 9). All of the data points fall near the line of equality suggesting that EI is relatively unchanged by an exercise bout lasting ≤ 2 h.

42

Figure 9: Average absolute energy intakes during both exercise and control trials as reported in 19 published repeated measures studies. Each data point represents the data collected within a study group as found in articles published between 1992 and 2017. Overall 34 study groups (where absolute energy intake during both control and exercise conditions was measured) were plotted. The diagonal black line depicts that the same amount of energy was consumed during both trials, with data points above the line portraying study groups who consumed more energy during the exercise trial, and those below the line portraying study groups who consumed more energy during the control trial [121, 123, 124, 132, 180-194].

Considering that EI was the same regardless of whether participants rested or exercised it is not surprising that relative EI (absolute EI - EE) is lower in the exercise trials (since exercise increases EE) in the same studies that are plotted in Figure 9. When adjusting EI for the amount of energy expended during exercise, the vast majority of study groups would have been in a more negative energy state during the exercise trials than during the control trials (Figure 10).

43

Figure 10: Average relative energy intakes (absolute energy intake - energy expenditure) during both exercise and control trials as reported in 19 published repeated measures studies. Each data point represents the data collected within a study group as found in articles published between 1992 and 2017. Overall 34 study groups (where absolute energy intake during both control and exercise conditions was measured and relative energy intake was reported or could be calculated) were plotted. The diagonal black line depicts that the same amount of relative energy was consumed during both trials, with data points above the line portraying study groups who consumed more relative energy during the exercise trial, and those below the line portraying study groups who consumed more relative energy during the control trial [121, 123, 124, 132, 180-194].

Even when assessing longer duration studies (<1.5 years) the overall impact of exercise on EI seems trivial. A systematic review evaluating the effects of exercise on EI reported that exercise (either acutely within 24 h of a single exercise bout, or over a 72 week span of exercise training) has no impact on EI [195].

While ample research indicates that exercise has no effect on EI, some studies have suggested that compensatory EI will occur following prolonged exercise (≥120 min). Although few studies employed prolonged exercise protocols, the two studies [124, 186] with the longest exercise bouts (2 h) did report an increase in absolute EI following exercise (thereby

44

compensating more completely for the completed exercise). It is possible that a larger exercise stimulus is required to impact EI.

Even when the amount of food consumed at the next meal is unchanged from that seen following rest, it appears that exercise results in a transient suppression of subjective appetite [196-198]. Although not always detected [132, 199, 200], the typical appetite progression following an acute bout of exercise results in decreased ratings of hunger and prospective eating and increased ratings of fullness and satisfaction during and immediately following exercise [123, 187, 201]. While relatively reproducible (in studies with adequate statistical power), this effect is short-lived and usually returns to baseline within an hour of exercise cessation [202].

The impact of exercise on appetite regulating hormones is still somewhat debatable, however the vast majority of studies report that exercise transiently decreases plasma acylated ghrelin levels, and increases plasma GLP-1, PYY, and PP levels [203]. These hormonal changes are also in agreement with what would be expected for appetite suppression, although correlations between hormonal changes and subjective appetite and/or EI are not always found [123, 131, 132, 194, 201, 204, 205].

2.5.4.1 Effects of Exercise on Appetite Regulating Hormones

Since ghrelin is the only clear orexigenic hormone to date, a lot of research has investigated the role ghrelin, specifically acylated ghrelin plays in appetite regulation. The majority of studies assessing the impact of exercise on acylated ghrelin concentrations have reported that acylated ghrelin concentrations decrease as a result of exercise. This response was found following various modes of exercise including running [187, 188, 194, 202, 204, 206], cycling [190, 198, 207], swimming [123], rope skipping [198], and resistance exercise [201]. Although frequently reported, this occurrence is not universal with some reporting no effect of exercise on plasma acylated ghrelin concentrations [131, 132, 200].

The effects of exercise on total and desacylated ghrelin are more ambiguous, for instance Erdmann et al. reported that low intensity cycling led to increases in total ghrelin concentrations [186], and Douglas et al. reported that desacylated ghrelin decreased as a result of treadmill exercise [131]; however these responses have yet to be repeated. Overall, exercise appears to transiently decrease acylated ghrelin levels, while the effects on total and desacylated ghrelin are equivocal (Appendix B: Table V).

45

Circulating levels of GLP-1 typically increase in response to exercise [131, 180, 200, 202, 207], although not always [198] (Appendix B: Table VI). It is not clear whether the GLP-1 increases following exercise are correlated with decreased sensations of appetite or EI as some have reported correlations between these variables [180], while others have reported no relationship [131, 205]. Similarly to GLP-1, most studies also suggest that acute exercise increases circulating PYY levels [131, 180, 190, 198, 200, 202], however others reported no effect [117, 207]. It seems that increased body fatness can suppress the rise in PYY following exercise considering that the increment in PYY is more pronounced in lean vs. obese individuals [131]. In addition, other factors including the mode of exercise, and sex appear to impact PYY concentrations differently. Although circulating PYY increases following aerobic exercise, there is no effect of resistance exercise on PYY concentrations [201]. Similarly the effect of exercise may be altered by sex, as Hagobian et al found no effect of exercise on PYY levels in men, but did report a transient increase in women [132](Appendix B: Table VII). The effects of exercise on PP concentrations have not been studied as often as PYY, GLP-1, and ghrelin (Appendix B: Table VIII), however exercise (especially aerobic exercise) has been found to increased plasma PP concentrations [192, 208]. Leptin, insulin, CCK, and OXM on the other hand are generally deemed to be uninvolved in exercise-induce anorexia. Leptin is occasionally measured, however since it is considered a marker of adiposity and a factor in long-term appetite regulation, it is usually omitted in studies assessing acute appetite responses following exercise. The available research suggests that acute exercise does not have any effect on circulating leptin levels [181, 188, 209, 210], as previously reported changes in leptin concentration (following acute exercise) can be explained by hemoconcentration or circadian rhythms [211]. Large changes in energy balance as a result of fasting or heavy exercise (e.g. marathon running) however seem to decrease leptin concentrations [211, 212]. Some have even suggested that leptin could be a potential marker for overtraining [213] (Appendix B: Table IV). Similarly insulin is rarely discussed in reference to appetite, partly because insulin has been suggested to both increase [148] and decrease [149] appetite, ultimately suggesting that the role of insulin in appetite is negligible [214]. CCK and OXM on the other hand are rarely considered as important agents in regards to exercise induced changes in appetite as their effects

46

are short-lived and often followed by compensatory behaviours thereby being poor candidates to explain exercise induced anorexia [162, 176].

2.5.5 Effects of Ambient Temperature on Appetite

It is thought that environmental temperature can also affect consumption, as food intake has been found to be low in the heat and high in cold environments in various animal models [215-217]. Although the data are sparse in human populations, earlier cross-sectional data on military subjects suggest that the same trend also appears in humans [218]. In human laboratory studies, acute (≤48 h) changes in ambient temperature 31°C vs. 22°C [192], 27°C vs. 22°C [219], ~19.5°C vs. ~26.5°C [220], 16°C vs. 22°C [221], or 18°C vs. 24°C [222] do not significantly impact subjective appetite or EI, although is some cases a trend in the expected direction appears [219].

Few studies have examined the effect of ambient temperature without exercise on appetite-regulating hormones. Acute (30-100 min) cold (2-6°C) exposure has previously been reported to decrease plasma leptin [223, 224], and CCK [223] concentrations, and increase plasma ghrelin levels [225]. Whereas acute (30-100min) heat (30-31°C) exposure has been reported to decrease plasma ghrelin levels [192, 225] and have no effect on PP, or CCK [192].

2.5.6 Effects of Ambient Temperature and Exercise on Appetite

To date, nine studies have in some form assessed the effects of exercise in varying temperatures on appetite [181, 192, 226-232] (Table 7). Of these, 6 assessed land-based exercise, 2 examined water-based exercise, and one evaluated the impact of water immersion (both cold and thermoneutral water) immediately post-exercise on EI, appetite sensation, and/or the concentration of appetite-regulating hormones.

2.5.6.1 Energy Intake

The effects of land-based exercise in different ambient temperatures on EI are ambiguous with only one study reporting an increase in EI following exercise in cold vs. temperate conditions [229]. While no other significant differences in EI following exercise in hot vs. temperate [192, 226] or cold vs. temperate [226] conditions were found, Wasse et al. report that there was a trend (p=0.08) towards higher intake in the cold and lower intake in the heat [226].

47

Comparatively, EI increases following water-based exercise in cool water (20-22°C) vs. in thermoneutral to warm water (33-34°C), exercise on land (24°C), or while resting at room temperature [230, 231]. The effects of exercise in thermoneutral (33-34°C) water however are not as clear, with some reporting decreases in EI [231], while others report no change in EI vs. control [230].

Despite the paucity and relative ambiguity of the data, a general tendency towards an increase in EI with exercise in the cold appears.

2.5.6.2 Appetite Sensation

There are only three studies that assess the impact of exercise on appetite sensation in varying ambient temperatures. Although few studies exist, an inverse relationship between appetite and temperature has been reported, with exercise in higher temperatures suppressing appetite and vice versa. Hunger and prospective food consumption decrease as a result of exercise in the heat vs. exercise in a thermoneutral environment [226], whereas exercise in the cold vs. in a thermoneutral environment generates lower satisfaction [226] and higher hunger and motivation to eat [227] scores. Even in the sole study where a significant difference between conditions could not be reached, the authors reported a trend (P=0.073) towards appetite suppression with exercise in the heat [192].

2.5.6.3 Appetite-Regulating Hormones

While the ambient temperature exercise is executed in does not seem to have any effect on plasma leptin or PP concentrations [181, 192, 228], the data regarding plasma ghrelin and PYY concentrations are equivocal.

Although some have reported that exercise in a hot environment decreases total ghrelin levels [192], and exercise in the cold increases acylated ghrelin levels [229] relative to exercise in a thermoneutral environment, most studies report no effect of ambient temperature on ghrelin’s response to exercise regardless of whether acylated ghrelin [181, 226, 228], desacylated ghrelin [228], or total ghrelin are measured [229].

Similarly PYY concentrations were reportedly unchanged regardless of whether exercise was completed in a: cold vs. thermoneutral environment [227, 229], or hot vs. thermoneutral environment [227] whereas Shorten et al. reported that PYY concentrations were higher when exercise was performed in the heat, vs. resting in a thermoneutral environment [181]. PYY 48

concentrations remained higher following the post-exercise buffet meal in the hot environment vs. both of the thermoneutral conditions (rest and exercise) [181].

Altering body temperature immediately post-exercise via cold, or thermoneutral water immersion, vs. control (resting at room temperature) has no effect on plasma leptin, acylated ghrelin, PYY, or PP concentrations [232].

Based on these results, the effects of exercise in hot and cold environments on appetite- regulating hormones are inconclusive. Few studies have been conducted, and none have yet assessed GLP-1 concentrations.

49

Table 7: Effects of exercise and ambient temperature on appetite regulating hormones. Subjects Timeline Exercise Temperature Energy Intake Appetite Leptin Ghrelin PYY PP CCK Reference - 07:00 arrival ↑ PYY in - 07:15 → Hot vs. either: control. -EI in 1) rest (Control) Hot: 36°C, Temperate was - ↓ leptin or Treadmill with 30% RH - No effect - ↑ PP 11 higher than after 2) exercise running of condition ↑ PYY after buffet healthy, control. buffet (Hot) or for 40 min Temperate: NA on acylated after meal; No NA [233] active - ↓ Relative EI meal; No 3) exercise at 70% 25°C with 30% ghrelin buffet difference males during Hot than difference (Temperate) V̇ O max RH meal, in between 2 it was in between hot vs. conditions control. conditions - 08:30 buffet- temperate type breakfast and provided control Appetite sensation - ↓ Total 06:30 arrival did not -No significant ghrelin in 1) - rest-22 Hot: 31°C with reach Cycling differences in EI hot ↑ PP ↑ CCK in 2) - rest-31 45% RH significanc 10 on a cycle between conditions during rest 3) - ex-22 e (p = healthy, ergometer conditions. vs. exercise conditions 4) - ex-31 0.073), but NA NA [192] active for 40 min - ↓ Relative EI temperate. conditions following -30 min Temperate: trended men at 60% during exercise - No effect vs. rest buffet recovery 22°C with 45% toward V̇O max conditions than of exercise conditions. meal. -buffet style 2 RH suppressio rest conditions. on total meal provided n ghrelin in the hot conditions. 09:00 arrival - ↓ AUC 1) exercise hot Hot: 30°C 50% values for 2) exercise Treadmill - trend towards 11 RH hunger and - No effect temperate running ↓ EI (~335kcal) healthy, prospective of condition for 60 min In hot vs. NA NA NA NA [226] active Temperate: food on acylated Cold buffet- at 65% temperate males 20°C 50% RH consumptio ghrelin style meals V̇ O max (p=0.08). 2 n in hot vs. provided at temperate. 11:00 and 14:30

50

Table 7: Effects of exercise and ambient temperature on appetite regulating hormones. Subjects Timeline Exercise Temperature Energy Intake Appetite Leptin Ghrelin PYY PP CCK Reference 09:00 arrival 1) exercise cold 2) exercise Treadmill Cold: 10°C - trend towards - ↓ AUC 10 - No effect temperate running 50% RH ↑ EI (~347kcal) values for healthy, of condition for 60 min in cold vs. satisfaction NA NA NA NA active on acylated Cold buffet- at 65% Temperate: temperate in cold vs. men ghrelin style meals V̇ O2max 20°C 50% RH (p=0.08). temperate. provided at 11:00 and 14:30 - ↓ Hunger and motivation to eat with exercise in all conditions - Greater ↓ Hot: 36°C 40% - ↓ ghrelin Hunger Cycling RH following and Exercise on a cycle exercise in - No 11 motivation conditions ergometer Temperate: all trials. effect of healthy NA to eat in NA NA NA [227] followed by 30 for 30 min 24°C 40% RH condition men hot and min recovery at 65% - No effect on PYY temperate V̇ O max Cold: 12°C of condition 2 than in 40% RH on ghrelin cold - ↑ Satiety with exercise in hot and temperate but not in cold.

51

Table 7: Effects of exercise and ambient temperature on appetite regulating hormones. Subjects Timeline Exercise Temperature Energy Intake Appetite Leptin Ghrelin PYY PP CCK Reference - No effect of condition Hot: 33°C, on leptin Cycling - No effect 60% RH on a cycle of condition Exercise - ↓ leptin ergometer on acylated 11 active followed by 3 h Temperate: 20 post- for 60 min NA NA or NA NA NA [228] males rest at 22°C, in °C, 60%RH exercise at 60% desacylated supine position and 3 h V̇ O max ghrelin 2 Cold: 7°C, post-

60%RH exercise compared to pre- exercise 45 min rest period at room - ↑ acylated temperature ghrelin 16 during cold overweig Exercise Treadmill Temperate: 20 - No vs. ht intervention walking °C, 40%RH - ↑ EI following effect of temperate participan for 45 min exercise in cold NA NA condition NA NA [229]

ts (10 45 min rest at 60% Cold: 8°C, vs. temperate on PYY - No effect male, 6 period at room V̇ O max 40%RH 2 of condition female) temperature on total followed by a ghrelin buffet-style meal - ↑ EI following exercise in cool Exercise/Rest Cool water: Cycling water vs. intervention 22°C on a cycle control and followed by an ergometer exercise on 6 trained hour of rest Warm water for 30 min land. NA NA NA NA NA NA [231] males during which an 34°C at 70% - ↓ EI following assortment of V̇ O max exercise in sweet food was 2 Thermoneutral warm water vs. available land 24°C all other conditions

52

Table 7: Effects of exercise and ambient temperature on appetite regulating hormones. Subjects Timeline Exercise Temperature Energy Intake Appetite Leptin Ghrelin PYY PP CCK Reference - ↑ EI following cold and 07:00 arrival Control (air): thermoneutral 07:15 exercise 20°C water Following Treadmill immersion vs. - No effect - No effect - No - No effect exercise running 10 active Cold water: control. of of condition effect of of control/water for 40 min NA NA [232] males 15°C No difference in condition on acylated condition condition immersion for at 70% EI between cold on leptin ghrelin on PYY on PP 20 min V̇ O max 2 Thermoneutral and 08:30 buffet- water: 33°C thermoneutral style meal immersion trials. - ↑ EI after exercise 30-min rest in cold water vs. Control (air): - followed by exercise in Cycling 25°C exercise/rest neutral water on a cycle intervention and resting ergometer Cool water: 11 males -20 min control NA NA NA NA NA NA [230] for 45 min 20°C recovery -No difference at 60% -followed by 1 in EI between V̇ O max Thermoneutral h buffet style 2 exercise in water 33°C meal neutral water and resting control

53

2.6 Summary of Literature Review

The preceding literature can be summarized in the following manner:

1) Maintaining a normal body weight is associated with improved health and performance outcomes while decrements in both health and performance are found with both increased weight (obese), and decreased weight (underweight) categories. EI and EE can both significantly affect body weight in either direction. In the military population weight loss is more likely to occur as military work can be very physically arduous.

2) Gradual weight loss (<3% body weight over 7 to 30 days) has minimal effects on perceptible health measures and working capacity, whereas both rapid (1%-3% body mass over <1 week), and considerable (>10% of initial weight) weight loss can impair physical performance and health. As a result weight losses should be kept below 3% of initial body weight, and soldiers should be given enough food and water to prevent dehydration.

3) In the majority of documented military field scenarios lasting 4-12 days (where the provided rations were meant to maintain energy balance), military rations do not lead to weight losses that exceed 3% of initial weight, however rapid weight loss (1%-3% body mass over <1 week) does occur. During longer military field operations (>60 days) weight losses above 3% body mass are more common.

4) Currently the caloric content of the food typically supplied to CAF during general field operations and training (the baseline 3 IMPs from which the rest of the field feeding program is formed around) is based on the Estimated Energy Requirement (EER) predictive equations. These equations were devised from a database that contained the EE of 407 healthy adults and specifically excluded military personnel and other highly active individuals (e.g. elite athletes).

5) The current procedures available to food service providers to increase the quantity of food supplied to CAF personnel in the field is based on financial rather than physiological factors. Considering the potential health and performance implications of under-supplying and the financial waste of over-supplying military rations, a physiological foundation from which food provision procedures can be determined for CAF field operations is needed.

6) Total energy expenditure (TEE) is composed of basal metabolic rate (BMR), thermic effect of food (TEF), thermoregulation, and physical activity. Of these physical activity and

54

thermoregulation may be of significant importance to the military population due to the physically demanding nature of the work and variety of environments that military personnel are required to operate in.

7) EE can vary greatly amongst military groups since the physical activities performed can differ greatly based on their trade and mission requirements. As such, EE of military personnel in the field has been reported to range between 2332 kcal/day and 6353 kcal/day in a temperate climate.

8) The thermal environment can also affect EE with potential increases in EE at both ends of the temperature spectrum. Although the effects of milder heat stress on EE are unclear, EE is increased when exercise is conducted in more severe heat stress conditions. In cold environments EE increases when individuals feel cold or start shivering. Usually however, soldiers wear enough clothing in cold environments; this clothing may affect the efficiency of movement due to its weight and bulkiness and consequently increase EE.

9) Research studies conducted with military personnel indicate that the average EE during field operations are as follows for selected environments: temperate climate: 3838 kcal/day (greater portion of females in this group as compared to the other two), hot environment: 4283 kcal/day, and cold environment: 4910kcal/day.

10) There are eight different types of EE determination methods each with varying degrees of accuracy: Questionnaires, intake-balance method, factorial method, motion sensors, heart rate recording, direct calorimetry, indirect calorimetry, and DLW.

11) The DLW technique is the most valid, reproducible, and non-intrusive method to use to estimate EE in military personnel during field operations. The cost associated with this technique may be prohibitive for applications to large numbers of studies or subjects. Alternative non-intrusive methods suffer from varying levels of validity and reproducibility, thus demonstrating and/or improving the validity of alternative non-intrusive techniques during field operations in military personnel is warranted.

12) Field feeding goes beyond difficulties with matching appropriate nutritional sustenance to physiological need; military personnel engaged in field operations consistently under eat regardless of the amount of food available to them. The under consumption of military

55

rations is currently accredited to: the palatability of food, inconvenient or lengthy food preparation, stereotypically negative views of rations, and intentional dieting. Although these factors certainly add to the under consumption of military rations, appetite, or rather the lack of appetite may also be a prevalent factor.

13) Appetite is assessed in three ways: 1) as a measure of food consumption, 2) as a subjective perception of various aspects of appetite such as a feeling of fullness or hunger, and by 3) the concentration of appetite regulating hormones - in an effort to explain food consumption and appetite perceptions.

13) Physiological appetite regulation occurs as a result of neural and hormonal signals that inhibit or stimulate neurons of the arcuate nucleus (ARC) of the hypothalamus. Seven different hormones (leptin, insulin, glucagon-like peptide-1 (GLP-1), oxyntomodulin (OXM), peptide YY (PYY), pancreatic polypeptide (PP) and cholecystokinin (CCK)) act to suppress appetite, and only one (ghrelin) hormone can stimulate appetite. While all eight hormones can impact appetite, three hormones (insulin, OXM, and CCK) seem to hold more minor roles.

14) Exercise has previously been reported to impact appetite in three ways: 1) although acute exercise is generally reported to have no effect on absolute caloric intake (thereby resulting in a negative relative energy intake), some studies have suggested that compensatory EI will occur following prolonged exercise (≥120 min). 2) In regards to the perception of appetite, the typical appetite progression following an acute bout of exercise results in decreased ratings of hunger and prospective eating and increased ratings of fullness and satisfaction during and immediately following exercise. 3) Acute exercise can also impact various appetite- regulating hormones; although the effects of exercise on leptin are unclear, circulating levels of GLP-1, PYY, and PP increase, whereas concentrations of acylated ghrelin decrease in response to exercise.

15) The impact of ambient temperature on appetite is ambiguous. Various animal models suggest that energy intake increases in the cold and decreases in hot environments. Sedentary human laboratory studies lasting (≤48 h) have not been able to replicate these findings instead suggesting that ambient temperatures between 16°C and 31°C have no effect on EI or appetite perception. Few studies have assessed the impact of ambient temperature on appetite-regulating hormones with some suggesting that acute cold exposure decreases leptin and CCK and

56

increases plasma ghrelin concentrations. Heat exposure on the other hand has been reported to decrease ghrelin concentrations and have no effect on PP or CCK.

16) Very little is also known about the impact of exercise in various thermal environments on appetite. Land-based exercise has been reported to have no impact on EI, whereas exercise in cold vs. thermoneutral water increases EI. Exercise in various thermal environments does however appear to impact appetite perception, signifying an inverse relationship between subjective appetite and temperature with higher temperatures suppressing appetite and vice versa. In regard to appetite-regulating hormones, exercise in different thermal environments seems to have no effect on plasma leptin, or PP concentrations, whereas the effects on plasma ghrelin and PYY concentrations are equivocal. To date, no study has assessed the impact of exercise in various thermal environments on GLP-1 concentration.

57

Chapter Three: Dissertation Objectives and Hypotheses

3.1 Energy Cost of Infantry Tasks

Having the ability to estimate EE in CAF members allows food service providers to more effectively determine how much food to provide military personnel during field operations. One potential method to estimate EE that can be employed either independently or in combination with other methods is the factorial method (a method that involves the accurate recording of all daily activities and matches the activity and its duration with previously documented energy costs of those activities from relevant and reliable sources such as scientific journals or technical reports). Unfortunately the energy costs of many military tasks are currently undefined, with the Compendium of Physical Activities (a compilation of physical activities and their measured energy costs as found in numerous published studies) [105] having sparse military relevant data, and additional sources having vaguely described activities and/or poorly defined results [106, 107] (i.e. data depicting only absolute energy costs without providing relative values that are more easily applied) thereby limiting the applicability of the data. As a result, the aim of the first study (The Energy Cost of Various Infantry Tasks) was to provide a compilation of military tasks and their associated energy costs in an effort to ease the estimation of EE in military personnel.

Every effort was made to estimate the energy cost of each military task using currently available sources, both the Compendium of Physical Activities, and military reports. It is hypothesized that where comparable activities are found, the energy costs of the completed military activities will be similar to the published values found in military reports, while the Compendium of Physical Activities is expected to underestimate the energy cost of military tasks. EE is expected to be higher than reported in the Compendium due to the heavy military clothing (~13kg) CAF are required to wear.

3.2 Ambient Temperature and Energy Expenditure

Military operations often include physically demanding activities of varying durations in thermally stressful environments. While it is currently unclear whether caloric supplementation is necessary in harsh thermal environments, others have speculated on the effects of thermally stressful environments on EE and the results of related research have been equivocal. Such knowledge is important for several reasons including understanding the nutritional consequences

58

and whether the provision of increased nutrition supplies is warranted. As a result, in the second study (The Effect of Ambient Temperature on the Energy Cost of Infantry Activities) the same military tasks were completed in three different environmental conditions: Hot (30.1 ± 0.2°C, 31.3± 1% RH), Temperate (21.0 ± 0.2°C, 32 ± 4% RH), and Cold (-10.4± 0.4°C, 56 ± 3% RH).

It is hypothesized that the energy cost of a range of activities executed in three different ambient temperatures: hot, temperate, and cold will be similar. EE measured during activities in the cold conditions might be significantly higher due to clothing weight, but the difference (if there is one) will likely not be large enough to be meaningful.

3.3 Military–Specific Algorithm

Although the factorial method can assist in the estimation of EE in military personnel, wearable technology has the potential to be more objective and accurate. As a result, the aim of the third study (Estimating Energy Expenditure in Military Personnel Using Accelerometry and Heart Rate) was to develop a military specific EE algorithm that could be applied to various military endeavours occurring in ambient temperatures ranging between -10°C and 30°C.

It is hypothesized that previously developed algorithms will underestimate EE in the current study due to the omission of clothing weight and ambient temperature data. As a result, the developed algorithm is expected to outperform past algorithms.

3.4 Appetite

Field feeding goes beyond difficulties with matching appropriate nutritional sustenance to physiological need; military personnel engaged in field operations consistently under eat regardless of the amount of food available to them [3, 11, 12]. Considering that voluntary anorexia can ultimately thwart any progress achieved by food service providers aiming to optimize nutritional adequacy, the role of appetite (behavioural, subjective, and hormonally implied) in this observed phenomenon was explored. Since both high intensity exercise [121, 206] and environmental stressors [181] have previously been shown to suppress hunger by physiological means, it is likely that voluntary anorexia is at least partly mediated by hormonal responses. As a result, the objective of the fourth study (The Effects of Exercise and Ambient Temperature on Appetite and Energy Intake) was to elucidate whether voluntary anorexia is a consequence of hormonal responses affected by the physical and environmental stressors imposed on military personnel. 59

3.4.1 Energy Intake

Since previous researchers found increases in EI with 120 min of exercise, resulting in partial [186] and complete [124] compensation for the amount of energy expended, it is hypothesized that completing 240 min of arduous physical activity will result in an increase in EI in the active conditions (Hot, Temperate, and Cold) vs. the Sedentary condition.

It is also hypothesized that absolute EI will be higher when participants are completing infantry tasks in the Cold environment as compared to when they are completing the same tasks in the Hot or Temperate environments, since exercise in the cold has previously been reported to increase energy intake [229].

3.4.2 Subjective Appetite

It is hypothesized that subjective appetite will be highest in the Cold environment and lowest in the Hot environment as previously reported [226].

3.4.3 Appetite Regulating Hormones

Leptin has previously been found to decrease following arduous physical activity [212], as a result, it is hypothesized that the military activities will lead to lower leptin concentrations during the active conditions vs. Sedentary. Lower acylated ghrelin concentrations and higher GLP-1, and PYY concentrations are expected during the active conditions vs. Sedentary.

60

Chapter Four: General Methods

4.1 Research Ethics Board approvals

The study was reviewed and approved by both the University of Toronto Research Ethics Board (REB) attached as Appendix R), and the DRDC Toronto Human Research Ethics Committee (HREC) (attached as Appendix S).

4.2 Sample size calculations

4.2.1 Study 1 2 and 3 Energy Cost of Activities and Differences Between Conditions.

It was calculated that a minimum of 6 participants were required in order to detect a 5% increase throughout the 8-h period in EE in the cold trial as a result of the heavy cold weather clothing [234] with a statistical power of 80% and a 0.05 level of significance. This study also served as a pilot study addressing the energy cost of completing a standardized set of common infantry military activities and for the development of an algorithm to estimate EE using accelerometry and heart rate data.

4.2.2 Study 4 - Appetite

It was calculated that a minimum of 20 participants were required in order to detect the following comparisons with a statistical power of 80% and a 0.05 level of significance:

(1) Assuming that acylated ghrelin concentrations are similar to those found in the 2011 study by King et al., it was calculated that twenty participants are required to detect a 13% decrease in acylated ghrelin concentration during the first two h following exercise vs. rest [235].

(2) Assuming that polypeptide YY (PYY) concentrations are similar to those found in the 2011 study by King et al., it was calculated that three participants are required to detect a 13% increase in PYY concentration during the first two h following exercise vs. rest [235].

(3) Assuming that EI is similar to that found by Shorten et al. 2009, it was calculated that twelve participants are required to detect a 10% difference in EI between the four conditions [233].

61

4.3 Participants

In order to aid in participant recruitment, posters (Appendix C) were posted around Defence Research Development Canada (DRDC) Toronto Research Centre, and emails (Appendix D) were sent to the CAF population. Twenty-seven (27) active, male (n=21) and female (n=6) individuals were recruited from a convenience sample of the CAF population to participate in this study. Seven withdrew voluntarily after starting the study due to time constraints (n=3), reconsideration (n=3), or loss of contact (n=1). Two participants were also excluded due to their non-compliance with the study restrictions. 18 participants (14 male, and 4 female) completed the study. Participants were between the ages of 18-60 years and were either Regular Force CAF or Class A reservists (Table 8).All participants were free from metabolic and cardiac disorders, were not taking any medications or natural health products and agreed to abstain from: 1) alcohol consumption for at least 24 h prior to each trial, 2) caffeine consumption for at least 15 h prior to each trial, and 3) exercise – activities above those of daily living, such as running, swimming, cycling, weight lifting, etc. for at least 24 h prior to each trial.

Table 8: Participant characteristics

Participants Sex Age Height Body BMI Body V̇ O2 Peak Peak Heart (years) (cm) Mass Fat (ml · min-1 · kg-1) Rate (bpm) (kg) (%) 14 males 33.5 ± 173.8 ± 80.4 ± 26.5 ± Completed 23 ± 8 44 ± 6 185 ± 14 4 females 10.8 10.3 15.7 3.8 7 males 28.9 ± 171.0 ± 75.3 ± 25.8 ± Withdrawn 21 ± 9 44 ± 6 189 ± 7 2 females 6.4 9.7 8.8 3.4 21 males 32.0 ± 172.8 ± 78.7 ± 26.3 ± All 22 ± 8 44 ± 6 187 ± 12 6 females 9.7 10.0 13.8 3.6 Data are presented as means ± SD. There were no significant differences between individuals who completed the study (completed), and those who did not (withdrawn).

62

4.4 Study Design

4.4.1 Study Overview

Figure 11: A graphical depiction of the study design. Black arrows denote visits to the laboratory whereas green arrows depict the days when participants collected data at home.

The study design is depicted in Figure 11. Participants were required to make 11 visits to the DRDC Toronto laboratory facilities –which include temperature and humidity-controlled chambers for human experimentation as well as an exercise physiology laboratory. The first visit consisted of the completion of the informed consent form, PAR-Q+, and other questionnaires.

During Visit 2, participants completed a maximal aerobic power test (V̇ O2max) on a treadmill. During Visit 3, participants gave a fasting blood sample, and had their body composition assessed via air-displacement plethysmography (BOD PODTM, COSMED, Rome, Italy). Visits 4-11 are the experimental trials, and the breakdown is as follows:

Visits 4 and 5 → Testing Condition 1 Visits 6 and 7 → Testing Condition 2 Visits 8 and 9 → Testing Condition 3 Visits 10 and 11 → Testing Condition 4 Each testing condition occurred on two consecutive days and was at least one week after the previous testing condition (Figure 11). The first day of each testing condition was a long 8-h day in the environmental chamber, and the second day of each testing condition was a short follow up visit. Participants completed a total of 4 different experimental conditions, one resting condition (Sedentary) where they rested for 8 h; and three active conditions where they completed two 2-h circuits composed of typical military tasks (covering a range of light, moderate and heavy work rates) dispersed with two bouts of 2-h rest. Each of these three active conditions was identical 63

except that it was completed at a different temperature: once at 30°C, RH 30% (Hot), once at 21°C, RH 30% (Temperate), and once at -10°C, RH 30% (Cold). Procedures for Testing Conditions 2, 3, and 4 were identical to Testing Condition 1, with the following exception: participants completed a different trial. If participants completed Hot in Testing Condition 1, they will complete Temperate, Cold, or Sedentary during Testing Condition 2. If they complete Sedentary during Condition 2, they will complete Temperate, or Cold, during Testing Condition 3, and so on. Whichever condition participants are not exposed to in Testing Conditions 1, 2, and 3 was the condition they underwent in Testing Condition 4. The Testing Conditions were completed in a repeated measures design; as a result, all of the participants underwent every condition. The order of these trials was assigned in randomized manner using a computerized number generator https://www.random.org/.

4.4.2 Visit 1: Questionnaire Completion

During the first visit all potential participants were fully informed of the details, discomforts and risks associated with the experimental protocol before being asked for their written informed consent. Participants were then asked to read and complete the written informed consent form (Appendix E) and the invasive procedures consent form (Appendix F). Once consent had been received, participants were asked to read and complete the Physical Activity Readiness Questionnaire-Plus (PAR-Q+) form to determine if they were fit for exercise without requiring secondary clearance by a physician (Appendix G). Participants were also asked to complete questionnaires regarding: participant demographics (Appendix H), Sleep Quality as per the Pittsburgh Sleep Quality Index (PSQI) (Appendix I) and physical activity (Appendix J).

4.4.3 Visit 2: Maximal Aerobic Power (V̇ O2max)

During Visit 2, maximal aerobic power (V̇ O2max) was determined using a progressive incremental exercise test to voluntary exhaustion on a treadmill (for details see "Measurements" section). Individual maximal heart rates (using transmitter/telemetry unit (Polar Electro PE3000, Finland)), and Borg’s scale [236] (Appendix O), ratings were also recorded during this test.

Upon completion of the treadmill exercise test, participants received both a 3-day weighed food record (3DwFR) and a smart-phone application (app) to record their food intake for three days. Participants were shown how to use the app and also received written instructions

64

on how to record their food intake (Appendix K). Participants were instructed to begin recording their food intake the morning following session 2, for a total of 3 days. During the first two days of the 3-day period, participants also collected 24-h urine samples for the measurement of urinary free cortisol, nitrogen, potassium and sodium. Participants were asked to store their urine containers in a cool place (ex. garage, fridge) until Session 3 when they were asked to bring their urine containers with them to the laboratory.

4.4.4 Visit 3: Body Composition and Baseline Data Collection:

Three days after session 2, participants reported to the laboratory after a 10 h overnight fast with their food records and their 24-h urine collection containers. One blood sample (13 ml) was taken via venipuncture for the measurement of serum ascorbic acid, urea, sodium, potassium, creatinine, cholesterol and lipids. Participants had their body weight measured using a standard scale. Then body composition was assessed via air-displacement plethysmography (BOD PODTM, COSMED, Rome, Italy), a commonly used technique to assess the percentages of body weight that consist of fat and lean body mass (for details see "Measurements" section).

During this visit, participants were also randomized to either a 3DwFR or a smart-phone app to record their food intake for three days. Participants were required to record their intake on several days throughout the study (as described below). Whether participants received a 3DwFR or a smart-phone app was determined in a randomized manner. Whichever method they received on this session, is the method they used for all subsequent visits.

4.4.5 Visits 4-11: Experimental Trials

For the two days prior to visit 4, 6, 8, and 10 participants were given military rations to sustain them for those two days (2 breakfast, 2 lunch, and 2 dinner rations). They were asked to document their food intake (using a 3DwFR or smart-phone app) and consume only the provided military rations. Participants were able to consume water ad libitum up until 10 pm the night prior to visit 4, 6, 8, and 10. One day prior to visit 4, 6, 8, and 10 and during session 4, 6, 8, and 10, participants also collected urine in 24-h urine collection containers. Participants brought their 24-h urine containers with them during visits 4-11. Participants brought the urine collected the day prior to visit 4 with them on visit 4; they brought the urine they collected during visit 4 and the evening of visit 4 with them on visit 5 and so on.

65

4.4.5.1 Visit 4, 6, 8, and 10: Chamber Days

Participants swallowed a telemetric core temperature capsule with 250ml of water at home at 6:00am and reported to the laboratory at 7:45am having fasted (not consumed food or liquids – except for the 250ml of water to ingest the core temperature capsule) since 10pm the previous night. Participants had their body weight measured using a standard scale and then they sat quietly (in the exercise laboratory) for 10 min and had their resting blood pressure measured using an automated blood pressure cuff. Their core temperature was also documented, since both a normal resting blood pressure (below 140/90) and a normal temperature (below 37.7°C) were required for the participants to participate on that day. If the participants fell within the normal ranges on that day, a venous catheter was inserted into an antecubital vein (or best available forearm vein) and a blood sample was obtained, an additional 8 blood samples were obtained throughout visit 4 (Figure 12). More detailed information about the amount of blood being drawn can be found in the section on Blood Samples. Visual analogue scales for appetite (VASA) (Appendix L), and the Profile of Mood States (POMS) were also administered (details can be found in the "Measurements" section).

Participants were given typical military rations to sustain them for the day (1 breakfast, 1 lunch and 1 dinner ration), providing ~4000 kcal and approximately 627g of carbohydrate, 191g of fat, and 108g of protein. There were 18 different rations to choose from (6 breakfast, 6 lunch, 6 dinner), and the participants selected 1 breakfast, 1 lunch, and 1 dinner ration which was kept the same for all of the chamber days (Visits 4, 6, 8, and 10) for that participant.

Following the fasting blood sample and fasting VASA assessment participants were able to consume breakfast. Rations could be consumed ad libitum from that point until 10:00pm that night. All food and drinks that were consumed were documented by the participant using the method (3DwFR or smart-phone app) they received during visit 3, and while in the lab by the investigators as well. Participants were also able to discard any portion of the rations. Discarded portions were given to the investigators for weighing and documentation.

Participants then outfitted themselves in the temperature appropriate military clothing for the environmental conditions they were going to endure that day. Participants also wore a portable metabolic measurement system (CORTEX Biophysik GmbH, Leipzig, Germany), a heart rate telemetry chest strap (Polar, Kempele, Finland), and a receiver for the core

66

temperature pill (Equivital™, Hidalgo, Cambridge, United Kingdom) (Details on all instrumentation can be found in the "Measurements" section).

The participants then entered the environmental chamber in which room temperature and relative humidity were controlled for the day’s duration. The environmental chamber was 5.9 metres long and 4.5 metres wide.

Every hour, participants were asked to rate their thermal comfort level (Appendix N). Following entry into the chamber, participants either rested for 8 h (S) or they underwent two 2- h circuits composed of typical military tasks (covering a range of light, moderate and heavy work rates) dispersed with two bouts of 2-h rest (C, T, and H) (Table 9).

67

Table 9: Battery of standardized infantry tasks that were completed in the environmental chamber. Activity Detailed Description of Activity Participants walked at the following 3 speeds: o 2.4 km/h, o 4 km/h, and o 5.6 km/h, At the following 3 grades: Walking on a motorized o 0%, treadmill* o 5% and

o 10%, And carrying the following 3loads: o No load, o 10 kg day pack, and a o 20 kg military rucksack. Participant lay prone with their head and shoulders off the ground resting on Drop and fire position their elbows and forearms Stooping/kneeling Participants kneeled on both knees, one knee, or crouched on their toes. Sitting Participants sat in a standard metal or plastic chair Standing Participants stood in one spot Walking from one side of the chamber to the Participants walked 4.5 metres, turned around and walked back continuously other Participants moved an ammo can from one 123.83cm high stand to another and Ammo can passing back continuously. This was done twice, once with a full ammo can (13.71kg) and once with an empty ammo can (3.16kg). Participants picked up two 20kg sandbags, carried them 7.44metres put them Moving sandbags down, walked back empty handed, picked up another 2 sandbags and so on. Participants picked up two Jerry cans, carried them 7.44metres turned around Moving Jerry cans and carried them back, turned around and so on. This was done twice, once with 2 full Jerry cans (20kg each) and once with two empty Jerry cans (2.5kg each). Participants lifted an ammo can from low stand (8.89cm) onto a high (174.63cm) Load and Unload Truck stand and brought it down again continuously. This was done twice, once with a Simulation full ammo can (13.71kg) and once with an empty ammo can (3.16kg). Activity Detailed Description of Activity Participants lifted an ammo can from the floor onto a low (123.83cm) stand and Ammo can lift brought it down again continuously. This was done twice, once with a full ammo

can (13.71kg) and once with an empty ammo can (3.16kg). Participants stacked and secured sandbags in an organized manner to build fortifications. - Participants picked up two 20kg sandbags, carried them 7.44metres put them down side by side, walked back empty handed, picked up another 2 sandbags and placed them beside the first two. Once 4 sandbags lay in Stacking and Tamping a row, participants flattened them down with a wooden handle (1kg). Participants would then walk 7.44metres, pick up another 2 sandbags, walk back and place the 2 sandbags on top of the first 2 sandbags. Every layer of 4 sandbags would be flattened before another row was added. Once a 4x4 sandbag fortification was built on one side of the room, another would be started. Flattening sandbags Participants hit and flattened sandbags with a 1kg wooden handle continuously with a wooden handle Flattening sandbags Participants stomped and flattened sandbags with their foot continuously with foot 68

Participants stepped up onto a 20.3cm step and then stepped back down continuously with: Step Up o No weight,

o 10kg day pack, and o 20kg rucksack ‘Escape to Cover’ army Participants ran as quickly as they could 4.5 metres, dropped to a prone position, drill got up and ran back as quickly as they could (4.5 metres) continuously. Leopard crawl Participants crawled across the ground staying as close to the ground as possible. Stretcher Carry Participants picked up two 22.5 kg dumbbells and walked from one side of the Simulation chamber (4.5 metres) to the other and back continuously Participants dragged an 85 kg mannequin 5.92metres turned around and dragged Dummy Drag them back continuously. *All combinations of speed, grade, and load were completed except for: 5.6km/h at 10 % grade with the 20 kg military rucksack. This combination was deemed to be uncharacteristic of the types to tasks military personnel would engage in.

EE was measured with the portable metabolic measurement system during each activity block (hours 0-2 and 4-6) in the environmental chamber. For each activity, participants rated their level of exertion on Borg’s Scale (Appendix O), and for each self-paced activity, researchers documented the number of repetitions the participant completed. Participants exited the chamber 8 h after entry, and they completed another POMS questionnaire. Participants took all unfinished food, their method of recording food intake (3DwFR or smart-phone app), and containers for 24-h urine collection home with them.

Figure 12: Schematic of each experimental condition. Grey boxes represent the 2-h ‘activity blocks’, although during the Sedentary condition the participants were inactive during these blocks. Participants arrived to the laboratory following a 10h overnight fast; the first visual analogue scale for appetite (VASA), as well as the first blood sample were collected upon arrival. Fasting and Post-Breakfast data points were collected outside of the environmental chamber prior to trial commencement. The 8-h trial began once the participant entered the environmental chamber; this occurred within minutes of the participants completing their breakfast.

4.4.5.2 Visit 5, 7, 9, 11: Follow-up Days

Participants reported to the laboratory with their food record and any unfinished food at 8:00am on the following day (the day after the 8-h chamber day). A fasting blood sample was 69

collected via venipuncture and participants completed a food satisfaction questionnaire (Appendix Q). On the last visit (visit 11) participants also completed an end of study questionnaire (Appendix P).

4.5 Measurements:

4.5.1 Basic Measurements

Body Composition was measured with air-displacement plethysmography (BOD PODTM, COSMED, Rome, Italy) during Visit 3. All participants wore tight-fitting spandex compression shorts and a lycra swim cap, and female participants also wore a spandex sport bra top. The tight-fitting clothing and cap is necessary in order to increase the accuracy of the test. The test requires the participant to sit inside a hollow chamber for approximately 2 min. Body composition was assessed two consecutive times to ensure accuracy. Thoracic gas volume was estimated from pre-determined values that are based on the participant’s height, weight, and age.

Body mass was measured at the beginning of each visit using a standard scale.

Height was measured at the beginning of the study using a stadiometer (Visit 2).

Blood pressure was measured at the beginning of each visit using an automated blood pressure monitor

Maximal aerobic power (V̇ O2max) was determined using a progressive incremental exercise test to voluntary exhaustion on a treadmill (Table 10) during Visit 2. Oxygen consumption was measured throughout the exercise by monitoring the inspired and expired respiratory gases using a mouthpiece assembly (nose clips, rubber mouthpiece, volume turbine and connecting hoses and lines) with a metabolic cart. Verbal encouragement was provided during the later stages of this test.

Table 10: Maximal aerobic power treadmill protocol adapted from [237]. Stage* Speed Grade 1 2.74 km/h (1.7mph) 10% 2 4.02 km/p (2.5mph) 12% 3 5.47 km/h (3.4mph) 14% 4 6.76 km/h (4.2mph) 16% 5 8.05 km/h (5.0mph) 18% 6 8.85 km/h (5.5mph) 20% 7 9.66 km/h (6.0mph) 22% *Each stage was maintained for a duration of 3 min. 70

Maximal heart rate: was determined during the maximal aerobic power test (Visit 2) (using transmitter/telemetry unit (Polar Electro PE3000, Finland).

Core Temperature was measured using a commercially-available personal activity monitor system (Equivital™, Hidalgo, Cambridge, United Kingdom). The system was used solely as a receiver for the core temperature pill. Core temperature was noted every 15 seconds from the moment the participant arrived to the laboratory, until they exited the environmental chamber (Visits 4, 6, 8, and 10).

Cortisol: Cortisol was assessed by urinary free cortisol (as measured by 24-hour urine collection).

Clothing weight was measured using a standard scale. Each item of clothing the participants wore in any conditions was weighed separately. On each chamber day (Visits 4, 6, 8, and 10), the clothing the participant wore in that condition was documented.

Mood was measured using the Profile of Mood States (POMS). The POMS is a 65 item computer questionnaire that provides measures on 6 mood states (tension, vigor, depression, anger, fatigue and confusion) and requires participants to score each question on a 5-point scale that ranges from 0 ("not at all") to 5 ("extremely") [238]. The POMS was completed twice on each chamber day (Visits 4, 6, 8, and 10), once before chamber entrance and once after chamber exit.

Sleep quality was determined using sleep quality questionnaires. Sleep was assessed in order to better understand the sleep patterns of the participants since studies have shown that food intake is affected by sleep deprivation. Both the Pittsburgh Sleep Quality Index (PSQI) which assesses sleep over the last month (Appendix I), and the Groningen Sleep Quality Scale (GSQS) which assesses how a participant slept the night before their visit (Appendix M) were employed.

Typical physical activity participation was determined through the physical activity questionnaire in order to understand how much physical activity the participants normally engage in.

71

Thermal comfort was assessed every hour during each chamber day (Visits 4, 6, 8, and 10), on the Thermal Comfort Scale (a 13-point scale ranging from 1 – ‘I am so cold I am helpless’ to 13 – ‘I am so hot I am sick and nauseated’) (Appendix N).

4.5.2 Energy Expenditure and Activity Measurements

Metabolic measurements: which include: oxygen consumption ( V&O2), carbon dioxide production ( CO2), EE (as calculated using Acheson’s RQ-based equation), metabolic -1 -1 equivalents (METs) (where 1 MET= 3.5 mL O2∙kg ∙min ), as well as the portion carbohydrate, fat, and protein oxidation, were all assessed simultaneously with the Metamax 3B portable oxygen uptake measurement system (CORTEX Biophysik GmbH, Leipzig, Germany). The Metamax 3B weighs about 1.5 kg and requires the participant to wear an oral-nasal mask over the face and an interface box strapped to the torso. The unit was worn during the activity portions (4 h) on each chamber day (Visits 4, 6, 8, and 10).

Heart rate: was measured continuously throughout each activity period (4 h) on each chamber day (Visits 4, 6, 8, and 10) using a heart rate strap (Polar Kempele, Finland), that was compatible with the Metamax 3B unit.

Movement: was assessed using the ActiGraph wGT3X-BT triaxial accelerometer (ActiGraph, Pensacola, Florida). The accelerometer was attached to a belt that the participant wore around their waist. The unit itself was placed between the right hip and the participant’s belly button. The accelerometer was initialized to collect data in 1 second epochs for the duration of each chamber day (Visits 4, 6, 8, and 10).

Subjective ratings of perceived exertion: were measured using Borg’s 6-20 scale [236]. This scale was presented to the participant at the end of each military task and they were asked to point to a number that reflects their ‘whole body’ perceived exertion level.

Activity repetitions: were counted during each self-paced repetitive military activity using a standard mechanical counter. Assessing activity repetitions was one way to determine whether participants completed a different amount of work in the different conditions.

Distance travelled per activity: was calculated from the following: the amount of time the participant spent doing the activity and the speed on the treadmill, or the repetitions and the distance covered per repetition. Distance travelled was another measure of ‘work done’.

72

Twenty-four-hour energy expenditure was estimated assuming that participants slept for 8 h and that they expended 0.95 METs while asleep [105]. Since participants were required to (and confirmed that they had) abstain from exercise for the 2 days prior to and on the chamber day, it was presumed that for the remaining 12 waking hours while EE was not measured, that participants expended energy at the same rate that they had during the Sedentary condition. 24 h EE was calculated as follows:

Where EE= energy expenditure, H=Hot condition, C=Cold condition, T=Temperate condition, S=Sedentary condition, BM=body mass in kg, and where 1 MET is equivalent to 1 kcal/kg/h [105].

4.5.3 Appetite Measurements

Appetite sensation: Visual Analogue Scales for Appetite (VASA) were used to assess appetite sensation while fasting, following breakfast, and every hour during each chamber day (Visits 4, 6, 8, and 10). The VASA assess four different sensations (hunger, fullness, satisfaction, and prospective eating), that together contribute to the assessment of appetite. Each sensation scale is a 100 mm line with the extremes of that sensation attached to each end of the line (e.g. ‘I am completely empty’ on the left end of the “satisfaction” scale and ‘I can’t eat another bite’ at the right end) (Appendix L).

Appetite hormone concentration: Peptide hormone responses to the various conditions (H, T, C, and R) were measured by the venous blood concentrations of the appetite-regulating peptide hormones: leptin, acylated ghrelin, glucagon-like peptide-1 (GLP-1), and polypeptide YY (PYY) at various time points on each chamber day (Visits 4, 6, 8, and 10). Refer to section 4.6 Blood Sampling for details.

Dietary intake: was assessed with a smartphone app and/or 3DwFR. While in the chamber, investigators also documented all feeding behaviours (selection, amount, and timing of food intake) and weighed all food that was discarded. The CAF Directorate of Food Services provided (from the manufacturer) the nutritional information for all of the presented rations some of which were based on chemical analysis. Dietary intake (EI and macronutrient intake) was determined by subtracting the energy and macronutrient content of the unconsumed food from the known quantities in the rations provided. 73

Relative energy intake (REI) was calculated by subtracting estimated 24 h EE from the energy consumed (kcal) throughout the day. From REI, it was determined whether a participant was in a positive or negative 24h energy balance.

Food satisfaction: The food satisfaction survey was used to assess the palatability of the rations that were provided. This information was used in conjunction with the food intake data in order to determine whether food that is reported as less palatable is eaten less often (Appendix Q).

4.6 Blood Sampling:

4.6.1 Blood sample timing

Blood samples were collected during Visits 3-11. During Visits 4, 6, 8, and 10 venous blood samples were drawn from the right or left antecubital vein using an indwelling catheter which was kept patent via infusion of saline (0.9% sodium chloride) following each sample taken. To determine how military activities in various environments differently affect appetite-regulating peptide hormones, blood samples during Visits 4, 6, 8, and 10 were taken immediately upon arrival to the laboratory (time point 0), at the end of the first exercise bout (beginning of the first rest period) (time point 2 h), 30 min into the first rest period (time point 2.5 h), 60 min into the first rest period (time point 3 h), 90 min into the first rest period (time point 3.5 h), 120 min into the first rest period (immediately before the second exercise bout) (time point 4 h), at the end of the second exercise bout (beginning of the second rest period) (time point 6 h), 60 min into the second rest period (time point 7 h), and 120 min into the second rest period (time point 8 h), while in the chamber. To determine how military activities in various environments differently affect nutritional status and health, one blood sample was collected immediately upon arrival to the laboratory on Sessions 3, 5, 7, 9, and 11 via venipuncture.

4.6.2 Collection procedures

Four (4) mL were taken at each blood sample time point listed above during Visits 4, 6, 8, and 10 for the measurement of leptin, acylated ghrelin, GLP-1, and PYY. Samples for the determination of GLP1, PYY, and leptin were collected in a chilled 2ml K2EDTA blood collection tube that was injected with a 167ul aprotinin solution containing 0.167mg of aprotinin (500KIU/ml blood) and 20ul of DPP-IV inhibitor (Millipore, Darmstadt, Germany). Samples for

74

the determination of acylated ghrelin were collected in a chilled 2ml K2EDTA blood collection tube that were injected with an 80uL 4-(2-Aminoethyl) benzenesulfonyl fluoride hydrochloride (AEBSF) solution containing 2mg of AEBSF. An additional three (3) mL was collected for the measurement of hemoglobin, and hematocrit in order to standardize for blood volume changes, and to ensure that the amount of blood that was taken from the participants, was not detrimental to their health. Therefore Visits 4, 6, 8, and 10 each required 9 blood samples, or 7 ml + 7 ml + 7 ml + 7 ml + 7 ml + 7 ml + 7 ml + 7 ml + 7 ml = 63 ml of blood (~4 tablespoons). Visits 3, 5, 7, 9, and 11 each required 1 blood sample, or 10 ml (<1 tablespoons) for the measurement of serum ascorbic acid, urea, sodium, potassium, creatinine, cholesterol and lipids and three (3) mL for the measurement of hemoglobin, and hematocrit). Thus, the total amount of blood sampled during the full duration of the study was: 13 ml + 73 ml + 13 ml + 73 ml + 13 ml + 73 ml + 13 ml + 73 ml + 13 ml = 357 ml, or about twenty-four tablespoons. Once drawn, blood samples were centrifuged immediately at 1000 x g for 15 min at 4ºC, and plasma was frozen at -70 ºC until subsequent biochemical analysis.

4.6.3 Blood Analysis

Immunoreactive plasma concentrations of leptin, GLP-1, PYY, and active ghrelin in the blood samples were analyzed with Meso Scale Discovery (MSD) 96-Well Ultra-Sensitive Human Immunoassay Kits, using electrochemiluminescent detection on an MSD Sector ImagerTM 6000 with Discovery Workbench software (version 3.0.18) (MSD®, Gaitherburg, MD, USA). Three different Meso Scale Diagnostics (MSD)assays were run, one for the determination of GLP-1 and leptin (Custom Human Metabolic Duplex -N45ZA-1, Rockville, Maryland), one for the determination of PYY (Human Total PYY - K151MPD-2, Rockville, Maryland), and one for the determination of active ghrelin (Custom Human Active Ghrelin - N45ZA-1, Rockville, Maryland). All assays were performed according to manufacturer’s instructions, in duplicates, and without alterations to the recommended standard curve dilutions. The sensitivity of the assays was 0.3pg/mL for GLP-1, 56pg/mL for leptin, 13pg/mL for PYY and 9pg/mL for acylated ghrelin. The inter- and intra-assay coefficients of variation were 12% and 9% for GLP- 1, 13% and <5% for leptin, 8% and 6% for PYY and 14% and 9% for acylated ghrelin respectively. All samples for one participant were run on the same plate, so as to limit inter- assay variation within participants.

75

All three kits were ‘sandwich immunoassays’. First biotinylated antibodies were added to the streptavidin coated MSD plates. Due to streptavidin’s high affinity for biotin, the biotinylated antibodies bound to streptavidin. Following this, the sample was added to the plate. The analyte of interest (leptin, GLP-1, PYY, or acylated ghrelin) in the sample, attached to the bound biotinylated antibodies. A solution containing antibodies labelled with electrochemiluminescent labels was then added to the plate. These labelled antibodies bound to the affixed analytes to form a sandwich complex. The MSD plates were then loaded into the MESO SECTOR S 600 plate reader. Voltage was applied to the plate’s electrodes and the electrochemiluminescent labels emitted light. The plate reader measured the intensity of the produced light and determined the concentration of the analyte in the sample (Figure 13).

Figure 13: The principal of electrochemiluminescent immunoassays.

4.7 Statistical Analysis

All data are expressed as means±SD unless otherwise stated. In order to compare the energy costs (METs) of our military tasks to the estimated costs (METs) taken from the Compendium of Physical Activities [105]or open source military studies, paired-samples t tests were used. Differences in EE, heart rate, core temperature, thermal comfort, clothing weight, perceived exertion, speed, and repetitions between conditions were examined using a one-factor repeated measures ANOVA between the Sedentary, Hot, Temperate and Cold trials; when necessary data were also examined by activity type (forced-pace, self-paced, and static).Repeated measures, two-factor ANOVAs were used to examine differences between trials over time for appetite sensation, EE, dietary intake, REI, and appetite-regulating peptide hormones. With each test, the assumption of sphericity was tested using Mauchly’s test. Where

76

Mauchly’s test was significant, the Greenhouse-Geisser estimate epsilon was assessed. If epsilon<0.75, than the degrees of freedom were corrected using the Greenhouse-Geisser correction, if epsilon>0.75, than the degrees of freedom were corrected using the Huynh-Feldt correction [239].Where significant main effects were found post hoc analysis was performed using the Bonferroni correction for multiple comparisons. Area under the curve calculations were made using the trapezoidal method. Pearson product moment correlation coefficients were also used to examine relationships between variables.

The algorithm was developed on a subset of data (Development group) and tested on the remaining data (Testing group). Algorithm development involved examining the strength of relationships found in the Development group data between EE (METs) and accelerometer output, HR, ambient temperature, and various personal variables (height, fat percentage, aerobic fitness etc.) using Pearson’s product-moment coefficient of correlation (r). Scatter plots were also utilized to further evaluate relationships between potential predictors and METs, where necessary transformations of the data were attempted to improve these relationships. The final EE algorithm was established using a fixed-effects model. The dependent variable for all attempted models was METs while, HR, accelerometry, ambient temperature, as well various personal variables (including height, sex, fat percentage, clothing weight, peak oxygen consumption, and age) were assessed as potential predictors of EE. Initially various mixed models were attempted with the inclusion of a random intercept for each participant. Mixed modelling when applied to repeated measures data considers how both within and between- individual factors affect EE. However, these mixed models all seemed to over-fit the data resulting in similar EE estimations (as compared to the final fixed effects model) when tested on the data from the Development group, and significantly worse estimations when tested on the data from the Testing group. This demonstrates that the variation between the participants in the Development group overly influenced the EE estimations. To limit this bias, fixed effects models are recommended [240] because unlike mixed models, fixed effects models are exclusively based on within-individual variation, and are not influenced by between-individual variation [240]. Overall this reduces the bias that can be introduced by personal variables (a necessity if this algorithm is to be applied to diverse military groups), but can potentially result in a less precise model [240]. The coefficient of determination (R2), and standard errors of the estimate (SEE) determined goodness of fit. Agreement between the measured METs (indirect calorimetry) and predicted METs (algorithm-determined) was assessed by Bland-Altman plots 77

which graphically depict the differences and 95% limits of agreement. Mean differences between predicted and measured METs were also determined by paired t-tests. Statistical significance was set at P<0.05. All statistical analysis was performed using the SPSS v. 22.0 software package.

78

Chapter Five: Study #1 - The Energy Cost of Various Infantry Tasks

Authors

I. Mandic,1 M. Ahmed,2 M. L’Abbe,2 L.S. Goodman,1,3 and I. Jacobs1

Affiliation

1 Faculty of Kinesiology and Physical Education, University of Toronto, Toronto ON M5S 2W6, Canada 2 Department of Nutritional Sciences, University of Toronto, Toronto ON M5S 3E2, Canada 3 Defence Research & Development Canada, Toronto Research Centre, Toronto ON M3K 2C9, Canada

Key terms: Compendium of Physical Activities; METs; energy expenditure; indirect calorimetry; clothing weight; military

Author contribution:

Mandic, I. contributed to research design, data acquisition, activity matching, analysis and interpretation of data, drafting of the manuscript and revision.

Ahmed, M. contributed to research design, data acquisition, activity matching, and manuscript revision.

Labe, M. contributed to research design and manuscript revision.

Goodman, L. contributed to data acquisition and manuscript revision.

Jacobs, I. was the principal investigator for the research contract that funded this research. He contributed to research design, analysis and interpretation of data, drafting of the manuscript and revision.

79

5.1 Abstract

Purpose: In this descriptive study, the metabolic cost of a standardized battery of infantry tasks was quantified using direct measurements of oxygen uptake in order to aid in future estimations of energy expenditure in military personnel in the field when such measurement techniques cannot be done.

Methods: The energy costs of 46 infantry tasks covering a range of light, moderate and heavy work rates were measured using indirect calorimetry in 21 Canadian Armed Forces (CAF) members (15 male, 6 female). Mean METs were calculated for each infantry activity and compared with similar activities reported in the Compendium of Physical Activities and in studies conducted in military personnel.

Results: When comparing the calculated MET values to published data, it was often difficult to effectively match the military tasks completed in this study to the activities described in the Compendium of Physical Activities. Many of the infantry tasks (20%) were not found in the Compendium and similar activities often significantly underestimated the energy cost of the military tasks. This prevailing underestimation is likely attributable to the clothing weight discrepancy found between military personnel (participants wore clothing weighing 13.5±1.4 kg) and the general population. In contrast, when comparing the infantry tasks in the current investigation with similar activities conducted in military personnel wearing similar clothing, no general underestimation was apparent.

Conclusion: This report has generated a compilation of the energy costs of common infantry tasks which can be used to estimate energy expenditure during field operations or training in the field.

80

5.2 Introduction

Knowledge of the energy demand associated with the activities of infantry personnel is important for several reasons. One reason may be to understand the aerobic physical fitness requirements to carry out missions and their component tasks so that individuals can be appropriately selected and/or trained. Another important reason for such knowledge is to be able to assess the nutritional energy requirements needed to sustain performance and health. When energy expenditure (EE) exceeds energy intake (EI) weight loss occurs, which if left unchecked can eventually lead to decrements in both health and physical performance [1, 2]. The nature of military operations and the physical activity components of missions can vary substantially based on the mission objectives and the environment in which the mission or training is being conducted. The diversity of duties can range from administrative jobs to construction to combat training, and consequently, it is important to recognize that the energy requirements in military personnel have also ranged from 2332 ± 373 kcal·day−1 for female administrative workers [3], to 6353± 478 kcal·day−1 for male Norwegian soldiers in the midst of field training [4] as measured via doubly labelled water (DLW).

As a result, commonly used estimations of energy requirements (like the Institute of Medicine’s Estimated Energy Requirement (EER) equations) are not particularly useful when it comes to estimating EE of military personnel in the field. Under such circumstances when there are no benchmarks for EER, and in the absence of measurement tools (DLW, indirect calorimetry, etc.) questionnaires and/or the factorial method could be considered. The latter is a method that involves the accurate recording of all daily activities and matches the activity and its duration with previously documented energy costs of the activity from relevant and reliable sources such as scientific journals or technical reports. The energy costs of many military tasks are undefined, with the Compendium of Physical Activities [5] having sparse military relevant data, and additional sources being limited in their usefulness due to vaguely described activities [6, 7].

In the current study the energy cost of several common infantry tasks was measured in order to provide an empirical basis for estimating dietary energy intake needs for Canadian Armed Forces (CAF) infantry personnel when they are deployed for field operations or training. Typical military tasks were compared to similar activities described in the Compendium of Physical Activities and studies conducted in military personnel. The Compendium is a valuable

81

instrument for researchers and the continued expansion and inclusion of different empirically measured tasks will only further increase the usability of this tool.

82

5.3 Methods

5.3.1 Participants

Twenty-seven healthy, male (n=21) and female (n=6) Canadian Armed Forces (CAF) members volunteered to participate in the study (Table 11). Six withdrew voluntarily after starting the study due to time constraints (n=3), reconsideration (n=3), or loss of contact (n=1). As a result, 21 participants (15 male, and 6 female) completed the study. All participants were free from metabolic and cardiac disorders and were not taking any medications or natural health products. All participants were fully informed of the details, discomforts and risks associated with the experimental protocol before being asked for their written informed consent. The study protocol was reviewed and approved by the institutional human research ethics committees at Defence Research and Development Canada (#2013-075), and the University of Toronto (#29914).

Table 11: Participant characteristics. Participant Height Body Body V̇ O Sex Age (y) BMI 2Peak Status (cm) Mass (kg) Fat (%) (mL·min-1·kg-1) 15 males 33.5 ± 171.9 ± 78.7 ± 26.5 Completed 24 ± 6 43 ± 6 6 females 10.1 10.3 15.4 ± 4.0 26.5 ± 176.2 ± 78.6 ± 25.8 Withdrawn 6 males 17± 6 47 ± 4 5.8 6.7 6.5 ± 3.4 21 males 32.0 ± 172.8 ± 78.7 ± 26.3 All 22 ± 8 44 ± 6 6 females 9.7 10.0 13.8 ± 3.6 There were no significant differences between individuals who completed the study (completed), and those who did not (withdrawn). Data are presented as mean ±SD.

5.3.2 Experimental Design

Participants made three initial visits to the lab, followed by one long physically arduous day in the lab.

5.3.2.1 Initial Visits

Visit 1 consisted of the completion of the informed consent form, PAR-Q+ [8], and physical activity questionnaires. During Visit 2, participants had their peak aerobic power

(V̇ O2peak) measured using indirect calorimetry during an incremental treadmill exercise test to exhaustion [9]. During Visit 3, participants reported to the laboratory after a 10-h overnight fast. Body mass was measured with a standard scale, and percent body fat was estimated via air- displacement plethysmography (BOD PODTM, COSMED, Rome, Italy).

83

5.3.2.2 Long Day

At 7:45am, participants reported to the laboratory and had their body weight measured using a standard scale, and their resting blood pressure measured using an automated blood pressure cuff. Participants then self-selected typical pre-packaged military field rations to sustain them for the day (1 breakfast, 1 lunch and 1 dinner ration). These rations are referred to by the CAF as Individual Meal Packages (IMPs). The participants were able to consume these rations ad libitum from that point up until 10pm that night. All food and drinks that were consumed in the lab were documented by the investigators. Participants consumed breakfast, and were then outfitted in typical military clothing, (which included: military fatigues, a tactical vest (~3-7kg), a fragmentation protection vest (2-3kg), combat boots (1.5kg- 2.5kg), and a helmet (~1.5kg)). Participants were also outfitted with a portable metabolic measurement system (Metamax 3B, CORTEX Biophysik GmbH, Leipzig, Germany) which weighed about 1.5 kg and required the participant to wear an oral-nasal mask over the face and an interface box strapped to the torso and a heart rate telemetry chest strap (Polar, Kempele, Finland), that was compatible with the Metamax 3B unit. Once outfitted in the appropriate clothing and equipment, the participants entered the environmental chamber that was maintained at 21°C, and 30% relative humidity and completed two 2-h circuits of a standardized set of typical military tasks, with two hours of rest following each set. The activities are described in Table 12. Activities were completed at scheduled specified times, and participants were asked to complete all non-treadmill tasks ‘as they would in the field’. While completing the various activities, activity repetitions were counted during each repetitive military task using a standard mechanical counter. Immediately following each activity participants were asked to rate their perceived exertion on Borg’s 6-20 scale [10].

84

Table 12: Battery of standardized infantry tasks that were performed by all participants in the environmental chamber and the approximate times that these activities were done. Block Time Activities completed - Walking on a motorized treadmill at 2.4 km/h, 4 km/h, and 5.6 km/h, at 0%, 5% and 10% grade with and without a 10 kg day 1 pack. (ACTIVE) - Moving full and empty ammo cans from one stand to another. 8:00-10:00 - Moving sandbags and full jerry cans from one corner to another - Stooping/kneeling - Simulating the ‘Escape to Cover’ army drill - Sitting in one spot - Leopard crawling 2 10:00- - Resting and Eating (REST) 12:00 - Walking on a motorized treadmill at 2.4 km/h, 4 km/h, and 5.6 km/h, at 0%, 5% and 10% grade with a loaded 20 kg military rucksack. - Moving empty jerry cans from one corner to another - Walking from one side of the chamber to the other with and without two 22.5 kg dumbbells - Lying in a drop and fire position 3 12:00- - Stepping up onto a 20.3cm step with no weight, with a 10kg day (ACTIVE) 14:00 pack, and with a 20kg rucksack - Stacking and Tamping - Flattening sandbags with a wooden handle or their foot continuously. - Standing in one spot - Lifting empty and full ammo cans from the ground - Dragging an 85 kg mannequin from one side of the chamber to the other. 4 14:00- - Resting and Eating (REST) 16:00

5.3.3 Indirect Calorimetry

Breath-by-breath data were collected and averaged over 30 second intervals with the Metamax 3B portable oxygen uptake measurement system. The unit was worn throughout the entire activity portions (4-h) on the long day and participants only took the mask off to drink water. Oxygen consumption (VO2) measurements were collected and displayed in real time throughout the activity portions. Although 1 MET is defined as the amount of oxygen consumed at rest (the resting metabolic rate (RMR)), [11], and can be measured accordingly, a reference value of 3.5 ml·kg·min−1 is often used when measurements cannot be made. The Compendium of Physical Activities calculated ‘Standard METs’ by using the reference value of 3.5 ml·kg·min−1 instead of measured RMR [5]. Even when measured RMR was available for some −1 −1 activities, VO2 values in ml·kg·min were divided by 3.5 ml·kg·min in order to maintain

85

−1 consistency [5]. METs in this study were calculated in the same manner (mean VO2 ml·kg·min for each military task was divided by 3.5 ml·kg·min−1 and termed ‘Calculated METs’) to facilitate comparisons among studies.

5.3.4 Data Reduction

Participants completed 50 different infantry tasks throughout the 4-h of activity. Some of these activities were short in duration (<2min) due to their high intensity (Mannequin Drag, Leopard Crawl etc.). Although these activities were completed in the manner and for the same duration that would be expected during operations, they were too short in duration for VO2 to reach a steady state. As a result, all activities lasting 210 seconds or less were visually inspected and if a plateau was reached, the last 30 seconds of data were used; if no plateau was attained, the activity was excluded. For all activities lasting longer than 211 seconds, the first 180 seconds of oxygen uptake data were excluded, and the remaining data were analyzed.

Two investigators (IM & MA) independently matched the completed activities with activities listed in the Compendium of Physical Activities. Investigators selected the same Compendium activity the majority of the time, and where discrepancies occurred, the activities were reviewed and discussed amongst the two investigators. During the review process, a Compendium activity was only matched if both investigators agreed that it was an acceptable match, otherwise no match was made. Reasonable matches were made for 80% of the activities. Match quality was ranked based on agreement between the two investigators:

i) If both investigators independently and confidently matched the same activity from the Compendium with an infantry task, then the match was deemed to be ‘excellent’ (E).

ii) If both investigators selected the same activity from the Compendium to match with an infantry activity but had some reservations (e.g. the Compendium activity is a match but the description seems too general), then the match was deemed to be ‘good’ (G).

iii) If there was initially a discrepancy between the Compendium activities selected by the investigators, but following further discussion an activity was agreed upon, or if both investigators found the Compendium activity description to be vague, then the match was deemed to be ‘adequate’ (A).

86

Open-source studies reporting the energy cost of activities completed by military personnel were assessed. If the activities completed in these studies were clearly described, and the reported oxygen uptake data were adequate then the activities in these studies were also matched to those completed in the current investigation.

5.3.5 Statistical Analysis

All data are expressed as mean ± standard deviation (SD) unless otherwise stated. Paired- samples t tests were used for each activity to compare the measured energy cost (METs) to the estimated cost (METs) taken from the Compendium of Physical Activities or the open source military studies. All statistical analysis was performed using the SPSS v. 22.0 software package and statistical significance was accepted at p<0.05.

87

5.4 Results

Out of the 50 different military tasks, four (Escape to Cover, Leopard Crawl, Stretcher Carry, and Mannequin Drag) were too short in duration for EE to reach a steady state, and these were excluded from analysis. Out of the remaining 966 individual tasks (activity*participants), 90 were excluded due to: 1) failure to complete the task (the task was too strenuous for the participant, one participant was too tall for the 10% grade on the treadmill, another was too short for the load and unload truck simulation, etc.) or 2) equipment problems. As a result, oxygen uptake measurements were available for the 876 individual activities displayed in Tables 13, 14, and 15.

5.4.1 Treadmill Activities

Treadmill activities were compared to similar activities found in the Compendium of Physical Activities (Table 13). Many of the treadmill activities completed in the current study (7) did not adequately match activities found in the Compendium of Physical Activities. In particular, many of the walking tasks that were completed at an incline were difficult to match. Nineteen treadmill activities were matched to activities in the Compendium. Of these 19 activities only two activities were accurately estimated by the Compendium; i.e. the calculated METs were not significantly different from those estimated by the Compendium (p>0.05). An additional four activities were found to have calculated METs that were within 10% of the Compendium METs. Considering that the SD of the energy cost of the tasks completed in the current study ranged between 6% and 30%, variations between calculated and Compendium METs that vary by less than 10% can also be deemed to be reasonably estimated by the Compendium. The Compendium of Physical Activities data underestimated the measured energy cost of nine treadmill tasks and overestimated the energy cost for four treadmill tasks. Five of the 13 tasks that were not accurately estimated by the Compendium (where calculated METs were not within 10% of the Compendium METs and were also significantly different from those estimated by the Compendium p<0.05) were still estimated within 1 MET of the measured value, and only one task was found to be more than 2 METs off.

88

Table 13: ‘Calculated METs’ from the treadmill activities compared to ‘Standard METs’ from similar activities found in the Compendium of Physical Activities. Treadmill Activities# VO 2 Calculated Compendium Compendium Match Load N ml·kg·min−1 Compendium description Speed Grade METs‡ code(s) METs QualityϞ carried (SD) walking, less than 2.0 mph, level, strolling, very 17151, 20036, 0% grade. 19 9.90 (1.48) 2.83 2.0* slow and or walking on job, less than 2.0 mph, very E 21045, 11791 2.4 km/h slow speed, in office or lab area (1.5mph) 5% grade. 19 12.51 (1.43) 3.57 hiking or walking at a normal pace through fields 10% grade. 20 15.69 (1.81) 4.48 17082 5.3* A and hillsides No load 0% grade. 19 13.51 (1.43) 3.86 17170 3.0* walking, 2.5 mph, level, firm surface E 4 km/h 5% grade 19 17.74 (1.39) 5.07 17088 4.5* marching, moderate speed, military, no pack A (2.5mph) 10% grade 19 23.77 (1.52) 6.79 17033 6.3* climbing hills, no load E 17200, 20038, walking, 3.5 mph, level, brisk, firm surface, 0% grade 18 20.29 (1.74) 5.80 4.3* E 5.6 km/h 21050 walking for exercise (3.5mph) 5% grade 19 26.84 (1.88) 7.67 17210 5.3* walking, 2.9 to 3.5 mph, uphill, 1 to 5% grade E 10% grade 16 34.83 (2.70) 9.95 17211 8.0* walking, 2.9 to 3.5 mph, uphill, 6% to 15% grade E walking and carrying small child, child weighing 15 0% grade. 19 11.24 (1.37) 3.21 05181 3.0* G 2.4 km/h lbs or more (1.5mph) 5% grade. 20 13.78 (1.43) 3.94 10% grade. 20 17.35 (1.74) 4.96 walking on job, 2.5 mph, slow speed and carrying Wearing a 0% grade. 20 14.77 (1.54) 4.22 11795, 21055 3.5* E 4 km/h light objects less than 25 pounds 10kg (2.5mph) 5% grade 20 19.73 (1.61) 5.64 17101 7.0* backpacking (Taylor Code 050) G (22lbs) 10% grade 18 26.48 (1.94) 7.57 17050 8.3* climbing hills with 21 to 42 lb load G day pack walking, 3.5 mph, briskly and carrying objects less 0% grade 19 22.98 (2.75) 6.57 11810, 21065 4.8* E than 25 pounds 5.6 km/h backpacking, hiking or organized walking with a (3.5mph) 5% grade 17 29.98 (3.43) 8.56 17012 7.8* A daypack 10% grade 12 39.55 (3.28) 11.30 Walking or walk downstairs or standing, carrying 0% grade. 15 12.92 (2.72) 3.69 11820 5.0* A 2.4 km/h objects about 25 to 49 pounds (1.5mph) 5% grade. 20 16.05 (2.28) 4.59 Wearing a 10% grade. 20 19.95 (2.30) 5.70 17060 9.0* climbing hills with 42+ lb load A Walking, 2.5 mph, slow speed carrying heavy 20kg 0% grade. 18 16.90 (2.82) 4.83 11797 3.8* E (44lbs) 4 km/h objects more than 25 lbs rucksack (2.5mph) 5% grade 19 22.97 (3.29) 6.56 17010 7.0 backpacking (Taylor Code 050) A 10% grade 16 30.37 (2.78) 8.68 17060 9.0 climbing hills with 42+ lb load G 5.6 km/h 0% grade 17 26.39 (4.26) 7.54 (3.5mph) 5% grade 16 33.87 (4.73) 9.68 # Participants wore a total of ~13.5±1.4kg of clothing and measurement equipment while performing all activities. ‡METs were calculated by taking the measured VO2 ml·kg·min−1 and dividing by 3.5 ml·kg·min−1. *Calculated METs were significantly different from Compendium METs p<0.05 Ϟ Match Quality qualitatively describes how well the Compendium activity matched the treadmill activity by categorizing it into one of 3 groups Excellent (E), Good (G), or Adequate (A). Compendium METs in red are estimated MET values.

89

The metabolic demands of activities performed on the treadmill in the current study were also compared to similar activities found in published studies conducted in military personnel wearing military clothing (Table 14). When comparing the current treadmill results to data from other reports describing military personnel wearing and carrying military gear, there was a much closer agreement in the metabolic response when the weight worn (clothing + equipment) and the speed of walking were similar between the compared activities. At lower intensities (when walking speed was <4.5 km·h−1 and load carried was <25 kg), METs reported in military studies were within 10% of the METs reported in the current investigation if the weight worn (clothing + equipment) was within 5 kg, and the speed was within 0.5 km·h−1. At higher intensities (when the load carried exceeded 25 kg, or the walking speed was >4.5 km·h−1), lower variations in load and/or speed at times resulted in higher variations in METs (Table 14). Nevertheless, METs calculated in the current study were within 17% of reported METs in all situations regardless of intensity, when the weight worn (clothing + equipment) was within 5kg, the grade was within 3%, and the speed was within 0.5km·h−1.

90

Table 14: ‘Calculated METs’ from the treadmill activities compared to ‘Calculated METs’ from similar activities found in military studies. Our Results Military Studies Total Total Load VO2 Load VO2 Treadmill Speed Calculated Military Study Speed Calculated (worn and N ml·kg·min−1 (worn N ml·kg·min−1 Reference Activity (km/h) METs‡ Activity (km/h) METs‡ carried) (SD) and (SD) carried) Women: Walk at 5% Treadmill walking at 3.7 km/h 4 19 17.74 (1.39) 5.07 17 18.8 (1.7) 5.37* grade 5% grade Men: 3.86 km/h 13.5±1.4kg 10kg [12] Women: Walk at 10% Treadmill walking at 5.79 km/h 5.6 16 34.83 (2.70) 9.95 17 40.8 (5) 11.66* grade 10% grade Men: 6.12 km/h Walk at 0% 3.5 15 13.39 (1.19) 3.83* 4 19 14.77 (1.54) 4.22 grade with 23.5±1.4kg 20kg 4.5 15 15.43 (2.42) 4.41 daypack 5.6 19 22.98 (2.75) 6.57 5.5 15 19.78 (2.70) 5.65* Treadmill walking at Walk at 0% 3.5 15 15.69 (2.46) 4.48 [13] 4 18 16.90 (2.82) 4.83 0% grade grade with 4.5 15 19.71 (3.04) 5.63* 33.5±1.4kg 35kg military 5.6 17 26.39 (4.26) 7.54 5.5 15 23.64 (3.37) 6.75* rucksack Walk at 0% Walk in the field ~2% grade with 23.5±1.4kg 5.6 19 22.98 (2.75) 6.57 24.4kg ~5.7 8 22.7 (3.4) 6.49 [14] grade daypack Move by foot on a 9 men MEN: mean (SE) level, hard surface at 4 19 13.51 (1.43) 3.86 9.3kg 4 and 11 Men:13.9(0.2) §4.05 1.11 m/s wearing women Women:14.4(0.3) Walk at 0% combat equipment. 13.5±1.4kg grade Move by foot on a 9 men MEN: mean (SE) level, hard surface at 5.6 18 20.29 (1.74) 5.80 9.3kg 5.33 and 13 Men:18.2(0.5) §5.34* 1.48 m/s wearing women Women: 19.0(0.6) combat equipment. Move by foot on a [15] level, hard surface at 9men MEN: mean (SE) 4 18 16.90 (2.82) 4.83 1.11 m/s wearing 4 and 12 Men: 14.8(0.5) §4.69 Walk at 0% combat equipment with women Women:17.6(0.7) grade with a 20 kg rucksack. 33.5±1.4kg 29.3kg military Move by foot on a rucksack level, hard surface at 9 men MEN: mean (SE) 5.6 17 26.39 (4.26) 7.54 1.48 m/s wearing 5.33 and 10 Men: 20.7(0.3) §6.52* combat equipment with women Women: 24.7(1.2) a 20 kg rucksack. ‡METs were calculated by taking the measured VO2 ml·kg·min−1 and dividing by 3.5 ml·kg·min−1. *Our calculated METs were significantly different from METs calculated from data presented in military studies p<0.05 §A weighted average of Men’s and Women’s METs was calculated and presented. 91

5.4.2 Non-Treadmill Activities

The results from the activities in the current investigation that were not carried out on a treadmill were compared to similar activities found in the Compendium of Physical Activities

(Table 15). Of the 20 non-treadmill activities, all but two were matched with a Compendium activity. Half of those activities (9) were accurately estimated by the Compendium (p>0.05), with an additional 5 falling within 1 MET of the measured METs. Only one activity differed by more than 2 METs from the estimated value. Eight activities were underestimated by the

Compendium, and one was overestimated.

92

Table 15:‘Calculated METs’ from non-treadmill activities compared to ‘Standard METs’ from similar activities found in the Compendium of Physical Activities.

Non Treadmill Activities# VO 2 Calculated Compendium Compendium Match (Weight carried) - mean±SD activity rate in N ml·kg·min−1 Compendium description METs‡ code(s) METs Quality RPM֎ or km/h (SD) Drop and fire position 20 6.03 (1.18) 1.72 04140 2.3* rifle exercises, shooting, lying down E Stoop/Kneel 20 6.39 (1.63) 1.83 20025 1.3* kneeling in church or at home, praying E Sit in one spot 20 5.37 (0.83) 1.53 07022 1.5 sitting quietly, fidgeting, general, fidgeting hands E Stand in one spot 20 5.66 (1.43) 1.62 07041 1.8 standing, fidgeting E Walk from one side of chamber to the other – walking on job, 3.0 mph, in office, moderate speed, 21 14.23 (3.71) 4.07 11792 3.5* E 3.8±0.8 km/h not carrying anything standing, light/moderate effort (e.g., assemble/repair Empty ammo can (3.2kg) passing – 41±15 RPM 20 9.86 (1.74) 2.82 11610 3.0 heavy parts, welding, stocking parts, auto repair, G standing, packing boxes, nursing patient care) standing, moderate effort, lifting items continuously, Full ammo can (13.7 kg) passing – 34±11 RPM 19 14.84 (2.37) 4.24 11615 4.5 G 10 – 20 lbs, with limited walking or resting Move sandbags (20 +20 kg) from one corner to 20 28.21 (4.27) 8.06 11050 8.0 carrying heavy loads (e.g., bricks, tools) G another – 3.8±0.7 km/h Empty Jerry can (2.5+2.5 kg) carry – 4.3±1.0 km/h 21 21.63 (4.90) 6.18 05121 5.0* moving, lifting light loads A Full Jerry can (20+20kg) carry – 3.5±1.0 km/h 20 27.87 (3.95) 7.96 11050 8.0 carrying heavy loads (e.g., bricks, tools) G Empty ammo can (3.2kg) load and unload truck 19 19.65 (3.66) 5.61 simulation - 27±11 RPM Full ammo (13.7kg) can load and unload truck truck driving, loading and unloading truck, tying down 15 27.85 (5.94) 7.96 11766 6.5* A simulation - 20±3 RPM load, standing, walking and carrying heavy loads Empty ammo can (3.2 kg) lift - 35±12 RPM 20 19.65 (4.15) 5.61 05121 5.0* moving, lifting light loads E standing, moderate effort, lifting items continuously, Full ammo can (13.7kg) lift - 31±8 RPM 21 25.85 (5.02) 7.39 11615 4.5* E 10 – 20 lbs, with limited walking or resting Stacking and Tamping – 0.9±0.3 RPM₡ 21 25.33 (3.94) 7.24 11477 6.5* manual or unskilled labor, general, vigorous effort A Flattening sandbags with a wooden handle (1 kg) forestry, ax chopping, slow, 1.25 kg axe, 19 21 16.38 (5.16) 4.68 11260 5.0 G continuously - 85±35 RPM blows/min, moderate effort Flattening sandbags with foot - 101±27 RPM 21 16.29 (4.41) 4.65 Step up no load - 52±10 RPM 21 24.33 (4.23) 6.95 03016 7.5 aerobic, step, with 6 - 8 inch step E Step up with daypack (10 kg) - 53±11 RPM 21 25.15 (4.17) 7.19 17027 6.0* carrying 16 to 24 lb load, upstairs E Step up with rucksack (20 kg) - 57±11 RPM 21 27.19 (4.62) 7.77 17028 8.0 carrying 25 to 49 lb load, upstairs E # Participants wore a total of ~13.5±1.4kg of clothing and measurement equipment while performing all activities. ‡METs were calculated by taking the measured VO2 ml·kg·min−1 and dividing by 3.5 ml·kg·min−1. *Calculated METs were significantly different from Compendium METs p<0.05. Ϟ Match Quality qualitatively describes how well the Compendium activity matched the non-treadmill activity by categorizing it into one of 3 groups Excellent (E), Good (G), or Adequate (A). ֎RPM – Repetitions per minute. Moving an object from one position to another = 1 repetition, moving it back to its initial position = 2 repetitions. Each hit/stomp = 1 repetition. Stepping up =1 repetition, stepping back down =2 repetitions. ₡Every 4 sandbags that are laid and flattened= 1 repetition Compendium METs in red are estimated MET values.

93

5.5 Discussion

A primary objective of the current investigation was to measure and describe the metabolic demand associated with the acute performance of various common infantry tasks. The results of this research provide an empirical basis that extends knowledge about the EE likely to be demonstrated by military personnel during infantry combat and training operations in the field. The measured data were compared to published values currently available for similar activities. The need for such research was reflected in our observations that many of the common tasks assessed in the current investigation could not be matched with activities identified in the Compendium of Physical Activities, and where matching could occur, the Compendium values were frequently underestimations of the energy cost as measured in the current investigation.

Nine of the 46 tasks in the current investigation could not be matched to a Compendium activity at all. From the remaining 37 infantry tasks, the energy costs of only 11 tasks were considered to be accurately estimated by the Compendium (p>0.05), with an additional four tasks having calculated METs within 10% of Compendium METs. The discrepancies between the Compendium METs and our calculated METs could be influenced by the subjective and qualitative nature of activity matching used in the current report; but this process is likely comparable to the procedure most researchers would use when estimating METs from self- reported questionnaires or determining METs for observed activities. Although two separate investigators matched the activities completed in the current investigation with those found in the Compendium, discrepancies in matching occurred 20% of the time. Even when the investigators selected the same Compendium activity as matching one of the tasks performed in the current investigation, an additional 17% of the time at least one investigator found the selected Compendium activity to be too vague or had reservations about the interpretation of the activity description. For example, Compendium activity 17060 – ‘climbing hills with 42+lb load’, does not specify a speed, or level of effort; therefore it was selected as the best match in the current study for both the 2.4 km·h−1 and 4 km·h−1 speeds when the participant was walking on a treadmill carrying a military rucksack at 10% grade. In one case (the 4 km·h−1 speed), the Compendium METs are the same as the calculated METs, whereas in the other instance (the 2.4 km·h−1 speed) the Compendium METs overestimate the energy cost by more than 3 METs.

94

Unfortunately, even when both investigators agreed and felt confident about the match selection from the Compendium listing of activities, this did not improve the MET estimation; only four of our 19 ‘excellent’ matches (21%) accurately estimated the energy cost of our infantry activities, with one additional “excellent” match having a calculated MET value within 10% of the Compendium MET value. Almost all other ‘excellent’ matches (13/14) underestimated the energy cost for the matching infantry activities. This consistent underestimation of our ‘excellent’ matches could be due to the heavy military clothing that our participants and more generally, military personnel wear when engaged in field training or deployed operations.

The energy cost of an activity (especially an activity that does not involve the support of body weight) increases as clothing weight increases [16], and the effect becomes more pronounced at higher exercise intensities [12]. Ricciardi et al. [12] reported that when walking at a slow pace (3.7 km·h−1 for women and 3.86 km·h−1 for men at 5% grade) oxygen consumption increased from 16.8±1.5 ml·kg·min−1 when wearing light clothing (t-shirt, short, and running shoes) to 18.8±1.7 ml·kg·min−1 when wearing body armour weighing 10kg. At higher intensities (5.79 km km·h−1 for women and 6.12mph for men at 10% grade) the increase in oxygen consumption as a result of the body armour was more dramatic, from 34.8±3.9 ml·kg·min−1 to 40.8±5 ml·kg·min−1 [12]

Although heavier clothing can significantly increase the energy cost of an activity, very few tasks listed in the Compendium of Physical Activities give any indication of the clothing that was worn while the energy cost of the task was assessed. Considering that the Compendium was created to ease the comparability of outcomes between studies [5], it mostly caters to the general population; as such it is logical to assume that the clothing worn during the energy cost assessment of the various activities would be typical of the general population and appropriate for the activity at hand unless clearly defined. Since the general population does not typically wear over 10kg of clothing and equipment, the overall underestimation of energy costs reported in the current investigation of infantry activities is expected.

As can be anticipated, comparisons of the results in the current investigation to other studies conducted with military personnel resulted in more similar values for the assessed activities. In particular our data corresponded very well with Pihlainen et al. [14] who found that marching at ~5.7 km·h−1 over various terrain (~2% grade) while wearing clothing and 95

equipment weighing 24.4kg resulted in oxygen consumptions of 22.7±3.4 ml·kg·min−1 whereas our participants consumed 22.95±2.75 ml·kg·min−1 of oxygen when walking with a daypack at 5.6 km·h−1 at 0% grade (with clothing/equipment/day pack weight = 23.5kg). Unfortunately, many published studies that assess the energy cost of various activities in military personnel do not display their results in values relative to participant weight, making the results more difficult to compare and apply. For the few studies that did include body weight, the activities described although not identical, can be considered close approximations of the activities performed in the current study. As such, the MET values for similar tasks reported in previous military studies (where the weight worn (clothing + equipment) was within 5kg, the grade was within 3%, and the speed was within 0.5 km·h−1) were all within 17% of the calculated METs in this investigation.

It is important to acknowledge that although the current quantification of the energy costs of infantry tasks will augment the utility of resources like the Compendium in estimating the expected EE in military personnel during infantry operations and training, there are inherent errors in estimating individual EE by using group mean values. For example, some EE estimation errors using group mean METs will be attributable to the normal distribution of inter- individual physiological and physical size differences, and some will be due to the range of ways that individuals choose to perform tasks. Moreover, our compilation of infantry tasks, like the Compendium is based on a reference value 3.5 ml·kg·min−1 being equivalent to RMR or 1 MET. This reference value, however, has previously been found to overestimate RMR in heavier, older, and less fit individuals [17]. As a result the data in the Compendium have been reported to underestimate the energy cost in these populations; and consequently correction factors have been developed to ‘correct’ for these issues [17]. The decision to use calculated vs. ratio (Activity EE/RMR) METs in this study was made in order to adequately compare our results with the Compendium. In addition, we acknowledge the likelihood that the widely used oxygen uptake of 3.5 ml·kg·min−1 will likely continue to be used as a 1 MET equivalent when EE estimation is conducted in military personnel, due to the cost and inconvenience of performing RMR measurements on a representative population sample

In conclusion, this study reports descriptive data on the metabolic cost of various common infantry tasks performed by the CAF. The Compendium of Physical Activities, although a valuable resource, was found to significantly underestimate the energy cost of most

96

of the infantry activities that were assessed in the current study using measurements of oxygen uptake. This general underestimation is in large part attributed to the heavier clothing and clothing accessories generally worn by military personnel in the field during deployed operations and training. This expanded empirically-based knowledge about the EE of common military tasks should be useful in improving the accuracy of estimates of energy expenditure during infantry operations, and therefore nutritional energy intake needs.

5.6 Acknowledgements

This research was funded by Defence Research & Development Canada (DRDC). This study was approved by the Canadian Forces Surgeon General’s Health Research Program. In accordance with the Department of National Defence (DND) policy, the paper was reviewed and approved for submission without modification by the DRDC Publications Office.

The authors acknowledge the valuable contributions of DRDC staff in data collection and participant recruitment, the Canadian Forces Environmental Medicine Establishment (CFEME) for medical support, and the CAF members who volunteered to participate in this study.

5.7 Conflict of Interest

The authors declare that they have no conflicts interests. The results of the study are presented clearly, honestly, and without fabrication, falsification, or inappropriate data manipulation. The results of the present study do not constitute endorsement by American College of Sports Medicine.

97

5.8 References 1. Moore, R.J., et al., Changes in soldier nutritional status and immune function during the Ranger training course. 1992, US Army Research Institute of Environmental Medicine: Natick, MA. 2. Shippee, R., et al., Nutritional and immunological assessment of Ranger students with increased caloric intake. 1994, US Army Research Institute of Environmental Medicine: Natick, MA. 3. Delaney, J.P., Determination of Total Daily Energy Requirements and Activity Patterns of Service Women. 2001, DTIC Document. 4. Hoyt, R.W., et al., Negative energy balance in male and female rangers: effects of 7 d of sustained exercise and food deprivation. Am J Clin Nutr, 2006. 83(5): p. 1068-75. 5. Ainsworth, B.E., et al., 2011 Compendium of Physical Activities: a second update of codes and MET values. Med Sci Sports Exerc, 2011. 43(8): p. 1575-81. 6. Goldman, R., Energy expenditure of soldiers performing combat type activities. Ergonomics, 1965. 8(3): p. 321-327. 7. Consolazio, C.F., Energy expenditure studies in military populations using Kofranyi- Michaelis respirometers. The American journal of clinical nutrition, 1971. 24(12): p. 1431-1437. 8. Warburton, D., et al., The 2014 Physical Activity Readiness Questionnaire for Everyone (PAR-Q+) and electronic Physical Activity Readiness Medical Examination (ePARmed-X+). Health & Fitness Journal of Canada, 2014. 7(1): p. 80. 9. Bruce, R.A., F. Kusumi, and D. Hosmer, Maximal oxygen intake and nomographic assessment of functional aerobic impairment in cardiovascular disease. Am Heart J, 1973. 85(4): p. 546-62. 10. Borg, G., Perceived exertion as an indicator of somatic stress. Scandinavian journal of rehabilitation medicine, 1970. 2(2): p. 92-98. 11. Jette, M., K. Sidney, and G. Blümchen, Metabolic equivalents (METS) in exercise testing, exercise prescription, and evaluation of functional capacity. Clinical cardiology, 1990. 13(8): p. 555-565. 12. Ricciardi, R., P.A. Deuster, and L.A. Talbot, Metabolic demands of body armor on physical performance in simulated conditions. Military medicine, 2008. 173(9): p. 817-824. 13. Christie, C.J. and P.A. Scott, Metabolic responses of South African soldiers during simulated marching with 16 combinations of speed and backpack load. Military medicine, 2005. 170(7): p. 619-622. 14. Pihlainen, K., et al., Cardiorespiratory responses induced by various military field tasks. Military medicine, 2014. 179(2): p. 218-224. 15. Patton, J.F., et al., Metabolic cost of military physical tasks in MOPP 0 and MOPP 4. 1995, US Army Research Institute of Environmental Medicine: Natick, MA. 16. Romet, T., et al., The metabolic cost of exercise in a cold air environment (-20 [degrees]C). Medicine & Science in Sports & Exercise, 1983. 15(2): p. 156. 17. Kozey, S., et al., Errors in MET Estimates of Physical Activities Using 3.5 ml. kg-1min-1 as the Baseline Oxygen Consumption. Journal of physical activity & health, 2010. 7(4): p. 508.

98

Chapter Six: Study #2 - The Effect of Ambient Temperature on the Energy Cost of Infantry Activities

Authors

Iva Mandic1, Mavra Ahmed2, Mary L’Abbe2, Len Goodman3, and Ira Jacobs1

Affiliation

1Faculty of Kinesiology & Physical Education, University of Toronto, Toronto ON M5S 2W6, Canada

2 Department of Nutritional Sciences, University of Toronto, Toronto ON M5S 3E2, Canada

3 Defence Research and Development Canada, Toronto ON M3K 2C9, Canada

Abbreviated Title: Ambient temperature effects on energy cost of infantry tasks

Key terms: Energy Expenditure; Indirect Calorimetry; Temperature; Self-Paced Exercise; Clothing Weight

Author contribution:

Mandic, I. contributed to research design, data acquisition, analysis and interpretation of data, drafting of the manuscript and revision.

Ahmed, M. contributed to research design, data acquisition, and manuscript revision.

L’Abbe, M. contributed to research design and manuscript revision.

Goodman, L. contributed to data acquisition and manuscript revision.

Jacobs, I. was the principal investigator for the research contract that funded this research. He contributed to research design, analysis and interpretation of data, drafting of the manuscript and revision.

99

6.1 Abstract

The purpose of this research was to assess the need for caloric supplementation of Canadian Armed Forces (CAF) field rations during operations in thermally stressful ambient conditions. The energy expenditure (EE) in a convenience sample of 18 CAF members was determined using indirect calorimetry during 8-h sessions in a thermally controlled chamber. One sedentary session (resting for 8 h at 21.1 ± 0.3 °C, RH 29± 2%) and three physically active sessions were completed on four different days. During the active sessions participants spent 4 h completing 50 different activities, once in a hot (30.1 ± 0.2°C, RH 31 ± 1%), once in a temperate (21.0 ± 0.2 °C, RH 32 ± 4%), and once in a cold (-10.4 ± 0.4°C, RH 56 ± 3%) environment. Overall, throughout 4 h of physical activity, participants expended ~10 kcal•h-1 more in the hot (mean EE=443±6kcal•h-1) and cold (mean EE=444±7kcal•h-1) conditions than in the temperate (mean EE=431±7kcal•h-1) condition (p<0.05). When only considering self-paced activities, participants self-selected lower work-rates in the heat and EE was not different between conditions; while in the cold, the increased EE disappeared when EE was expressed relative to clothed (instead of nude) weight. Overall, minor differences between conditions were found for EE, but were not apparent when forced-paced activities were excluded. These results suggest that minimal caloric supplementation may only be warranted during short-term operations in harsher thermal environments (-10°C, 30°C) when an activity rate is imposed rather than self-selected.

New & Noteworthy:

This study demonstrates that energy expenditure is marginally increased when completing infantry tasks in hot and cold environments as compared to completing the same tasks in a thermoneutral environment. In cold environments increased energy costs were associated with the heavier clothing that was worn; whereas in hot environments the increased energy expenditure was likely because of thermoregulatory strain.

100

6.2 Introduction

Military operations may include physically demanding activities of varying durations in thermally stressful environments. Moreover, there may not be sufficient time for physiological acclimation or acclimatization after deployment before actual operations commence. Others have speculated on the effects of thermally stressful environments on energy expenditure (EE) and the results of related research have been equivocal [1]. Such knowledge is important for several reasons including understanding the nutritional consequences of operations in harsh environments, such as whether the provision of increased daily nutritional energy content in field rations is warranted during operations or training in thermally stressful conditions.

When comparing exercise in the heat vs. in a temperate environment, previous research has found EE to increase [2-5], remain relatively unchanged [6], and decrease [7, 8]. When -1 exercising on a cycle ergometer at 55% of V̇ O2max at 30°C, EE was ~73kcal∙h lower than when completing the same exercise at 20°C [8]. Alternatively, Consolazio et al. (1963) reported that EE increased by ~20kcal∙h-1when moderate cycling was completed in a hot environment (37.7°C) as compared to a warm (29.4°C) or temperate environment (21.2°C) - with no differences found between the warm and temperate conditions [3]. Increases in EE during exercise as a result of heat exposure are thought to be mediated by the increased energy cost of heat dissipation. To further convolute the issue, it has previously been found that when completing self-paced exercise in hot environments, participants self-select lower work-rates than in thermoneutral or cool environments [9, 10]. Taken together it remains equivocal as to whether one can expect EE during military field operations in a hot environment to be increased, decreased, or maintained as compared to a temperate environment.

In cold environments, if an individual feels cold, a ~7% increase in EE occurs as a result of non-shivering thermogenesis [11], once shivering is initiated, EE can increase as much as five times above the resting metabolic rate [12]. When dressed appropriately for a cold ambient temperature, there does not seem to be any difference in EE. During exercise, EE is similar between cold and temperate conditions [8, 13] and any increases in EE above that expended in temperate conditions have been interpreted as being the result of the weight [14] and hobbling effect of cold weather clothing [7]. Furthermore, there is currently no research assessing whether the pace of self-paced exercise is changed as a result of cold exposure or heavy winter clothing. As a result, the purpose of this study was to evaluate EE while simulating a day of standardized

101

army field activities in a hot (30°C), temperate (21°C) and cold (-10°C) environment in which participants wear clothing that is appropriate for the environmental condition they will endure, and where they complete tasks as they would in the field.

102

6.3 Methods

6.3.1 Participants

Twenty-seven healthy, male (n=21) and female (n=6) individuals were recruited from a convenience sample of Canadian Armed Forces (CAF) members to participate in the study (Table 16). Seven withdrew voluntarily after starting the study due to time constraints (n=3), reconsideration (n=3), or loss of contact (n=1). Two participants were also withdrawn due to their non-compliance with the study restrictions. Eighteen participants (14 male, and 4 female) completed the study.

Table 16: Participant characteristics Peak Body Body Age Height V̇ O2 Peak Heart Participants Sex Mass BMI Fat (yrs) (cm) (ml · kg-1 · min-1) Rate (kg) (%) (bpm) 14 males 33.5 ± 173.8 ± 80.4 ± 26.5 ± 185 ± Completed 23 ± 8 44 ± 6 4 females 10.8 10.3 15.7 4.0 14 7 males 28.9 ± 171.0 ± 75.3 ± 25.8 ± 189 ± Withdrawn 21± 9 44 ± 6 2 females 6.4 9.7 8.8 3.4 7 21 males 32.0 ± 172.8 ± 78.7 ± 26.3 ± 187 ± All 22 ± 8 44 ± 6 6 females 9.7 10.0 13.8 3.6 12 There were no significant differences between individuals who completed the study (completed), and those who did not (withdrawn). Data are presented as mean ± SD.

All participants were fully informed of the details, discomforts and risks associated with the experimental protocol before being asked for their written informed consent. The study protocol was reviewed and approved by the institutional human research ethics committees at Defence Research and Development Canada (#2013-075), and the University of Toronto (#29914).

6.3.2 Experimental Design

The study design is depicted in Figure 29. Participants made three initial visits to the lab, followed by an additional four visits for the experimental conditions.

103

Figure 14: Graphical depiction of the study design.

6.3.2.1 Initial Visits

Visit 1 consisted of the completion of the informed consent form, and PAR-Q+ ([15].

During Visit 2, participants completed a maximal aerobic power test (V̇ O2max) on a treadmill using a progressive incremental exercise test to voluntary exhaustion [16]. Oxygen consumption was measured throughout the exercise with a metabolic cart. During Visit 3, participants reported to the laboratory after a 10-h overnight fast, had their body mass measured using a standard scale, and had their body composition assessed via air-displacement plethysmography (BOD PODTM).

6.3.2.2 Experimental Conditions

Participants completed a total of 4 different experimental conditions; each condition started in the morning and lasted 8 h in the environmental chamber. At least one week intervened between conditions (Figure 29). For the two days prior to each condition, participants were given military field rations to consume at home for those two days (2 breakfast, 2 lunch, and 2 dinner rations). They were asked to document their food intake, consume ad libitum only the provided military rations and water, and bring back all unconsumed and partially eaten items. Participants were asked to refrain from ingesting food or water after 10 pm the night prior to each experimental condition. There were four experimental conditions: participants completed one resting condition (Sedentary) where they sat upright for 8 h in the environmental chamber with the ambient temperature controlled at 21.1 ± 0.3°C, and the relative humidity (RH) maintained at 29 ± 2%. Participants also completed three active conditions where they executed two 2-h circuits composed of a standardized set of typical military tasks (covering a range of light, moderate, and heavy work rates) with 2-h rest following each circuit. Each of

104

these three active conditions was identical except that the maintained temperature in the environmental chamber was different so that the protocol was completed: once at 30.1 ± 0.2°C, RH 31 ± 1% (Hot), once at 21.0 ± 0.2 °C, RH 32 ± 4% (Temperate), and once at -10.4 ± 0.4°C, RH 56 ± 3% (Cold). All of the participants underwent every condition and the order of these trials was assigned in a randomized manner using a computerized number generator https://www.random.org/.

6.3.2.2.1 Order of Events during each Experimental Condition

At 6:00am on the morning of each experimental condition, participants swallowed a telemetric core temperature capsule with 250ml of water at home. At 7:45am, participants reported to the laboratory for initial measurements of body mass measured using a standard scale, resting blood pressure using an automated blood pressure cuff, and core temperature using the Equivital™ (Hidalgo, Cambridge, United Kingdom). Participants were then given typical military rations to sustain them for the day (1 breakfast, 1 lunch, and 1 dinner ration was provided) which they were able to consume ad libitum from that point up until 10pm that night. All food and drinks that were consumed in the chamber were documented by the investigators. Participants consumed breakfast, and then donned the temperature appropriate military clothing for the environmental condition on that day. Participants also wore the following measurement equipment: i) a portable metabolic measurement system (Metamax 3B, Cortex, Leipzig, Germany) which weighed about 1.5 kg and required the participant to wear an oral-nasal mask over the face and an interface box strapped to the torso; ii) a heart rate strap (Polar Kempele, Finland), that was compatible with the Metamax 3B unit; and iii) a commercially-available personal activity monitor system (Equivital™, Hidalgo, Cambridge, United Kingdom) used solely as a receiver for the core temperature pill. The participants then entered the environmental chamber in which room temperature and RH were controlled for the day’s duration. Following entry into the chamber, participants either rested for 8 h (Sedentary) or they underwent two 2-h circuits composed of typical military tasks (Table 17) with 2-h rest following each circuit (Cold, Temperate, and Hot). Activities were completed during specified times, and participants were asked to complete all non-treadmill tasks ‘as they would in the field’. Immediately following each activity participants were asked to rate their perceived exertion on Borg’s 6-20 scale [17], and every hour, they were asked to rate their thermal comfort level on the Thermal Comfort Scale (a 13 point scale ranging from 1 – ‘I am so cold I am helpless’ to 13 – ‘I am so hot I am sick and nauseated’). 105

Table 17: Battery of standardized infantry tasks that were performed by all participants in the environmental chamber and the approximate times that these activities were done. Block Time Activities completed - Walking on a motorized treadmill at 2.4 km/h, 4 km/h, and 5.6 km/h, at 0%, 5% and 10% grade with and without a 10 kg day pack. 1 (ACTIVE) - Moving full and empty ammo cans from one stand to another. - Moving sandbags and full jerry cans from one corner to 8:00-10:00 another - Stooping/kneeling - Simulating the ‘Escape to Cover’ army drill - Sitting in one spot - Leopard crawling 2 10:00-12:00 - Resting and Eating (REST) - Walking on a motorized treadmill at 2.4 km/h, 4 km/h, and 5.6 km/h, at 0%, 5% and 10% grade with a loaded 20 kg military rucksack. - Moving empty jerry cans from one corner to another - Walking from one side of the chamber to the other with and without two 22.5 kg dumbbells - Lying in a drop and fire position - Stepping up onto a 20.3cm step with no weight, with a 10kg 3 (ACTIVE) 12:00-14:00 day pack, and with a 20kg rucksack - Stacking and Tamping - Flattening sandbags with a wooden handle or their foot continuously. - Standing in one spot - Lifting empty and full ammo cans from the ground - Dragging an 85 kg mannequin from one side of the chamber to the other. 4 14:00-16:00 - Resting and Eating (REST)

EE (as calculated using Acheson’s RQ-based equation), was assessed continuously with the Metamax 3B portable oxygen uptake measurement system. The unit was worn during the entire activity portions (4 h) on each chamber day and participants only took the mask off to drink water. Three different types of activities were performed in the environmental chamber: forced-pace (treadmill activities), self-paced (non-treadmill dynamic activities that include moving ammo cans, walking from one side of the chamber to the other, flattening sandbags etc.), and static (standing, sitting, kneeling, and maintaining the drop and fire position) activities.

While completing the self-paced activities, activity repetitions were counted using a standard mechanical counter. Eight hours after entry participants exited the chamber and took all

106

unfinished food home with them. They continued documenting their food intake until they went to sleep that night.

6.3.3 Statistical Analysis

Repeated measures, one-factor ANOVAs were used to examine differences between the Sedentary, Hot, Temperate, and Cold trials; when necessary data were also examined by activity type (forced-pace, self-paced, and static). Mauchly’s Test was used to test the assumption of sphericity, and where Mauchly’s test was significant, the Greenhouse-Geisser estimate epsilon was assessed. If epsilon<0.75, then the degrees of freedom were corrected using the Greenhouse-Geisser correction; if epsilon>0.75, then the degrees of freedom were corrected using the Huynh-Feldt correction [18]. Where significant main effects were found, post hoc analysis was performed using the Bonferroni correction for multiple comparisons. All data are expressed as mean± standard deviation (SD) unless otherwise stated. All statistical analysis was performed using the SPSS v. 22.0 software package and statistical significance was accepted at p<0.05.

107

6.4 Results

6.4.1 Missing Data

Due to the intermittent nature of the exercise performed during the active conditions (Hot, Temperate, and Cold), participants were expected to be active for a total of 2 h and 36 min during the 4 h of scheduled activity. However, an average of 10 min of activity data are missing (8.4 min (5%), 14.7 min (9%), and 7.0 min (4%), in the Hot, Cold, and Temperate conditions respectively) as a result of equipment problems or participant exhaustion. The equipment problems were primarily because of the portable oxygen uptake system sample line freezing, pump failure, and sample line blockage. As a result, all data presented ‘per activity’ includes only data where a participant completed the activity in all three active conditions and the data were successfully acquired for all three conditions.

6.4.2 Chamber Temperature.

All temperatures were maintained within 0.5ºC of the desired temperature as seen in Table 18. Although every effort was made to maintain RH at 30%, this level of control was not feasible for the Cold trials.

Table 18: Temperature and humidity attained during the trial conditions Cold Hot Temperate Sedentary Temperature (°C) -10.4± 0.4 30.1 ± 0.2 21.0 ± 0.2 21.1 ± 0.3 Relative Humidity (%) 56 ± 3 31.3± 1 32 ± 4 28.7± 2 Values are displayed as mean ± SD

6.4.3 Clothing Weight

During the three active conditions (Hot, Temperate, and Cold), participants wore standardized military clothing items that are mandatory during field operations; these included: a tactical vest (~3-7kg), a fragmentation protection vest (2-3kg), combat boots (1.5kg- 2.5kg), and a helmet (~1.5kg). Additionally, all participants in all conditions (Hot, Temperate, Cold, and Sedentary) wore the following measurement equipment - the portable oxygen uptake system (~1.5kg), and a heart rate telemetry chest strap (~0.004kg). Other than these required clothing and equipment, participants decided which military clothing to wear or not wear in order to be comfortable in the condition that they would be enduring. Clothing weight is defined as the sum of all mandatory clothing, measurement equipment, and optional clothing participants chose to wear.

108

There was a significant main effect of condition on clothing weight worn (p<0.05). Clothing weight was significantly lighter during the Sedentary trial than it was during all other conditions (Sedentary clothing weight= 4.89±0.48kg vs. Temperate clothing weight=13.49±1.37kg; Hot clothing weight =13.16±1.37kg; Cold clothing weight= 14.71±1.99kg) p<0.05. Participants wore significantly heavier clothing in the Cold condition than they did in any other condition (p<0.05). There was no significant difference in clothing weight between the Hot and Temperate conditions (p>0.05).

6.4.4 Core Temperature

Core temperature was recorded every ~15 seconds throughout the Temperate, Hot and Cold trials. Seven (7) of the 54 testing days were missing complete and reliable data; this is the result of recording errors and the odd ‘bad pill’. The data depicted here includes only the 11 participants whose data were complete for all three active conditions (Temperate, Hot, and Cold). The data were averaged over 1 min segments and compared between the conditions (Figure 15).

Figure 15: Average core temperature (°C) for each condition (Cold – black solid line; Hot - dotted line; Temperate – grey solid line) as transmitted by the participants whose ingestible core temperature pill data was complete for all three trials (n=11). Mean core temperatures measured during the 8 h in the environmental chamber were significantly different between conditions.* indicates significantly different from all other conditions p<0.05.

109

Core temperatures measured in each condition were significantly different from all other conditions p<0.05. Participants core temperature was the highest in the Hot condition (37.68±0.31°C), followed by the Temperate condition (37.46±0.24°C) and it was the lowest in the Cold condition (37.42±0.26°C) (p<0.05).

6.4.5 Thermal Comfort

The highest thermal comfort rating that was given throughout the study was a 12 – ‘I am almost as hot as I can stand’, and the lowest rating that was given throughout the study was a 3 – ‘I am very cold’. The majority of the thermal comfort ratings were either a 6- ‘I am cool but fairly comfortable’, 7- ‘I am comfortable’, or an 8 – ‘I am warm but fairly comfortable’. Thermal comfort ratings reported in each condition were significantly different than all other conditions in the following manner: Hot>Temperate>Sedentary>Cold (p<0.05) (Table 19).

Table 19: The average thermal comfort score reported for each condition. Condition Thermal Comfort Score Scoring Descriptor Hot 8.0±0.7* ‘I am warm but fairly comfortable’ Temperate 7.2±0.3* ‘I am comfortable’ Sedentary 6.6±0.4* ‘I am comfortable’ Cold 5.9±0.6* ‘I am cool but fairly comfortable’ *denotes significantly different than ALL other conditions (p<0.05) Data are presented as Mean±SD.

There was a significant main effect of time on thermal comfort, (p<0.05). Participants felt significantly warmer at 1 h, 2 h, 5 h, and 6 h than they did during 3 h, 4 h, 7 h, and 8 h (p<0.05). The hours the participants felt warmer (1 h, 2 h, 5 h, and 6 h) correspond to the hours that the participants were actively completing the military activities during the Hot, Temperate, and Cold conditions.

There was also a significant interaction between condition and time on thermal comfort, p<0.05. How thermal comfort ratings changed over time was significantly affected by the condition the participants were in. As expected, the Sedentary condition was the least like the other conditions since the participants remained in a thermoneutral environment and did not partake in military activities during this condition. Unsurprisingly, thermal comfort scores remained similar throughout the day during the Sedentary condition, while they changed from one hour to the next throughout the day for the other three conditions (Figure 16).

110

Figure 16: Average thermal comfort ratings reported for each condition (Sedentary – dashed grey line; Temperate – solid grey line; Hot – dotted line; Cold – black line) throughout the 8 h in the environmental chamber. The horizontal grey rectangle depicts a range of thermal comfort scores that are deemed to fall within a subjectively comfortable temperature range. The horizontal dashed line at the top of this range highlights a thermal comfort score of 8 which represents ‘warm but fairly comfortable’, while the bottom dashed line represents a thermal comfort score of 6 ‘cool but fairly comfortable’. The middle- dashed line signifies a thermal comfort score of 7 or ‘comfortable’. The lighter vertical rectangles signify the time periods during which military tasks were completed in the Hot, Temperate, and Cold conditions. Data are presented as mean±SEM.

A Pearson product-moment correlation was run to determine the relationship between the thermal comfort rating reported at each hour, and the 5-min average core temperature measured at each hour (when the thermal comfort rating was reported). A moderate, positive, statistically significant correlation was found between thermal comfort and core temperature (r=0.432, n=364, p<0.05) (Figure 17).

111

Figure 17: The 5-min average core temperature (°C) measured at each hour vs. the thermal comfort rating reported at the same time for each participant. Different symbols indicate the different conditions (Cold – ; Temperate – ; Hot – ).

When running Pearson product-moment correlations by condition between thermal comfort ratings and the 5-min average core temperatures measured at each hour, statistically significant correlations are only found for the Temperate and Hot conditions. The strongest correlation (just missing the 0.5 cut-off for a strong correlation) was found during the Temperate condition (r=0.487, n=112, p<0.05). For the Hot condition, there was also a moderate, statistically significant, positive correlation between thermal comfort and core temperature (r=0.410, n=116, p<0.05). When looking only at the data gathered during the Cold condition on the other hand, there was a trend towards a weak positive correlation between thermal comfort and core temperature (r=0.152, n=136, p=0.077).

6.4.6 Heart Rate

Heart rate was significantly different in all conditions p<0.05. Heart rate was the highest in the Hot condition (129±26bpm), followed by the Temperate condition (120±25bpm), followed by the Cold condition (115±24bpm), and lowest in the Sedentary condition (68±12bpm) (Figure 18).

112

Figure 18: Average heart rate during each condition (Hot – black dotted bar; Cold – dark grey bar; Temperate – light grey bar; Sedentary – white dotted bar) throughout the 4 h (during both ‘activity blocks’) that the Metamax was worn in the environmental chamber. * indicates significantly different from all other conditions p<0.05. Data are presented as mean±SEM

6.4.7 Rate of Perceived Exertion

Perceived exertion was significantly different in all conditions p<0.05. The military tasks felt more difficult in the Hot condition (RPE=11.2±3.2) than they did in either the Cold (RPE=10.9± 2.7) or Temperate (RPE=10.5±3.0) conditions p<0.05. Similarly, the military tasks felt significantly easier in the Temperate condition than in either the Hot or Cold conditions p<0.05 (Figure 19).The same trends emerged when activities were segmented into forced-pace activities and self-paced activities (p<0.05), during static conditions however, perceived exertion was not different between conditions (p>0.05).

Figure 19: Average Borg Scale rating during each active condition (Hot – black dotted bar; Cold – dark grey bar; Temperate – light grey bar) throughout the 4 hours of activity in the environmental chamber. Dashed horizontal lines indicate the rating descriptors for values 9 (Very Light), 11 (Light), and 13 (Somewhat Hard). * indicates significantly different from all other conditions p<0.05. Data are presented as mean±SEM.

113

6.4.8 Speed

When considering all self-paced travelling activities (activities where participants moved about the chamber - i.e. sandbag carry from one corner to another, leopard crawl, mannequin drag etc.), there was an effect of condition on speed; participants travelled significantly slower during the Hot conditions (3.45±1.01km·h-1) then they did during either the Temperate (3.61±1.14 km/h), or Cold (3.68±1.25km·h-1) conditions p<0.05. This discrepancy was large enough that even when including the forced-pace travelling activities (treadmill activities), this difference between conditions remained (Hot (3.37±1.53km· h-1) >Temperate (3.41±1.55km·h-1) and Cold (3.42±1.55km·h-1) (p<0.05). There were no differences in speed between the Cold and Temperate conditions in either case (p>0.05).

6.4.9 Repetitions

The pace of non-travelling activities (ammo can passing, flattening sandbags with a wooden handle etc.) was determined by counting the number of repetitions completed in each condition. When assessing the absolute number of repetitions completed in each condition, the number of repetitions completed in the Hot condition (38.2±57.9 reps) was significantly lower than the number completed in the Temperate condition (40.3±61.8 reps) p<0.05. The number of repetitions completed in the Cold condition (39.8±58.1 reps) however was not significantly different from either the Hot or Temperate conditions. Considering that each activity was completed within the same timeframe for all conditions, the same pattern remains when assessing repetition rate (reps·h-1): Hot (1177.2±1318.5 reps·h-1) < Temperate (1242.7±1374.8 reps·h-1) p<0.05, with no other significant differences.

6.4.10 Energy Expenditure by Time

Since our activity blocks consisted of intermittent exercise, assessing EE by time considers both the activity and recovery EE during that activity block. All trends discussed below were the same regardless of whether absolute EE or EE relative to nude weight were assessed.

When comparing 30-s averages of all of the data in the 4 conditions there was a significant main effect of condition on EE (p<0.05). EE in the Sedentary condition (98.24±31.53kcal·h-1 (1.29±0.44 kcal·kg-1·h-1)) was significantly lower than all of the other conditions (Cold EE=379.47±206.55 kcal·h-1 (4.86±2.50 kcal·kg-1·h-1); Hot

114

EE=380.04±194.67kcal·h-1 (4.88±2.40 kcal·kg-1·h-1); Temperate EE=365.14±197.47 kcal·h-1 (4.67±2.39 kcal·kg-1·h-1)) P<0.05. The amount of energy expended in the Hot and Cold conditions was the same and significantly higher than the amount of energy expended in the Sedentary and Temperate conditions (p<0.05).

6.4.11 Total Energy Expenditure

Considering that exercise was intermittent in nature in the current study, and since there is currently no gold standard for the estimation of anaerobic EE [19], similar measures to those conducted by Novas et al. were taken in order to account for the anaerobic contribution to EE [20]. Incorporation of the excess post-exercise oxygen consumption (EPOC) measured during the 15 min of recovery was included in the following manner:

Where “Activity Block EE” is the EE measured during both 2-h activity blocks (4 h); where “Recovery EE” is the EE measured during 15 min following one of the two activity blocks during the Hot, Cold and Temperate conditions, and where “Sedentary Recovery EE” is the EE measured during the identical time period during the Sedentary condition. Although the accuracy of this method is uncertain, it is suggested that the omission of any anaerobic measure is worse than the inherent error introduced by the estimate [19]. In addition, recovery EE during the active conditions (Hot, Cold, and Temperate), did not reach the values seen during the Sedentary condition by the 15th min (with resting values remaining 10-30 kcal∙h-1 above that seen during the 15th min of Sedentary recovery); and as a result, this adjustment is likely a highly conservative estimate. When 4-h EE was calculated for each participant in this way, there were no significant differences between the active conditions, the only condition that remained significantly different was the Sedentary condition (Figure 20). By looking at the data using this method (analyzing EE by participant) it is possible that the missing data are swaying the results.

115

Figure 20: Total energy expended during four hours of activity (Hot, Temperate, Cold), or rest (Sedentary). Each symbol represents a participant. * denotes significantly different from all other conditions p<0.05

6.4.12 Energy Expenditure by Activity

Assessing EE by activity ensures that EE is not influenced by missing data. Considering that some activities were self-paced, while others were forced-paced, we decided to also consider these activity groups separately. When comparing all activities (without stratifying the activities by type) the amount of energy expended in the Temperate condition was significantly lower than that expended in both the Hot and Cold conditions (Temperate activity EE=431.05±176.29 kcal·h-1 vs. Hot activity EE= 442.70±174.47 kcal·h-1; Cold activity EE=443.47±185.04 kcal·h-1) (p<0.05), with no significant differences between the Hot and Cold conditions (Figure 21). When considering only the forced-pace activities (treadmill activities) the same trends were found with EE remaining higher during the Hot (466.61±177.74 kcal•h-1) and Cold (473.38±183.27 kcal•h-1) conditions as compared to the Temperate (451.81±169.53 kcal•h-1) condition (p<0.05), with no other significant differences between Hot and Cold conditions. No significant differences were found between conditions when assessing self-paced or static activities (p>0.05).

116

Figure 21: Average energy expended (kcal·h-1) during all activities and activity subtypes (forced-pace, self-paced, and static activities), completed throughout the three active conditions (Hot – dotted bar; Cold – dark grey bar; Temperate – light grey bar). § indicates significantly different from all other conditions when comparing all activities p<0.05. ǂ indicates significantly different from all other conditions when comparing self-paced activities p<0.05. Data are presented as mean±SEM. When assessing EE relative to clothed weight as opposed to nude weight, EE is higher during the Hot condition (4.82±1.77 kcal∙kg-1∙h-1) than during either the Cold (4.70±1.79 kcal∙kg-1∙h-1) or Temperate (4.65±1.73 kcal∙kg-1∙h-1) conditions (p<0.05), with no differences between Cold and Temperate conditions (p>0.05) (Figure 22).

117

Figure 22: Average energy expended (kcal∙kg-1∙h-1) relative to nude or clothed weight throughout the three active conditions (Hot – dotted bar; Cold – dark grey bar; Temperate – light grey bar). ǂ indicates significantly different from all other conditions when comparing EE relative to nude body mass p<0.05. § indicates significantly different from all other conditions when comparing EE relative to clothed body mass p<0.05. Data are presented as mean±SEM.

118

6.5 Discussion

This study was conducted to clarify if EE is likely to be increased when CAF infantry are operating or training in thermally stressful field environments. For this purpose, the EE during a standardized set of infantry tasks was performed in a temperature controlled thermal chamber. If EE was found to be affected, then the magnitude of the effect could be used to inform decisions about whether caloric supplementation of existing field rations should be provided. The results of our research indicate that EE is likely minimally affected when personnel are operating in hot or cold environments when they are wearing the military clothing ensembles for that particular environment.

In the current study, the physiological impact of the ambient temperature was partially mitigated as participants were dressed appropriately for the thermal conditions. Thus, it is not surprising that for the great majority of the time participants felt ‘fairly comfortable’ while in the environmental chamber. Most thermal comfort scores were either a 6- ‘I am cool but fairly comfortable’, 7- ‘I am comfortable’, or an 8 – ‘I am warm but fairly comfortable’. Even still, thermal comfort ratings in all of the conditions were significantly different from one another with Hot>Temperate>Sedentary>Cold p<0.05. Similarly, although core temperature was well defended in the different ambient temperatures, there were statistical differences among the Hot, Temperate, and Cold conditions in the expected direction, with core temperature being the highest in the Hot condition and lowest in the Cold condition p<0.05. As expected core temperature and thermal comfort were correlated, however during the Cold condition this association was weakened. This is not surprising considering that thermal comfort is equally affected by both core and skin temperature [21]. Skin temperature was not measured in the current study, but it is possible that skin temperature was more affected in the Cold condition (than in either the Temperate or Hot conditions) thereby deteriorating the relationship between core temperature and subjective thermal comfort.

Although adequate clothing can temper the environmental impact on thermoregulation, the advantage of clothing selection diminishes as the level of heat stress increases. Similarly, high ambient temperatures also seem to impact EE additively. Previous studies assessing the effect of high ambient temperature on EE have been inconsistent in their findings with reported decreases [7, 8], no effect [6], and increases [2-5, 22] of EE in high ambient temperatures as compared with EE when the same physical activity is performed at comfortable temperatures.

119

Inconsistencies seem to arise as a result of methodological differences between studies (e.g. exercise employed, duration of heat exposure, participant acclimatization status, etc.) as well as discrepancies in the definitions used (e.g. "temperate" describes 16°C in some studies [7] and 25°C [5] in others, while "hot" refers to temperatures ranging from 30°C [8] to 46.2°C [22].

Studies that report decreased EE in the heat explain the EE reductions as being due to increased muscle efficiency [6] or increased metabolic efficiency in the heat [23], secondary to a preventative central nervous system-mediated mechanism by which cutaneous thermoreceptor afferents stimulate the preoptic anterior hypothalamus leading to decreased metabolic heat production [8]. It is important to note however that these studies also tend to subject their participants to a lower heat stress, such that heart rate responses between temperate and hot conditions are not different [8], or the heat exposure used was relatively short in duration, resulting in limited increases in both core temperature, and thermoregulatory processes such as sweating and skin blood flow [24].

Others report that total EE is unchanged in hot environments, but the aerobic contribution to EE (as determined by oxygen consumption) decreases, resulting in an increased anaerobic contribution to EE [25]. This commonly reported increase in the anaerobic component as a result of exercise in the heat has been postulated to be due to reduced muscle blood flow commensurate with increased skin blood flow, and reduced oxygen delivery to muscle [4, 25], increased plasma epinephrine concentrations [26], and increased enzymatic activity or Q10 effect [6].

When heat stress conditions are more severe due to higher temperature, humidity, and/or exposure duration, increased EE is often reported [2, 3, 5]. Such an increase is often attributed to an additive energy cost of heat dissipation – increased blood circulation, increased sweat gland activity, and elevated body temperature [2].

In the current study, it is not surprising that EE was marginally increased (~3%) in the heat considering that participants had a long exposure (8 h) to thermal stress and wore military clothing with textile characteristics that would interfere with heat transfer. The highest thermal comfort ratings, heart rate, and core temperature values were, as expected, found in the Hot condition, suggesting that the higher observed EE in the heat could potentially be attributed to heat dissipation.

120

Although a slight increase in EE was found in the heat in the current study, this observation was not very robust. When 4-h EE was calculated for each participant, there were no significant differences between the Hot, Temperate, and Cold conditions. Although these results may be affected by missing data, when assessing EE by activity and splitting the data by activity-type (forced-paced, self-paced, and static), EE is only significantly higher in Hot and Cold conditions (as compared to Temperate) for forced-pace (treadmill) activities, with no significant differences between conditions for self-paced or static activities further suggesting that the impact of ambient temperature on EE is at most minimal between -10°C and 30°C in appropriately clothed individuals.

It has previously been proposed that pacing (self-paced activities) is regulated by a central anticipatory mechanism that considers both the task requirements and physiological limitations, such that work intensity and speed decrease in a hot vs. cool environment in order to prevent the likelihood of reaching a critical body temperature prior to the cessation of activity [10, 27]. In the current study, participants were instructed to complete all non-treadmill activities “as they would in the field”; this resulted in participants voluntarily choosing to work at a slower rate in the heat. In the Hot condition, participants travelled ~4-6% slower than they did in either the Temperate or Cold conditions, and they completed ~5% fewer repetitions in the Hot vs. Temperate condition. Even when all activities were considered together (including all forced-pace treadmill activities) participants were still ~1% slower in the Hot condition. This was not surprising considering that lower work rates are typical in hot environments when exercise is self-paced, with: lower power outputs at 32°C vs. 23°C in cyclists [28], and slower running speeds at 35°C vs. 15°C in runners [29]. Although the decrease in speed and work rate found in the heat was small, the EE of self-paced activities was not different between the Hot, Temperate, and Cold conditions. Interestingly, even though participants were the slowest in the heat, the military activities were perceived as the most difficult (by Borg’s 6-20 scale) in the Hot conditions and the easiest in the Temperate conditions. Although this difference was small (3% between each condition), it endured even when only considering self-paced activities with only static activities perceived as the same between conditions. Overall the current results suggest that the increase in EE (that is often found when heat exposure is lengthy and results in higher heart rate and core temperature) is mitigated when activities are self-paced, even if participants feel that they are working harder in the heat.

121

When considering EE in low ambient temperatures, previous research suggests that when dressed appropriately for the cold so that shivering thermogenesis is obviated, increases in EE are due to the weight of heavy winter clothing [14] and the ‘hobbling effect’ or inefficiency of moving while wearing that clothing [7]. In the present study, EE was marginally increased (~3%) in the Cold and participants wore significantly heavier clothing in the Cold condition, than they did in any other condition (~15kg in the Cold condition vs. ~13kg in the Temperate and Hot conditions, and ~5kg in the Sedentary condition). Clothing weight was still a significant factor during the Hot and Temperate conditions due to the weight of mandatory military equipment. Since every effort was made to simulate a typical field trial, participants were expected to wear the protective equipment that would be required in the field. Although EE was significantly increased during the Cold condition as compared to the Temperate condition (when considering EE by time, or activity), when looking at EE relative to clothed body mass, there were no significant differences in EE between the Cold and Temperate conditions. This suggests that the higher EE in the cold can potentially be explained by clothing weight alone. This may especially be true since participants chose to wear more layers such as long johns, warmer trousers, and shirts, in lieu of heavy parkas that would more obviously affect movement. Additionally, when assessing speed or repetitions completed, there were no differences between the Cold and Temperate conditions, suggesting that neither cold exposure, nor clothing weight significantly affected the pace of the self-paced tasks in our study. Considering this, it is not clear why EE did not differ for self-paced tasks between Cold and Temperate and did for forced- pace tasks. Perhaps the distribution of the clothing weight differentially impacted treadmill and non-treadmill activities, considering that weight on the extremities increases the energy cost of tasks that necessitate a larger range of movement to a greater extent [30]. Whereas the treadmill activities had constant leg and arm movements, some self-paced tasks did not require the participant to move their legs (i.e. moving ammo cans from one stand to another, lifting ammo cans from the floor, flattening sandbags) and could potentially explain a smaller effect of clothing weight on these activities.

In conclusion, the results of this investigation indicated that when participants were dressed in temperature appropriate clothing, and completed military tasks, the energy costs of the military activities increased by about 10kcal∙h-1 in the hot (30°C), and cold (-10°C) environments, as compared to when they were completed in a thermoneutral (21°C) environment. These differences were statistically significant, but only amounted to an additional 122

80-120kcal∙day-1 which is equivalent to an additional slice of bread, or tablespoon of peanut butter. In the cold, the minimal increase in EE was associated with the weight of the clothing that was worn; whereas in the heat this increase was likely the result of the increased cost of heat dissipation. Moreover, this small difference disappears when forced-paced activities are excluded suggesting that minimal caloric supplementation may only be necessary during short- term operations in harsher thermal environments (-10°C, 30°C) when an activity rate is imposed.

123

6.6 References 1. Banerjee, B. and N. Saha, Effect of temperature variation in a climatic chamber on energy cost of rest and work. Environmental research, 1972. 5(2): p. 241-247. 2. Consolazio, C.F., et al., Energy Requirements of Men in Extreme Heat. The Journal of Nutrition, 1961. 73(2): p. 126-134. 3. Consolazio, C.F., et al., Environmental temperature and energy expenditures. Journal of Applied Physiology, 1963. 18(1): p. 65-68. 4. Fink, W., D. Costill, and P. Van Handel, Leg muscle metabolism during exercise in the heat and cold. European journal of applied physiology and occupational physiology, 1975. 34(1): p. 183-190. 5. Sun, Y. and N. Zhu, Experimental research on human metabolism in hot environment. Advanced Materials Research, 2012. 599: p. 241-244. 6. Rowell, L.B., et al., Human metabolic responses to hyperthermia during mild to maximal exercise. J Appl Physiol, 1969. 26(4): p. 395-402. 7. Gray, E.L., F.C. Consolazio, and R.M. Kark, Nutritional Requirements for Men at Work in Cold, Temperate and Hot Environments. Journal of Applied Physiology, 1951. 4(4): p. 270- 275. 8. Yamashita, Y., et al., Decreased energy expenditure during prolonged sub-maximal exercise in a warm environment. European Journal of Sport Science, 2005. 5(4): p. 153-158. 9. Periard, J.D., et al., Cardiovascular strain impairs prolonged self-paced exercise in the heat. Exp Physiol, 2011. 96(2): p. 134-44. 10. Tucker, R., et al., Impaired exercise performance in the heat is associated with an anticipatory reduction in skeletal muscle recruitment. Pflugers Arch, 2004. 448(4): p. 422-30. 11. Dauncey, M.J., Influence of mild cold on 24 h energy expenditure, resting metabolism and diet-induced thermogenesis. Br J Nutr, 1981. 45(2): p. 257-67. 12. Jacobs, I., L. Martineau, and A.L. Vallerand, Thermoregulatory thermogenesis in humans during cold stress. Exerc Sport Sci Rev, 1994. 22: p. 221-50. 13. Leblanc, J.A., Effect of environmental temperature on energy expenditure and caloric requirements. J Appl Physiol, 1957. 10(2): p. 281-3. 14. Romet, T., et al., The metabolic cost of exercise in a cold air environment (-20 [degrees]C). Medicine & Science in Sports & Exercise, 1983. 15(2): p. 156. 15. Warburton, D., et al., The 2014 Physical Activity Readiness Questionnaire for Everyone (PAR-Q+) and electronic Physical Activity Readiness Medical Examination (ePARmed-X+). Health & Fitness Journal of Canada, 2014. 7(1): p. 80. 16. Bruce, R.A., F. Kusumi, and D. Hosmer, Maximal oxygen intake and nomographic assessment of functional aerobic impairment in cardiovascular disease. Am Heart J, 1973. 85(4): p. 546-62. 17. Borg, G., Perceived exertion as an indicator of somatic stress. Scandinavian journal of rehabilitation medicine, 1970. 2(2): p. 92-98. 18. Girden, E.R., ANOVA: Repeated measures. 1992: Sage. 19. Scott, C.B., Estimating energy expenditure for brief bouts of exercise with acute recovery. Appl Physiol Nutr Metab, 2006. 31(2): p. 144-9. 20. Novas, A.M., D.G. Rowbottom, and D.G. Jenkins, A practical method of estimating energy expenditure during tennis play. J Sci Med Sport, 2003. 6(1): p. 40-50. 21. Frank, S.M., et al., Relative contribution of core and cutaneous temperatures to thermal comfort and autonomic responses in humans. J Appl Physiol (1985), 1999. 86(5): p. 1588-93. 22. Al-Haboubi, M.H., Energy expenditure during moderate work at various climates. International Journal of Industrial Ergonomics, 1996. 17: p. 379-388. 124

23. Brouha, L., et al., Physiological reactions of men and women during muscular activity and recovery in various environments. Journal of Applied Physiology, 1961. 16(1): p. 133-140. 24. Smolander, J., et al., Aerobic and anaerobic responses to incremental exercise in a thermoneutral and a hot dry environment. Acta Physiol Scand, 1986. 128(1): p. 15-21. 25. Williams, C.G., et al., Circulatory and metabolic reactions to work in heat. J Appl Physiol, 1962. 17: p. 625-38. 26. Febbraio, M.A., et al., Muscle metabolism during exercise and heat stress in trained men: effect of acclimation. J Appl Physiol (1985), 1994. 76(2): p. 589-97. 27. Tucker, R. and T.D. Noakes, The physiological regulation of pacing strategy during exercise: a critical review. British Journal of Sports Medicine, 2009. 43(6): p. e1-e1. 28. Tatterson, A.J., et al., Effects of heat stress on physiological responses and exercise performance in elite cyclists. Journal of Science and Medicine in Sport, 2000. 3(2): p. 186-193. 29. Marino, F.E., M.I. Lambert, and T.D. Noakes, Superior performance of African runners in warm humid but not in cool environmental conditions. J Appl Physiol (1985), 2004. 96(1): p. 124-30. 30. Dorman, L.E. and G. Havenith, Effects of simulated clothing weight distribution on metabolic rate. 2007, Loughborough University: Loughborough, UK.

125

Chapter Seven: Study #3 - Estimating Energy Expenditure in Military Personnel Using Accelerometry and Heart Rate

Authors

Iva Mandic1, Mavra Ahmed2, Paul Corey3, Mary L’Abbe2, and Ira Jacobs1

Affiliation

1Faculty of Kinesiology and Physical Education, University of Toronto, Toronto ON M5S 2W6, Canada

2Department of Nutritional Sciences, University of Toronto, Toronto ON M5S 3E2, Canada

3Department of Statistical Sciences, University of Toronto, Toronto ON M5T 3M7, Canada

Abbreviated Title: Military-Specific Energy Expenditure Algorithm

Key terms: Accelerometry, Heart Rate, Energy Expenditure, Military

Author contribution:

Mandic, I. contributed to research design, data acquisition, analysis and interpretation of data, drafting of the manuscript and revision.

Ahmed, M. contributed to research design, data acquisition, and manuscript revision.

Corey, P. contributed to the statistical analysis of the data

Labe, M. contributed to research design and manuscript revision.

Jacobs, I. was the principal investigator for the research contract that funded this research. He contributed to research design, analysis and interpretation of data, drafting of the manuscript and revision.

126

7.1 Abstract

An algorithm was developed to estimate energy expenditure (EE) during infantry field operations. Since ambient temperature and clothing weight may significantly affect EE, the relevance of these factors was also addressed. The model was developed from data acquired from 17 Canadian Armed Forces (CAF) members (13 male, 4 female) who completed one sedentary, and three active 8-h sessions in a thermally controlled chamber. During the sedentary sessions, participants rested at 21°C whereas during the active sessions, they spent 4 of the 8 h completing 50 infantry tasks (covering a range of sedentary to heavy work rates), once in a hot (30°C), once in a temperate (21°C), and once in a cold (-10°C) environment. During the trials, participants wore the army clothing ensemble mandated for operations at each ambient temperature and were fitted with an accelerometer, a portable oxygen uptake measurement system, and a heart rate telemetry unit. The algorithm was developed using the activity data from the three-active sessions and was then tested on data collected from an additional 8 CAF members (6 male, 2 female) who completed at least one of the four sessions. A two-regression model was developed using fixed-effects analysis and identified two predictors of EE, expressed as METs: accelerometry vector magnitude (VM) and heart rate (HR)-principally the percentage of maximum heart rate (%MHR). Based on a composite score of (√푉푀 + 퐻푅), a different equation was used: such that when (√푉푀 + 퐻푅) was ≤ 150 then EE(METs)= - 0.636+(0.046*√푉푀)+(4.357*%MHR) and when (√푉푀 + 퐻푅) was > 150 then EE(METs)= - 2.961+(0.051*√푉푀)+(8.211*%MHR). The model was able to account for 82% of the variation in EE, had a mean bias of -0.21 METs, and a SEE of 1.04 METs. Predicted EE for each condition fell within 15% of measured EE. Neither the environmental temperature, nor the clothing worn meaningfully improved the prediction of EE. In conclusion, this algorithm is appropriate for calculating EE for ambient temperatures ranging between -10°C and 30°C, when dressed appropriately for the thermal environment, and the clothing ensemble weighs between 5kg and 17kg.

127

7.2 Introduction

For populations such as military personnel during operations or training in the field, the ability to estimate energy expenditure (EE) has important strategic and operational implications related to advance planning and decisions about the quantity and quality of food to provide to sustain field operations. In military personnel, EE can vary greatly depending on the mission at hand, and potentially as a result of the environment encountered [1]. The most valid EE estimation methods in the field can be expensive and intrusive to administer, such as continuous measurements of metabolic rate via indirect calorimetry or doubly labelled water (DLW) techniques. This has led to several investigations into the validity of a variety of other methods to estimate EE in the field in military personnel. Accelerometers in particular have been at the forefront of this trend, with more and more algorithms emerging in the literature in recent years [2-4].

Accelerometers are used to quantify cumulative daily movement through measurements of uni-axis or multi-axis acceleration. Although accelerometers cannot detect load carriage, are often unable to distinguish between the terrain traversed (e.g. flat vs. sloped, solid vs. swampy/snow covered), and can only detect acceleration of the body part they are attached to [5], many accelerometry-based algorithms have adequately estimated EE [6], especially when estimating EE for repetitive activities (marching, running, etc.), while taking height, body mass and load carried into account [7, 8]. Not all studies however, have such favourable results; algorithms based solely on accelerometry and personal variables have previously underestimated EE by 50%-60% in older persons [9], were unable to adequately predict EE in military populations due to the variability in military tasks completed by different military groups [10], and when 9 commonly used accelerometry-based algorithms were recently evaluated, none were considered to be equally accurate across a broad range of activity categories [11].

While EE estimation through the sole use of heart rate (HR) monitors is accompanied by its own set of limitations, mainly revolving around the many factors that can affect HR-EE relationships such as emotion, caffeine intake and sleep deprivation [12, 13]; the combined use of accelerometry and HR for EE estimation appears to be promising. The accelerometers allow verification of movement with increases in HR, thereby decreasing the likelihood of EE overestimations as a result of non-movement increases in HR. With this combination method,

128

one study found group EE estimations for running and walking to be very accurate (within 0.54%) when compared to direct calorimetry, whereas individual estimations were less exact (within 14%) [14]. In free living situations in sub-Saharan Africa, the combined method was less ideal in predicting physical activity EE by accounting for only 16% of the total variance measured by DLW [15]. Lastly in a study measuring EE in young European men, estimations using the combination method were very comparable to DLW when the relationship between HR and EE was individually calibrated [16]. The estimations for all of these studies are more accurate on a group level than on an individual level, making this combination method useful for making larger scale EE inferences. For best results, it is generally accepted that the algorithm used to estimate EE should be developed in a similar population performing similar activities to the target population [5, 17, 18].

Given our interest in military operations in the field the influence of ambient temperature is relevant. Current EE algorithms do not typically take ambient temperature, or clothing weight into account. Since both (ambient temperature and clothing weight) are capable of significantly impacting EE [19-23], and military personnel wear heavy clothing and are frequently involved in physical labour outdoors, we think it is important to consider these variables in EE algorithms for field use. Thus, this study involves the simultaneous measurement of EE using a portable metabolic unit, accelerometry and HR during a day of simulated infantry activities performed in different temperatures while wearing temperature specific clothing. The specific purpose of this study was to develop an EE algorithm for field use with infantry personnel that would involve minimally intrusive measurements such as HR and accelerometry, and could be applied to military operations in ambient temperatures ranging between -10°C and 30°C.

129

7.3 Methods

7.3.1 Participants

Twenty-seven healthy, male (n=21) and female (n=6) Canadian Armed Forces (CAF) members volunteered to participate in the study. Of the initial 27, 25 participants completed at least one active session in the environmental chamber; the remaining 2 withdrew due to time constraints (n=1) or reconsideration (n=1) prior to completing any activities in the environmental chamber. The data from all 25 participants was analyzed and included in this study.

Of the 25 participants with usable data, 5 withdrew voluntarily after starting the study due to time constraints (n=2), reconsideration (n=2), or loss of contact (n=1). Two participants were also withdrawn due to their non-compliance with the study restrictions. As a result18 participants completed the entire study. Of these 18, no useful data were obtained from 1 participant because of equipment problems in the hot condition. Therefore, the data of 17 participants (13 male, and 4 female) were available for the development of the predictive equations generated in this study. The remaining 8 participants (6 male, and 2 female) with partial data were used to validate the developed algorithm.

All participants were fully informed of the details, discomforts and risks associated with the experimental protocol before being asked for their written informed consent. The study protocol was reviewed and approved by the institutional human research ethics committees at Defence Research and Development Canada (#2013-075), and the University of Toronto (#29914).

7.3.2 Experimental Design

The study design is depicted in Figure 23. Participants made three initial visits to the lab, followed by an additional four visits for the four experimental conditions.

130

Figure 23: Graphical depiction of the study design.

7.3.2.1 Initial Visits

Visit 1 consisted of the completion of the informed consent form and the PAR-Q+ [24].

During Visit 2, participants completed a maximal aerobic power test (V̇ O2 max) on a treadmill using a progressive incremental exercise test to voluntary exhaustion [25]. Oxygen consumption was measured throughout the exercise with a metabolic cart. Individual maximal heart rates (using transmitter/telemetry unit (Polar Electro PE3000, Finland)) were also recorded during this test. During Visit 3, participants reported to the laboratory after a 10-h overnight fast, had their body mass measured using a standard scale, and had their body composition assessed via air-displacement plethysmography (BOD PODTM).

7.3.2.2 Experimental Conditions

Participants completed a total of 4 different experimental conditions; each condition started in the morning and lasted 8 h in the environmental chamber. At least one week intervened between conditions (Figure 23). For the two days prior to each condition, participants were given military field rations to consume at home for those two days (2 breakfast, 2 lunch, and 2 dinner rations). They were asked to document their food intake, consume ad libitum only the provided military rations and water, and bring back all unconsumed and partially eaten items. Participants were asked to refrain from ingesting food or water after 10 pm the night prior to each experimental condition. There were four experimental conditions: participants completed one resting condition (Sedentary) where they sat upright for 8 h in the environmental chamber with the ambient temperature controlled at 21.1 ± 0.3°C, and the relative humidity (RH) maintained at 29 ± 2%. Participants also completed three active conditions where they executed two 2-h circuits composed of a standardized set of typical military tasks (covering a 131

range of light, moderate, and heavy work rates) with 2-h rest following each circuit. Each of these three active conditions was identical except that the maintained temperature in the environmental chamber was different so that the protocol was completed: once at 30.1 ± 0.2°C, RH 31 ± 1% (Hot), once at 21.0 ± 0.2 °C, RH 32 ± 4% (Temperate), and once at -10.4 ± 0.4°C, RH 56 ± 3% (Cold). All of the participants underwent every condition and the order of these trials was assigned in a randomized manner using a computerized number generator https://www.random.org/.

7.3.2.2.1 Order of Events during each Experimental Condition

At 6:00am on the morning of each experimental condition, participants swallowed a telemetric core temperature capsule with 250ml of water at home. At 7:45am, participants reported to the laboratory for initial measurements of body mass measured using a standard scale, resting blood pressure using an automated blood pressure cuff, and core temperature using the Equivital™ (Hidalgo, Cambridge, United Kingdom). Participants consumed breakfast, and then donned the temperature appropriate military clothing for the environmental condition on that day. Participants also wore the following measurement equipment: i) an ActiGraph wGT3X-BT triaxial accelerometer (ActiGraph, Pensacola, Florida) that was initialized to collect data in 1-s epochs and was placed between the right hip and the participant’s navel; ii) a portable metabolic measurement system (Metamax 3B, Cortex, Leipzig, Germany) which weighed about 1.5 kg and required the participant to wear an oral-nasal mask over the face and an interface box strapped to the torso; iii) a heart rate strap (Polar Kempele, Finland), that was compatible with the Metamax 3B unit; and iv) a commercially-available personal activity monitor system (Equivital™, Hidalgo, Cambridge, United Kingdom) used solely as a receiver for the core temperature pill. The participants then entered the environmental chamber in which room temperature and relative humidity were controlled for the day’s duration. The environmental chamber was 5.9 metres long and 4.5 metres wide. Following entry into the chamber, participants either rested for 8 h (Sedentary) or they underwent two 2-h circuits composed of typical military tasks (Table 20) with 2-h rest following each circuit (Cold, Temperate, and Hot).

132

Table 20: Battery of standardized infantry tasks that were performed by all participants in the environmental chamber and the approximate times that these activities were done. Block Time Activities completed - Walking on a motorized treadmill at 2.4 km/h, 4 km/h, and 5.6 km/h, at 0%, 5% and 10% grade with and without a 10 kg day pack. - Moving full and empty ammo cans from one stand to another. 1 8:00- - Moving sandbags and full jerry cans from one corner to another (ACTIVE) 10:00 - Stooping/kneeling - Simulating the ‘Escape to Cover’ army drill - Sitting in one spot - Leopard crawling 10:00- 2 - Resting (REST) 12:00 - Walking on a motorized treadmill at 2.4 km/h, 4 km/h, and 5.6 km/h, at 0%, 5% and 10% grade with a loaded 20 kg military rucksack. - Moving empty jerry cans from one corner to another - Walking from one side of the chamber to the other with and without two 22.5 kg dumbbells - Lying in a drop and fire position 12:00- 3 - Stepping up onto a 20.3cm step with no weight, with a 10kg day pack, and (ACTIVE) 14:00 with a 20kg rucksack - Stacking and Tamping - Flattening sandbags with a wooden handle or their foot continuously. - Standing in one spot - Lifting empty and full ammo cans from the ground - Dragging an 85 kg mannequin from one side of the chamber to the other. 14:00- 4 - Resting (REST) 16:00

The Metamax 3B unit collected breath-by-breath oxygen uptake data that were averaged over 30-s intervals for each subject and expressed as metabolic equivalents (METs), where 1 -1 -1 MET= 3.5 mL O2∙kg ∙min . The unit was worn throughout the activity portions (4-h) on each chamber day and participants only removed the face-mask to drink water. Eight hours after entry participants exited the chamber.

7.3.3 Analysis

7.3.3.1 Data Cleaning for Algorithm Development

The algorithm was developed from the activity data of the development group (n=17). Mean values for EE, HR, and accelerometry were calculated for each completed activity each participant executed in each condition. The data from the Sedentary trials were not used to develop the algorithm. The following data cleaning procedures were also employed to ensure that only the relevant data were included in algorithm development.

133

7.3.3.1.1 Metabolic and Heart Rate Data

Participants completed ~50 different military tasks throughout the 4 h of activity during the Hot, Cold, and Temperate trials. Depending on the task duration, different data cleaning procedures were employed for EE and HR data collected:

i) Some of the activities were very high intensity and therefore so short in duration (<120s) that a steady state for oxygen uptake and HR were not achieved (e.g. mannequin drag, leopard crawl). Therefore tasks that lasted <120 s were excluded from algorithm development.

ii) The data for activities that lasted between 120-210 s were visually inspected and if a plateau was reached for both EE and HR, the last 30 s were analyzed; otherwise the activity was excluded from analysis. A steady state was attained for over half of the activities that lasted between 120-150 s.

iii) When task duration was >210 s, the first 180 s were excluded to ensure that a steady state was in fact attained for both EE and HR and the remaining portion was analyzed.

7.3.3.1.2 Accelerometry

Vertical axis (Axis 1), horizontal axis (Axis 2), perpendicular axis (Axis 3), and vector magnitude (VM) counts·s−1 were first multiplied by 60 to yield axis 1 counts·min−1, axis 2 counts·min−1, axis 3 counts·min−1, and VM counts·min−1. EE algorithms typically refer to either the vertical counts which are derived from Axis 1, or VM which takes all three axes into account and is equal to:

Pearson correlations between METs and the various accelerometry outputs were strongest between METs and VM (r=0.726, n=2178, p<0.05) as opposed to correlations between METs and axis 1 (r=0.663, n=2178, p<0.05), METs and axis 2 (r=0.477, n=2178, p<0.05), and METs and axis 3 (r=0.551, n=2178, p<0.05). As a result, our algorithm was based on the VM variable, instead of either Axis 1, 2, or 3.

Since the activities were repetitive in nature, VM was also relatively cyclical throughout the activity period for each activity (Figure 24).

134

Figure 24: Accelerometry and metabolic data that was used to develop the algorithm. This graph shows the energy expenditure and accelerometry data for one participant during the ‘Step Up’ activity with no load. The dashed black lines denote the time that the participant was engaged in the activity.

Therefore additional measures were taken in order to guarantee that analyzed data for each activity were indeed during the period of time that the participant was exercising (this ensured that if there were signal timing synchronization misalignments between the accelerometers and HR monitors caused by technical problems, that they would not affect the results). As a result, the following data cleaning measures were taken:

i) If the activity lasted less than 180 s, the middle third of the data was analyzed (i.e. if the activity lasted 120 s, the middle 30 s (seconds 30-90) were analyzed).

ii) If the activity lasted longer than 180 s, the first and last minute were omitted and the rest of the data were analyzed.

7.3.3.2 Statistical Analysis

Preliminary data analysis involved examining the strength of relationships between EE (METs) and accelerometer output, HR, ambient temperature, and various personal variables (height, fat percentage, aerobic fitness etc.) using Pearson’s product-moment coefficient of correlation (r). Scatter plots were also utilized to further evaluate relationships between potential

135

predictors and METs. In addition, transformations of the data were attempted to improve these relationships.

The final EE algorithm was established using a fixed-effects model. The dependent variable for all attempted models was METs while, HR, accelerometry, ambient temperature, as well as various individual physical variables (including height, sex, fat percentage, clothing weight, peak oxygen consumption, and age) were assessed as potential predictors of EE. The coefficient of determination (R2), and standard errors of the estimate (SEE) were used to assess goodness of fit. Agreement between the measured METs (indirect calorimetry) and predicted METs (algorithm-determined) was assessed by Bland-Altman plots which graphically depict the differences and 95% limits of agreement. Mean differences between predicted and measured METs were also determined by paired t-tests. All data are expressed as mean± standard deviation (SD) unless otherwise stated. All statistical analysis was performed using the SPSS v. 22.0 software package and statistical significance was accepted at p<0.05.

136

7.4 Results

7.4.1 Participants

7.4.1.1 Sample used for algorithm development:

Steady-state EE and accelerometry data in all three active conditions (Hot, Temperate, and Cold) were obtained from each of the 17 participants. The participant characteristics are displayed in Table 21. The algorithm was developed on participants who ranged in age from 18 - to 53 years, weighed between 51 and 117kg, and had a V̇ O2 peak between 31 and 53 ml·kg 1·min-1.

7.4.1.2 Sample used for algorithm validation:

Participants whose data the algorithm was validated on were very similar in every respect to the participants who were part of the sample used for the development of the prediction equations (Table 21). No significant differences emerged when comparing any of the personal characteristics of the two groups. The participants in this group were however more similar to each other with an age range between 20 and 39 years, a body mass between 61 and 87 kg, and a -1 -1 V̇ O2 peak between 34 and 49 ml·kg ·min .

Table 21: Participant characteristics. Age Height Body Mass Body Fat V̇ O Peak Participants Sex BMI 2 (yrs) (cm) (kg) (%) (ml · kg-1 · min-1) Used for 13 males 33.7 ± 173.4 ± 26.5 ± algorithm 80.1 ± 16.1 23 ± 8 44 ± 6 4 females 11.1 10.5 3.9 development Used for 6 males 29.0 ± 169.3 ± 26.7 ± algorithm 76.5 ± 10.2 23 ± 9 43 ± 6 2 females 6.6 9.9 3.5 validation 19 males 32.3 ± 172.2 ± 26.6 ± All 79.0 ± 14.5 23 ± 8 43 ± 6 6 females 10.1 10.3 3.7 Data are presented as means ± SD. There were no significant differences between individuals whose data were used to develop the algorithm and those whose data were used to validate the algorithm.

7.4.2 Correlations

Relationships between METs and many potential predictor variables using Pearson’s product-moment coefficient of correlation (r) were assessed (Table 22). Large positive correlations were found between METs and HR, as well as METs and VM. Interestingly no other variables were significantly correlated with METs.

137

Table 22: Correlations between variables of interest Sex (coded as Body Clothing METs HR VM Temperature Age Height V̇ O2 peak 1=male Fat Weight # (BPM) (counts·min−1) (°C) (years) (cm) (ml · min-1) and (%) (kg) 2=female) METs# 1 .683** .726** .003 .034 -.012 -.038 -.041 -.032 .031 HR (BPM) .683** 1 .548** .121** -.384** .097** -.078** .133** -.077** .039

VM ** ** ** ** ** ** (counts·min−1) .726 .548 1 .031 -.066 -.085 .117 -.036 .114 -.012 Temperature ** ** (°C) .003 .121 .031 1 .007 -.007 .000 -.017 .002 -.357 Age (years) .034 -.384** -.066** .007 1 .105** -.338** .251** -.273** -.124**

Sex (1=M and ** ** ** ** ** ** ** 2=F) -.012 .097 -.085 -.007 .105 1 -.696 .665 -.630 -.272 - Height (cm) -.038 -.078** .117** .000 -.338** -.696** 1 .857** .386** .486** Body Fat (%) -.041 .133** -.036 -.017 .251** .665** -.486** 1 -.283** -.058** VO peak - 2 -.032 -.077** .114** .002 -.273** -.630** .857** 1 .512** (ml · min-1) .283** Clothing - .031 .039 -.012 -.357** -.124** -.272** .386** .512** 1 Weight (kg) .058** # Body mass was not included in the correlations as METs are calculated from oxygen consumption and body mass; METs = (VO2 (L/min) / body mass (kg)*1000) /3.5 **. Correlation is significant at the 0.05 level (2-tailed).

7.4.3 Prediction Models

Considering the wide range of activity types that characterize common infantry tasks, a wide variety of activities were included in the current study for the development of the algorithm. A strong linear relationship is not always found between VM and METs for less cyclic activities like vacuuming, mowing the lawn, moving boxes etc.[26]. As a result, the relationship between VM and METs was weakened in the current study in two ways due to the large variety of activities that were included: 1) a distinct cluster of data points near 0 VM emerged as additional stationary activities were included, and 2) the dispersion of the VM data was not uniform throughout each level of METs. As METs increased the variation in VM counts increased as well (Figure 25:(A)). Visual inspection of residual scatter plots confirmed that the assumption of homoscedasticity was not met. Therefore, several transformations were attempted to normalize the variances across METs; square root transformation of VM counts (√푉푀) was the most successful and also significantly improved the coefficient of determination (R2) (Figure 25:(B)).

It was quickly determined that a single equation was not capable of adequately estimating EE for the wide range of activities performed. This is not surprising as previous researchers

138

have divided their data based on activity type, due to the varied relationships different activity classes have with EE [6, 18, 27]. Unfortunately classifying data by activity type proved ineffective in our dataset as largely different activities could not be distinguished from one another using the accelerometry and HR output available to us. When looking at the raw data for example, a stationary repetitive activity with a heavy load (e.g. full ammo can passing), had similar accelerometry and HR data as a treadmill activity with no load (walk with no load at 2.4km/h at 5% grade) (VM=1379 counts·min−1, HR=107bpm, METs=5.3; vs. VM=1388 counts·min−1, HR=113bpm, METs= 4.0).

Although dividing the data by activity type could not effectively be done, it was beneficial to separate stationary/low intensity data from the rest, similarly to previous researchers parsing out "activities of daily living" or "other activities" [18]. In addition, considering that the HR-EE relationship is also not the same during light activities and more strenuous ones [12] the data were divided in a manner that considered the output gathered from both accelerometry and HR.

This was done by the addition of HR to √푉푀 which resulted in a composite score (√푉푀 + 퐻푅) that was plotted against METs (Figure 25:(C)). A distinct grouping was visible when the composite score (√푉푀 + 퐻푅) was below 150 (Figure 25:(C)) as a result, the data were split in the following manner and a regression model was developed for each group resulting in a two-regression model for the prediction of EE:

If the composite score (√푉푀 + 퐻푅) is ≤ 150 then:

EE (METs) = -0.636 + (0.046 x √푽푴) + (4.357 x %MHR)

If the composite score (√푉푀 + 퐻푅) is > 150 then:

EE (METs) = -2.961 + (0.051 x √푽푴) + (8.211 x %MHR)

Where EE= Energy Expenditure; VM=Vector magnitude in counts·min−1; and %MHR=Percentage of maximal heart rate.

Algorithms that incorporate HR often require the researcher to determine individual HR- EE relationships in order to accommodate for various factors including age, physical fitness, and efficiency of movement between participants [28, 29]. The determination of individual HR-EE relationships requires additional time that is often difficult to obtain during infantry field operations. Thus in an effort to mitigate these factors without requiring individual HR-EE

139

relationships, the incorporation of a percentage of maximal heart rate (%MHR – as measures on visit 2) instead of raw HR, [30], was included in the present algorithm.

Figure 25: Regression plots of (A) A accelerometer counts per minute (vector magnitude (VM)) vs. Energy expenditure (metabolic equivalents (METs)). The smaller oval indicates the distinct cluster of points that emerge around 0 VM symbolizing stationary activities. The larger oval indicates the dynamic activities; (B) the square root transformation of accelerometer B counts per minute (√VM)) vs. Energy expenditure (metabolic equivalents (METs)); (C) The composite score of the square root transformation of accelerometer counts per minute plus heart rate in beats per minute (√VM + HR). The dashed line depicts how the data were divided. C

The final EE algorithm was established using a fixed-effects model. Initially various mixed models were attempted with the inclusion of a random intercept for each participant. Mixed modelling when applied to repeated measures data considers how both within and between-individual factors affect EE. However the personal variables (such as age, height, and fat percentage) included in these mixed models all seemed to over-fit the data resulting in similar EE estimations (as compared to the final fixed effects model) when tested on the data from the Development group, and significantly worse estimations when tested on the data from the Testing group. This demonstrates that the variation between the participants in the Development group overly influenced the EE estimations. To limit this bias, fixed effects models are recommended [31] because unlike mixed models, fixed effects models are exclusively based on within-individual variation, and are not influenced by between-individual variation [31]. Overall this reduces the bias that can be introduced by personal variables (a

140

necessity for the application of this algorithm on diverse military groups), but can potentially result in a less precise model [31].

7.4.4 Validation

Figure 26 displays the measured METs and the predicted METs using the two-regression model in the validation participants for each activity (A) and each minute (C) in all of the conditions. When considering the activity data (where each data point represents a participant's mean METs for each activity in each condition), the two-regression model was able to account for 74% of the variation in the activity EE, had a standard error of the estimate (SEE) of 1.20 METs (Figure 26 (A)), and had a mean bias of -0.15 METs (Figure 26 (B)).When considering the minute by minute data (where each data point represents the participant's mean METs for each minute in each condition), the two-regression model was able to account for 82% of the variation in EE, had an SEE of 1.04 METs (Figure 26 (C)), and had a mean bias of -0.21 METs. Suggesting that overall our model slightly overestimates EE by 0.21 METs, and that 95% of data points fall within 2.04 METs of measured EE.

141

Figure 26: Regression and Bland-Altman plots. (A) Regression plot of Measured METs vs. Predicted METs for the validation participants in all conditions (Hot, Temperate, Cold, and Sedentary) where each data point represents a participant's mean METs for each activity in each condition. (B) Bland-Altman plot showing residuals (actual-predicted) for the two-regression model where each data point represents a participant’s mean METs for each activity in each condition. The solid line depicts the mean difference (bias) between methods; the dashed lines depict the limits of agreement (LOA; 1.96 SD). (C) Regression plot of Measured METs vs. Predicted METs for the cross-validation participants in all conditions (Hot, Temperate, Cold, and Sedentary) where each data point represents a participant's mean METs for each minute. (D) Bland-Altman plot showing residuals (actual-predicted) for the two-regression model where each data point represents a minute of data in each condition. The solid line depicts the mean difference (bias) between methods; the dashed lines depict the limits of agreement (LOA; 1.96 SD).

7.4.4.1 Predicting the Energy Cost of Activities

Of the 46 different activities, the two-regression model predicted within 1 MET of measured values for all but 6 activities (1 treadmill (wearing a 20 kg rucksack at 5.6km/h at 5% grade) and 5 non-treadmill (Empty ammo can (3.2kg) passing; Full ammo can (13.7 kg) passing; Empty ammo can (3.2kg) load and unload truck simulation; Empty ammo can (3.2 kg) lift; and Flattening sandbags with a wooden handle (1 kg) continuously) activities. For the treadmill activities, all but 2 (wearing a 10 kg day pack at 2.4km/h at 0% grade; and wearing a 10 kg day

142

pack at 4km/h at 0% grade) activities were predicted within 15% of the actual EE. When considering the spread of the data, only 3 treadmill activities (wearing a 10 kg day pack at 2.4km/h at 0% grade; wearing a 10 kg day pack at 4km/h at 0% grade; and wearing a 10 kg day pack at 5.6km/h at 10% grade) had mean predictions outside of the standard deviation of the measured EE (Figure 27). For the 20 non-treadmill activities, the two-regression model over- estimated the EE for 8 activities (Sit in one spot; Stand in one spot; Walk from one side of chamber to the other; Empty ammo can (3.2kg) passing; Full ammo can (13.7 kg) passing; Empty ammo can (3.2kg) load and unload truck simulation; Empty ammo can (3.2 kg) lift; and Flattening sandbags with a wooden handle (1 kg) continuously) by more than 15%. Only 4 activities on the other hand, had mean predictions outside of the standard deviation of the measured EE (Empty ammo can (3.2kg) passing; Full ammo can (13.7 kg) passing; Empty ammo can (3.2kg) load and unload truck simulation; and Empty ammo can (3.2 kg) lift) (Figure 28).

When including all activity data there was a significant difference between the measured (5.56±2.39METs) and predicted (5.71±1.92METs) EE t(517) = -2.81, p<0.05. However when excluding the three empty ammo can activities (Empty ammo can passing, Empty ammo can load and unload truck simulation, and Empty ammo can lift), no significant difference between measured (5.61±2.43METs) and predicted (5.69±1.93METs) EE is found t(485) = -1.55, p>0.05.

143

Figure 27: Measured METs vs. Predicted METs for the cross-validation participants in all active conditions (Hot, Temperate, and Cold) for each treadmill activity. The white bars ( ) depict the average measured METs for each activity, and the black bars ( ) depict the average predicted METs for those same activities. The outlined, light grey vertical boxes represent the standard deviation of the measured METs, and the darker grey vertical boxes represent the standard deviation of the predicted METs.

144

Figure 28: Measured METs vs. Predicted METs for the cross-validation participants in all active conditions (Hot, Temperate, and Cold) for each non-treadmill activity. The white bars ( ) depict the average measured METs for each activity, and the black bars ( ) depict the average predicted METs for those same activities (when the average is the same for measured and predicted, only the white bar is visible). The outlined, light grey vertical boxes represent the standard deviation of the measured METs, and the darker grey vertical boxes represent the standard deviation of the predicted METs.

145

7.4.4.2 Predicting Energy Expenditure by Time

When predicting EE over 4-h of data collection, most individual estimates (63 of all (development and validation) 85 trial days (74%); or 11 of the 17 days in the validation participants (65%) fell within 15% of the measured values (Table 23). At a group level, estimates were significantly better; when assessing the trials by condition; average predictions for Cold were within 2% when considering all 25 participants, and within 1% when considering only the 8 validation participants. Average predictions for Hot were within 8% when considering all participants, and within 9% when considering only the validation group. Mean predictions for Temperate were within 5% for all participants, and within 12% when considering only the validation group, and lastly Sedentary trial predictions were within 2% for all and within 13% for the validation participants. The 13% error during the sedentary trials averages out to just over 10kcal per hour.

146

Table 23: Measured 4 h energy expenditure (kcal) vs. predicted 4 h energy expenditure (kcal) per participant per trial. COLD HOT TEMPERATE SEDENTARY Measured Difference Measured Difference Measured Difference Measured Difference Measured Predicted Measured Predicted Measured Predicted Measured Predicted Participant – from – from – from – from 4 h EE 4 h EE 4 h EE 4 h EE 4 h EE 4 h EE 4 h EE 4 h EE Number Predicted Measured Predicted Measured Predicted Measured Predicted Measured (kcal) (kcal) (kcal) (kcal) (kcal) (kcal) (kcal) (kcal) (kcal) (%) (kcal) (%) (kcal) (%) (kcal) (%) D1 1298 1078 -220 -17% 1405 1608 203 14% 1262 1298 36 3% 455 389 -66 -15% D2 1033 1188 155 15% 1132 1206 73 6% 981 1144 164 17% 366 363 -3 -1% D3 1214 1323 109 9% 1229 1483 254 21% 1150 1461 311 27% 328 293 -35 -11% D4 1155 1029 -126 -11% 1078 1028 -50 -5% 1014 1020 6 1% 351 272 -79 -22% D5 1803 1971 168 9% 1778 2594 817 46% 1660 2210 550 33% 355 525 170 48% D6 1556 1354 -202 -13% 1579 1507 -72 -5% 1586 1394 -192 -12% 510 568 58 11% D7 1487 1512 25 2% 1457 1691 234 16% 1367 1586 219 16% 319 405 85 27% D8 1510 1162 -348 -23% 1471 1144 -328 -22% 1350 1056 -294 -22% 406 283 -122 -30% D9 1914 1962 49 3% 1713 2202 488 29% 1911 1987 76 4% 470 491 21 4% D10 1682 1546 -135 -8% 1783 1790 7 0% 1642 1505 -138 -8% 434 453 19 4% D11 1527 1607 79 5% 1440 1410 -30 -2% 1348 1389 41 3% 378 389 11 3% D12 1197 1080 -118 -10% 1213 1035 -178 -15% 1188 1189 0 0% 318 337 19 6% D13 1436 1331 -104 -7% 1630 1512 -119 -7% 1657 1540 -117 -7% 467 401 -66 -14% D14 1500 1610 109 7% 1889 2253 365 19% 1843 1773 -70 -4% 490 466 -24 -5% D15 1767 1830 62 4% 1890 2110 220 12% 1853 1925 72 4% 493 473 -20 -4% D16 2150 1908 -242 -11% 1657 1886 229 14% 1588 1737 148 9% 409 363 -46 -11% D17 1752 1745 -7 0% 1801 1827 25 1% 1853 1707 -146 -8% 333 312 -21 -6% Mean±SD 1528 ± 1484 ± -44 ± -3% ± 1538 ± 1664 ± 126 ± 7% ± 1485 ± 1525 ± 39 ± 3% ± 399 ± -1% ± for 405 ± 67 -6 ± 68 296 322 155 11% 263 449 274 17% 302 336 204 14% 87 18% Development

V1 1636 1639 4 0% 1465 1929 464 32% 434 541 107 25%

V2 312 346 34 11% V3 983 1018 35 4% 1091 1195 105 10% 237 280 43 18%

V4 1579 1799 220 14% V5 1465 1571 106 7% 442 417 -25 -6% V6 2125 1942 -183 -9% 2023 2003 -21 -1%

V7 1340 1198 -141 -11% 1228 1243 15 1% V8 1187 1450 263 22% 1251 1492 240 19% 263 313 50 19% Mean±SD 1572 ± 1557 ± -14 ± -1% ± 1459 ± 1578 ± 119 ± 9% ± 1312 ± 1485 ± 173 ± 12% ± 379 ± 13% ± for 337 ± 96 42 ± 47 413 314 201 15% 448 428 131 9% 185 340 199 14% 104 12% Validation Mean±SD 1537 ± 1498 ± -38 ± -2% ± 1523 ± 1647 ± 124 ± 8% ± 1452 ± 1517 ± 5% ± 395 ± 2% ± 65 ± 205 390 ± 77 5 ± 66 for ALL 310 314 159 11% 294 436 250 15% 288 329 14% 89 18% “D” (in “Participant Number” column) denotes that the participant was part of the algorithm development group; whereas “V” denotes that the participant was part of the validation group

147

7.5 Discussion

In order for a prediction algorithm to be useful in the military population it needs to be relatively simple, and non-intrusive. Algorithms that require individual calibration, resting metabolic rate (RMR) measurements, or excessive equipment are limited in the ease and agility with which they can be applied in infantry field operations or training scenarios. As a result, the purpose of this study was to develop and validate a simple algorithm using inexpensive devices that could be used under field conditions during infantry operations to predict EE. The current model was developed considering factors which we feel are integral to military operations but others have omitted (ambient temperature and clothing weight) and included the largest number of infantry activities reported in a single study to date. The developed two-regression model was able to account for 82% of the variation in EE and had a SEE of 1.04 METs. Estimations of group EE fell within 9% of measured EE (when considering all 17 days endured by the validation group); and when considered by condition, average predictions were within 9% of measured EE for the Hot trials, within 12% of measured EE for the Temperate trials, within 1% of measured EE for the Cold trials, and within 13% of measured EE for the Sedentary trials.

Many prediction equations used to estimate EE from accelerometry measurements have been based on data collected from and/or applied to the general population [6, 32, 33]. As such, these equations are typically based on walking, running, office (filing papers, working at a computer), and household (sweeping, washing dishes) activities [6, 33, 34]. Considering that the most effective algorithm for a specific population is one that is developed on a similar population engaged in similar activities to the target population, a handful of military-specific EE algorithms have been developed over the years [7, 8, 10, 35, 36] each with their own set of limitations.

The vast majority or military-specific EE algorithms were developed solely on males [7, 8, 35, 36], and often males with highly similar physical characteristics (age, height, weight, fat percentage etc.) [8, 36] implying that these algorithms may not be applicable to military groups whose characteristics vary more widely. In addition, many military-specific EE algorithms are reported to have good agreement between predicted and measured values, however most of these results are based on tests conducted on the same data that the algorithms were developed on [7, 8, 10, 36] suggesting that the true agreement is lower than reported. Lastly, the limitations of use of some of these military-specific EE algorithms are highly restrictive for the intended

148

population: the algorithm by Hoyt et al. is predominantly meant for walking and running activities with declining relevance as the addition of other activities increases; whereas other algorithms cannot differentiate between load-free and load carriage situations [8, 10, 36] a rather problematic limitation considering the frequency of load carriage in military domains.

In contrast, the current algorithm was developed on a diverse sample, validated on data from participants not included in algorithm development, and used a combination of accelerometry and HR to better detect load carriage vs. using accelerometry alone.

To the best of our knowledge, only one algorithm utilizing both accelerometry and HR has previously been developed in the military population [18, 35]. This algorithm requires the data analyst to apply a different linear regression based on which of 5 activity classes (1- walking, 2-marching with a backpack, 3-materials handling, 4- running, and 5-other activities) the completed activity falls into. Although this method seemed promising (R2=0.88 and SEE=0.42 METs, as calculated from provided values), when applied to the current data set regardless of whether the data were classified by the equipment output [18], or by categorizing the activities by their name and description, the predicted values were not nearly as accurate (Table 24).

Table 24: Comparison between the current algorithm and Wyss & Mader 2011. Algorithm Data applied to Classification Determinant R2 SEE [35] Wyss & Mader, 2011 testing by Output as per [18] 0.88 0.42* data ͌ [35] Current validation data by by Output# as per [18] 0.63 2.12 activity ǂ [35] Current validation data by By activity name and 0.65 1.85 activity ǂ description Current algorithm Current validation data by N/A 0.74 1.20 activity ǂ *data originally displayed in kJ/min converted to METs for comparison ͌The data Wyss & Mader, (2011) validated their algorithm on: 12 Swiss Army males (Age= 21.2±3.0 years; Body mass=72.6±10.7kg; Height=179.3±4.5cm) ǂ Validation data from the current study analyzed by each completed activity from: 8 CAF members (6 male, 2 female; Age= 29.0 ± 6.6; Body mass=76.46 ± 10.2kg; Height=169.3 ± 9.9cm) # Considering the current study equipped participants with only 1 accelerometer, discrimination between walking with or without a backpack was done manually by activity name vs. by backpack accelerometry as was suggested by [18].

The misclassification of activities certainly worsened prediction accuracy, considering that the estimation improved when activities were classified by activity name and description instead of equipment output. Although some researchers have had some success with classifying 149

sedentary activities (standing, sitting, lying down) and some walking and running activities [37, 38], as the number of activity classifications increases, the likelihood of misclassification of activities seems to increase as well especially when tested on a new sample, or when classifying data from unsupervised activity. Ermes et al. (2008) for example only correctly classified activities 55% of the time when attempting to classify data collected outside of a laboratory setting [39]. Consequently, it is not surprising that the current activities were misclassified when using Wyss & Mader's algorithm, and that the resulting estimation suffered accordingly.

In the current study activity classification was not done as activities could not be easily divided based on the accelerometery and HR output gathered. A larger number of activities were performed in the current study as compared to previous studies; this made classifying the current activities problematic even when considering the descriptions of these activities. For example, should the same type of activity using a different load (i.e. full ammo can lift vs. empty ammo can lift) be classified in the same group? Although the accelerometry output was practically the same, light, arm-dominant activities (including all activities utilizing the empty ammo can) were found to have a different relationship with HR than their heavier weight counterparts (the full ammo can activities). This suggests that their activity groups should not be the same, however considering the equipment output alone (without objectively knowing the load of the items moved), useful activity classification is unattainable.

Considering that the military personnel wear significantly heavier clothing, and are required to work in various thermal environments, the current study considered whether the inclusion of clothing weight and/or the thermal temperature significantly improved the estimate. When either variable was included in the model the actual contributions (although statistically significant) were negligible. In addition, when considering correlations between METs and the many potential predictor variables (Table 22) only accelerometry and HR had statistically significant relationships with METs, further validating the inclusion of solely these two predictors. Remarkably, EE on a group level was accurately estimated during the sedentary trial (within 2% of measured EE when considering all 25 study participants) even though the algorithm was developed solely on activity data collected during the Hot, Temperate and Cold trials; clothing worn during the sedentary trial was also on average ~9kg lighter than clothing worn during the other 3 trials (Sedentary clothing = 4.9±0.5kg vs. Hot clothing = 13.2±1.4kg;

150

Cold clothing = 14.7±2.0kg; Temperate clothing = 13.5±1.4kg) yet EE estimation was not affected by this difference.

Although at the group level EE by condition fell within 13% of measured values, individual predictions were less accurate with individual assessments misestimating EE by as much as 34% in the validation group. As a result, this algorithm is best applied to groups as opposed to individuals; if individual estimates are required, the inclusion of individualized HR- EE relationships should be highly considered

In regard to the activities, the current algorithm was off by more than 1 MET for 6 of the 46 (13%) performed activities. The vast majority of the activities that were poorly predicted were activities where lightweight tools/loads were used/moved (activities utilizing the wooden handle (1 kg), or the empty ammo can (3.2 kg)). While our algorithm overestimated the energy cost of these low weight, arm-dominant activities, once the weight of the load increased (full ammo can instead of the empty ammo can for example), the predicted EE was considerably more accurate. Arm dominant activities are known to have a different HR-EE relationship compared to leg dominant or whole-body activities [34]. Activities that elicit the same HR will have lower energy costs when the arms are predominantly used, as opposed to larger muscle groups such as the legs or whole-body activities [34, 40]. As a result, this algorithm should be used warily if it is expected that participants will engage in primarily low intensity arm exercises.

From a goodness of fit perspective, our model has a larger reported SEE compared to some other published EE algorithms [7, 8, 35]. The increased spread of the data can be explained by two factors: 1) the current algorithm was tested on data collected from a separate group of participants from those who the algorithm was developed on, whereas several studies report goodness of fit using the same data for development and testing [8, 10]; and 2) the large number of included activities, the more varied the activities are, the more variable the data will be . When assessing the data by activity, mean bias was -0.15 METs and the SEE was 1.20 METs. When the lightweight, arm dominant activities (all empty ammo can and wooden handle activities) are removed from this data set the mean bias falls to -0.02 METs, and the SEE falls to 0.97 METs. Activities were not excluded from validation in order to provide a more realistic example of algorithm performance.

151

In conclusion, the current military-specific EE algorithm estimates group level EE within 15% of actual values without individualized HR-EE relationships, RMR measurements, or expensive devices. Although the great number of diverse activities increased the variability in the data, it allowed for a more realistic example of the movement variation expected from the military population. Remarkably only HR and ACC were significantly correlated with EE, and no other potential predictors (including environmental temperature and clothing weight) meaningfully improved the prediction of EE. This simple algorithm can resultantly be applied to various military endeavours occurring in ambient temperatures ranging between -10°C and 30°C when dressed appropriately for the thermal environment endured.

152

7.6 References 1. Tharion, W.J., et al., Energy requirements of military personnel. Appetite, 2005. 44(1): p. 47-65. 2. Howe, C.C., H.J. Moir, and C. Easton, Classification of Physical Activity Cut-Points and the Estimation of Energy Expenditure During Walking Using the GT3X+ Accelerometer in Overweight and Obese Adults. Measurement in Physical Education and Exercise Science, 2017: p. 1-7. 3. Montoye, A.H., et al., Comparison of linear and non-linear models for predicting energy expenditure from raw accelerometer data. Physiological Measurement, 2017. 38(2): p. 343. 4. White, T., et al., Estimation of Physical Activity Energy Expenditure during Free-Living from Wrist Accelerometry in UK Adults. PloS one, 2016. 11(12): p. e0167472. 5. Hendelman, D., et al., Validity of accelerometry for the assessment of moderate intensity physical activity in the field. Med Sci Sports Exerc, 2000. 32(9 Suppl): p. S442-9. 6. Crouter, S.E., K.G. Clowers, and D.R. Bassett, Jr., A novel method for using accelerometer data to predict energy expenditure. J Appl Physiol (1985), 2006. 100(4): p. 1324-31. 7. Kinnunen, H., et al., Wrist-worn accelerometers in assessment of energy expenditure during intensive training. Physiological Measurement, 2012. 33(11): p. 1841. 8. Hoyt, R.W., et al., Total energy expenditure estimated using foot-ground contact pedometry. Diabetes technology & therapeutics, 2004. 6(1): p. 71-81. 9. Starling, R.D., et al., Assessment of physical activity in older individuals: a doubly labeled water study. J Appl Physiol, 1999. 86(6): p. 2090-2096. 10. Horner, F., et al., Development of an accelerometer-based multivariate model to predict free-living energy expenditure in a large military cohort. Journal of sports sciences, 2013. 31(4): p. 354-360. 11. Lyden, K., et al., A comprehensive evaluation of commonly used accelerometer energy expenditure and MET prediction equations. European journal of applied physiology, 2011. 111(2): p. 187-201. 12. Li, R., P. Deurenberg, and J.G. Hautvast, A critical evaluation of heart rate monitoring to assess energy expenditure in individuals. Am J Clin Nutr, 1993. 58(5): p. 602-7. 13. Silva, A.M., et al., Accuracy of a combined heart rate and motion sensor for assessing energy expenditure in free-living adults during a double-blind crossover caffeine trial using doubly labeled water as the reference method. European journal of clinical nutrition, 2015. 69(1): p. 20. 14. Brage, S., et al., Branched equation modeling of simultaneous accelerometry and heart rate monitoring improves estimate of directly measured physical activity energy expenditure. J Appl Physiol, 2004. 96(1): p. 343-351. 15. Assah, F.K., et al., Accuracy and validity of a combined heart rate and motion sensor for the measurement of free-living physical activity energy expenditure in adults in Cameroon. Int J Epidemiol, 2011. 40(1): p. 112-120. 16. Villars, C., et al., Validity of combining heart rate and uniaxial acceleration to measure free-living physical activity energy expenditure in young men. J Appl Physiol, 2012. 17. Härtel, S., et al., Estimation of energy expenditure using accelerometers and activity-based energy models—validation of a new device. European Review of Aging and Physical Activity, 2010. 8(2): p. 109. 18. Wyss, T. and U. Mader, Recognition of military-specific physical activities with body-fixed sensors. Mil Med, 2010. 175(11): p. 858-64.

153

19. Consolazio, C.F., et al., Energy Requirements of Men in Extreme Heat. The Journal of Nutrition, 1961. 73(2): p. 126-134. 20. Consolazio, C.F., et al., Environmental temperature and energy expenditures. Journal of Applied Physiology, 1963. 18(1): p. 65-68. 21. Dauncey, M., Influence of mild cold on 24 h energy expenditure, resting metabolism and diet-induced thermogenesis. British Journal of Nutrition, 1981. 45(02): p. 257-267. 22. Romet, T., et al., The metabolic cost of exercise in a cold air environment (-20 [degrees]C). Medicine & Science in Sports & Exercise, 1983. 15(2): p. 156. 23. Gray, E.L., F.C. Consolazio, and R.M. Kark, Nutritional Requirements for Men at Work in Cold, Temperate and Hot Environments. Journal of Applied Physiology, 1951. 4(4): p. 270-275. 24. Warburton, D., et al., The 2014 Physical Activity Readiness Questionnaire for Everyone (PAR-Q+) and electronic Physical Activity Readiness Medical Examination (ePARmed- X+). Health & Fitness Journal of Canada, 2014. 7(1): p. 80. 25. Bruce, R.A., F. Kusumi, and D. Hosmer, Maximal oxygen intake and nomographic assessment of functional aerobic impairment in cardiovascular disease. Am Heart J, 1973. 85(4): p. 546-62. 26. Matthews, C.E., Calibration of accelerometer output for adults. Medicine and science in sports and exercise, 2005. 37(11 Suppl): p. S512-22. 27. Brage, S., et al., Reexamination of validity and reliability of the CSA monitor in walking and running. Med Sci Sports Exerc, 2003. 35(8): p. 1447-54. 28. Hills, A.P., N. Mokhtar, and N.M. Byrne, Assessment of physical activity and energy expenditure: an overview of objective measures. Frontiers in nutrition, 2014. 1. 29. Strath, S.J., et al., Evaluation of heart rate as a method for assessing moderate intensity physical activity. Med Sci Sports Exerc, 2000. 32(9 Suppl): p. S465-70. 30. Saltin, B., et al., Physical training in sedentary middle-aged and older men II. Oxygen uptake, heart rate, and blood lactate concentration at submaximal and maximal exercise. Scandinavian journal of clinical and laboratory investigation, 1969. 24(4): p. 323-334. 31. Gunasekara, F.I., et al., Fixed effects analysis of repeated measures data. International journal of epidemiology, 2013. 43(1): p. 264-269. 32. Freedson, P.S., E. Melanson, and J. Sirard, Calibration of the Computer Science and Applications, Inc. accelerometer. Med Sci Sports Exerc, 1998. 30(5): p. 777-81. 33. Heil, D.P., Predicting activity energy expenditure using the Actical® activity monitor. Research quarterly for exercise and sport, 2006. 77(1): p. 64-80. 34. Strath, S.J., et al., Simultaneous heart rate-motion sensor technique to estimate energy expenditure. Med Sci Sports Exerc, 2001. 33(12): p. 2118-23. 35. Wyss, T. and U. Mader, Energy expenditure estimation during daily military routine with body-fixed sensors. Mil Med, 2011. 176(5): p. 494-9. 36. Westerterp, K.R., G.J. Rietjens, and L. Wouters, Measurement of Exercise Intensity with a Tri-Axial Accelerometer during Military Training. 2009, Maastricht, Netherlands. 37. Bonomi, A.G., et al., Detection of type, duration, and intensity of physical activity using an accelerometer. Med Sci Sports Exerc, 2009. 41(9): p. 1770-1777. 38. Karantonis, D.M., et al., Implementation of a real-time human movement classifier using a triaxial accelerometer for ambulatory monitoring. IEEE Trans Inf Technol Biomed, 2006. 10(1): p. 156-67. 39. Ermes, M., et al., Detection of daily activities and sports with wearable sensors in controlled and uncontrolled conditions. IEEE Trans Inf Technol Biomed, 2008. 12(1): p. 20-6.

154

40. Vokac, Z., et al., Oxygen uptake/heart rate relationship in leg and arm exercise, sitting and standing. J Appl Physiol, 1975. 39(1): p. 54-9.

155

Chapter Eight: Study #4 - The Effects of Exercise and Ambient Temperature on Appetite and Energy Intake

Authors

Iva Mandic,1 Mavra Ahmed,2 Shawn G. Rhind,1,3 Len S. Goodman,1,3 Mary L’Abbe,2 and Ira Jacobs1

Affiliation

1Faculty of Kinesiology and Physical Education, University of Toronto, Toronto ON M5S 2W6, Canada

2 Department of Nutritional Sciences, University of Toronto, Toronto ON M5S 3E2, Canada

3Defence Research & Development Canada, Toronto Research Centre, Toronto ON M3K 2C9, Canada

Abbreviated Title: Effects of Exercise and Temperature on Appetite

Key terms: Appetite, Exercise, PYY, GLP-1, Acylated Ghrelin

Word count: <7000 without tables and figures

Number of figures and tables: 8 figures and 8 tables

This study was submitted to “Appetite” and is currently under review (January 2018).

156

Author contribution:

Mandic, I. contributed to research design, data acquisition, analysis and interpretation of data, drafting of the manuscript and revision.

Ahmed, M. contributed to research design, data acquisition, and manuscript revision.

Rhind, S. contributed to data analysis and manuscript revision.

Goodman, L. contributed to data acquisition and manuscript revision.

L’Abbe, M. contributed to research design and manuscript revision.

Jacobs, I. was the principal investigator for the research contract that funded this research. He contributed to research design, analysis and interpretation of data, drafting of the manuscript and revision.

Disclosure Statement: This research was funded in part by Defence Research & Development Canada (DRDC). This study was approved by the Canadian Forces Surgeon General’s Health Research Program. In accordance with the Department of National Defence (DND) policy, the paper was reviewed and approved for submission without modification by the DRDC Publications Office. The authors declare that they have no competing interests.

157

8.1 Abstract

Acute exercise in a comfortable ambient temperature typically changes the hormonal environment towards appetite suppression, which is usually not followed by a compensatory drive to eat. The result is a negative energy balance that lasts for at least 24 h. Moreover, ambient temperature and energy intake are thought to be inversely related. As a result, it is not known whether the frequently reported phenomenon of exercise-induced anorexia is further exacerbated in a warm environment, and/or blunted in the cold. This study investigated the effects of exercise in three different environmental temperatures vs. rest, on appetite regulating hormones, perceptions of appetite, and food intake. In a randomized repeated-measures design, 18 moderately fit Canadian Armed Forces members (14 male, 4 female) volunteered to complete four 8-h trials in a thermally controlled chamber: one resting trial (Sedentary) where participants rested for 8 h at 21°C; and three additional 8-h trials where they completed two 2-h circuits of standardized military tasks interspersed with two 2-h rest periods, once in an ambient temperature of 30°C (Hot), once at 21°C (Temperate), and once at -10°C (Cold). Participants consumed military field rations ad libitum during the trials and were provided with a dinner ration pack to consume at home. Visual analogue scales were used to assess appetite and venous plasma samples were obtained for measurements of GLP-1, PYY, acylated ghrelin, and leptin. While neither exercise nor environmental temperature had significant effects on circulating leptin or GLP-1 levels, both exercise alone and exercise plus cold exposure caused significant increases in blood concentrations of PYY and suppressed acylated ghrelin concentrations (p<0.05). Additionally, acylated ghrelin suppression was further exacerbated in the warmer trials (Temperate, Hot). Thus, the acylated ghrelin concentration changes can be summarized as Sedentary>Cold>Temperate=Hot. Similarly, appetite was perceived as being suppressed in the heat compared to the cold (p<0.05). Irrespective of the hormonal and subjective appetite measurements, dietary intake remained remarkably similar regardless of the ambient temperature. In this study where food was freely available, variations in ambient temperature, exercise vs. rest, appetite hormone concentrations, and subjective appetite sensation had no effect on dietary intake within 24 h of acute, prolonged exercise.

158

8.2 Introduction

Several studies have assessed the acute effects of exercise on appetite, appetite-regulating hormone concentrations, and energy intake (EI) [1-10]. This literature indicates that acute exercise typically shifts the hormonal milieu towards appetite suppression during and for about 30 min following exercise, and is associated with decreases in the concentrations of the appetite stimulating hormone acylated ghrelin [1-6] and increases in various appetite suppressing hormones including peptide YY (PYY) [3, 7, 8, 10] and glucagon-like peptide-1 (GLP-1) [3, 7, 8, 10]. These hormonal shifts have been reported in association with a decrease in subjective appetite [4, 9] although not always [5, 11].

Interestingly, this “appetite suppressing” hormonal response is typically not followed by a subsequent change in EI [1, 12]. Some studies reported a decrease in EI [7, 8], while others reported minor increases in EI [10, 13]. Regardless, relative energy intake (REI) (which is equivalent to total EI minus total energy expenditure (EE)) decreases as a result of exercise causing a short-term negative energy balance [14].

It is unclear how the interaction between exercise and ambient temperature impacts appetite and EI [15]. Some studies suggest that appetite and the hormonal responses associated with appetite suppression are augmented with higher ambient temperature [13], and blunted with lower ambient temperature [16], while other studies are equivocal [2, 17]. To our knowledge, only two studies [2, 18] have reported the impact of exercise in different (hot, temperate, and cold) ambient temperatures on appetite in the same participants, but neither study assessed actual food intake.

An understanding of how exercise in varying temperatures affects appetite and EI can be important for those who participate in outdoor sports, in occupational, and military operational settings requiring prolonged outdoor physical exertion. In military personnel for example, body weight loss due to under-eating during field training and deployed operations is not uncommon even when food is plentiful [19-25]. This ‘voluntary anorexia’ can at times result in rapid weight loss that can exceed 10% of initial body weight and lead to immune suppression, decreased physical performance and cognitive function [26, 27]. Although this‘voluntary anorexia’ is often explained by disliked food palatability, insufficient time to eat, and inconvenient or lengthy food preparation [28-30], it is possible that voluntary anorexia in military personnel is also mediated by hormonal responses. 159

As such, while a benefit of exercise for those who wish to lose body mass may indeed be appetite suppression, a sustained energy deficit could be detrimental to health and performance. Hence, the purpose of this study was to investigate the effects of ambient temperatures on appetite sensation, food intake, and appetite-regulating hormone levels in military personnel completing a standardized set of infantry tasks throughout an 8-h day.

160

8.3 Methods

8.3.1 Participants

Twenty-seven healthy, male (n=21) and female (n=6) Canadian Armed Forces (CAF) members volunteered to participate in the study (Table 25). Seven withdrew voluntarily after starting the study due to time constraints (n=3), reconsideration (n=3), or loss of contact (n=1). Two participants were also excluded due to their non-compliance with the study restrictions. As a result, 18 participants (14 male, and 4 female) completed the study. All participants were free from metabolic and cardiac disorders and were not taking any medications or natural health products. All participants were fully informed of the details, discomforts and risks associated with the experimental protocol before being asked for their written informed consent. The study protocol was reviewed and approved by the institutional human research ethics committees at Defence Research and Development Canada (#2013-075), and the University of Toronto (#29914).

Table 25: Participant characteristics. Participant Height Body Body V̇ O Sex Age (y) BMI 2Peak Status (cm) Mass (kg) Fat (%) (mL·min-1·kg-1) 14 males 33.5 ± 173.8 ± 80.4 ± 26.5 Completed 23 ± 8 44 ± 6 4 females 10.8 10.3 15.7 ± 4.0 7 males 28.9 ± 171.0 ± 75.3 ± 25.8 Withdrawn 21± 9 44 ± 6 2 females 6.4 9.7 8.8 ± 3.4 21 males 32.0 ± 172.8 ± 78.7 ± 26.3 All 22 ± 8 44 ± 6 6 females 9.7 10.0 13.8 ± 3.6 There were no significant differences between individuals who completed the study (completed), and those who did not (withdrawn). Data are presented as mean ±SD.

8.3.2 Experimental Design

The study design is depicted in Figure 29. Participants made three initial visits to the lab, followed by an additional four visits for the four experimental conditions.

161

Figure 29: Graphical depiction of the study design.

8.3.2.1 Initial Visits

Visit 1 consisted of the completion of the informed consent form, PAR-Q+ [31], and the Pittsburgh Sleep Quality Index (PSQI) [32]. During Visit 2, participants had their peak aerobic power (V̇ O2peak) measured using indirect calorimetry during an incremental treadmill exercise test to exhaustion [33]. During Visit 3, participants reported to the laboratory after a 10-h overnight fast, after having completed a 3-day weighed food record to assess each participant’s typical dietary intake. Body mass was measured with a standard scale, and percent body fat was estimated via air-displacement plethysmography (BOD PODTM, COSMED, Rome, Italy). Measurements were completed two consecutive times to ensure accuracy.

8.3.2.2 Experimental Conditions

Participants completed a total of 4 different experimental conditions; each condition started in the morning and lasted 8 h in the environmental chamber. At least one week intervened between conditions (Figure 29). For the two days prior to each condition, participants were given military field rations to consume at home for those two days (2 breakfast, 2 lunch, and 2 dinner rations). They were asked to document their food intake, consume ad libitum only the provided military rations and water, and bring back all unconsumed and partially eaten items. Participants were asked to refrain from ingesting food or water after 10 pm the night prior to each experimental condition. There were four experimental conditions: participants completed one resting condition (Sedentary) where they sat upright for 8 h in the environmental chamber with the ambient temperature controlled at 21.1 ± 0.3°C, and the relative humidity (RH) maintained at 29 ± 2%. Participants also completed three active conditions where they executed two 2-h circuits composed of a standardized set of typical military tasks (covering a

162

range of light, moderate and heavy work rates) with 2-h rest following each circuit. Each of these three active conditions was identical except that the maintained temperature in the environmental chamber was different so that the protocol was completed: once at 30.1 ± 0.2°C, RH 31 ± 1% (Hot), once at 21.0 ± 0.2 °C, RH 32 ± 4% (Temperate), and once at -10.4 ± 0.4°C, RH 56 ± 3% (Cold). All of the participants underwent every condition and the order of these trials was assigned in a randomized manner using a computerized number generator https://www.random.org/.

8.3.2.2.1 Order of Events during each Experimental Condition (Figure 30) At 6:00am on the morning of each experimental condition, participants swallowed a telemetric core temperature capsule with 250mL of water at home. At 7:45am, participants reported to the laboratory for initial measurements of body mass measured using a standard scale, resting blood pressure using an automated blood pressure cuff, and core temperature using the Equivital™ (Hidalgo, Cambridge, United Kingdom). The Visual Analogue Scales for Appetite (VASA) [34] were also administered while fasting. Participants consumed breakfast, and then donned the temperature appropriate military clothing for the environmental condition on that day. Participants were also fitted with: 1) a portable metabolic measurement system (Metamax 3B, CORTEX Biophysik GmbH, Leipzig, Germany) which weighed about 1.5 kg and required the participant to wear an oral-nasal mask over the face and an interface box strapped to the torso; 2) a heart rate telemetry chest strap (Polar, Kempele, Finland), that was compatible with the Metamax 3B unit; and 3) a receiver for the core temperature pill (Equivital™, Hidalgo, Cambridge, United Kingdom). The participants then entered the environmental chamber in which room temperature and relative humidity were controlled for the day’s duration. The environmental chamber was 5.9 metres long and 4.5 metres wide.

163

Figure 30: Schematic of each experimental condition. Grey boxes represent the 2-h ‘activity blocks’, although during the Sedentary condition the participants remained inactive during these blocks. Participants arrived to the laboratory following a 10-h overnight fast; the first visual analogue scale for appetite (VASA), as well as the first blood sample were collected upon arrival. Fasting and Post-Breakfast data points were collected outside the environmental chamber prior to trial commencement. The 8-h trial began once the participant entered the chamber; this occurred within minutes of the participants completing their breakfast.

Following entry into the chamber, participants either rested for 8 h (many watched neutral movies, read books, worked quietly, or played chess to pass the time) (Sedentary) or completed two 2-h circuits of a standardized set of typical military tasks (Cold, Temperate, and Hot), with two hours of rest following each set. The infantry activities completed during the circuits included resting activities (sitting, kneeling, standing, lying down), and both aerobic (treadmill walking at various speeds, inclines, and while carrying different loads) and strength / muscular endurance (moving sandbags, ammo cans and jerry cans, dragging an 85 kg mannequin across the floor, building a sandbag barrier, etc.) activities. After 8 h, participants exited the environmental chamber and took their dinner ration – and any leftover items from their breakfast and lunch rations – home with them. Participants continued to document all food that was eaten upon leaving the laboratory and returned all waste and uneaten food the following day.

8.3.3 Measurements

Sleep quality: Sleep quality was assessed with questionnaires in order to better understand the sleep patterns of the participants since studies have shown that food intake is affected by sleep deprivation [35]. Both the PSQI -which assesses sleep over the last month (conducted once during Visit 1) [32], and the Groningen Sleep Quality Scale (GSQS) [36] – which assesses how a participant slept the night before their visit (conducted the morning of each chamber day) were used.

164

Energy expenditure (EE): Four hour EE (as calculated using Acheson’s RQ-based equation [37]), was assessed continuously with a portable oxygen uptake measurement system (Metamax 3B, CORTEX Biophysik GmbH, Leipzig, Germany). The unit was worn during the entire activity portions (4 h) on each chamber day and participants only took the mask off to drink water or on rare occasion consume food. Twenty-four-hour EE was also estimated assuming that participants slept for eight hours and that they expended 0.95 METs while asleep [38]. Additionally, since participants were required to (and confirmed that they had) abstain from exercise for the 2 days prior to and on the chamber day, it was presumed that for the remaining 12 waking hours while EE was not measured, that participants expended energy at the same rate that they had during the Sedentary condition. As a result, 24-h EE was calculated as follows:

Where EE= energy expenditure, H=Hot condition, C=Cold condition, T=Temperate condition, S=Sedentary condition, BM=body mass in kg, and where 1 MET is equivalent to 1 kcal/kg/h [38].

Dietary intake: On the experimental chamber days, participants received 1 breakfast, 1 lunch, and 1 dinner ration, providing ~4000 kcal and approximately 645g of carbohydrate, 121g of fat, and 116g of protein. There were 18 different rations to choose from (6 breakfast, 6 lunch, 6 dinner), and the participants selected 1 breakfast, 1 lunch, and 1 dinner ration which was kept the same for all of the chamber days (Hot, Temperate, Cold, Sedentary) for that participant. Investigators documented all feeding behaviours (selection, amount, and timing of food intake) and weighed all food that was not consumed. The CAF Directorate of Food Services provided (from the manufacturer) the nutritional information for all of the presented rations some of which were based on chemical analysis. Dietary intake (both EI and macronutrient intake) was determined by subtracting the energy and macronutrient content of the unconsumed food from the known quantities in the rations provided. Relative energy intake (REI) was calculated by subtracting estimated 24-h EE from the energy consumed (kcal) throughout the day. From REI, it was determined whether a participant was in a positive or negative 24-h energy balance.

Subjective appetite sensation: Ten-cm VASA were used to assess appetite sensation. Four indices were assessed: hunger, satiety, fullness and prospective eating. Subjective appetite 165

was assessed while fasting, immediately following breakfast, and every hour throughout the 8 h in the environmental chamber (Figure 30).

8.3.4 Blood Sampling

Blood Sample Timing: During each experimental trial, venous blood samples were drawn from the right or left antecubital vein using an indwelling catheter, which was kept patent via infusion of saline (0.9% sodium chloride) following each sample taken. Blood samples were taken immediately upon arrival to the laboratory (time point 0), at the end of the first exercise bout (beginning of the first rest period; time point 2 h), 30, 60, and 90 min into the first rest period (time point 2.5 h, 3 h, and 3.5 h respectively), 120 min into the first rest period (immediately before the second exercise bout; time point 4 h), at the end of the second exercise bout (beginning of the second rest period; time point 6 h), 60 and 120 min into the second rest period (time point 7 h and 8 h respectively) while in the chamber (Figure 30).

Collection Procedures: Four mL samples were taken at each blood sample time point and assayed for the plasma concentrations of leptin, acylated ghrelin, GLP-1, and PYY. Samples for the determination of GLP-1, PYY, and leptin were collected into chilled 2mL K2EDTA blood collection tubes that were pre-spiked with 167uL of a protein enzyme inhibition cocktail solution, containing 0.167mg of aprotinin (a competitive serine protease inhibitor; 500KIU/mL blood) and 20uL of dipeptidyl peptidase-IV (DPP-IV; a serine exopeptidase inhibitor) (Millipore, Darmstadt, Germany). Samples for the determination of acylated ghrelin were collected in chilled 2mL K2EDTA blood collection tubes that were injected with an 80uL 4-(2- aminoethyl) benzenesulfonyl fluoride hydrochloride (AEBSF) solution containing 2mg of AEBSF (an irreversible serine protease inhibitor) [39]. Once drawn, blood samples were centrifuged immediately at 1000 x g for 15 min at 4ºC, and then frozen at -70 ºC until subsequent biochemical analysis. An additional three (3) mL was collected at each time point and analyzed immediately for the measurement of haemoglobin and hematocrit in order to standardize for plasma volume changes [40].

Blood Analysis: Immunoreactive plasma concentrations of leptin, GLP-1, PYY, and active ghrelin were analyzed with Meso Scale Discovery (MSD) 96-Well Ultra-Sensitive Human Immunoassay Kits, using electrochemiluminescence detection on an MSD Sector Imager™ 6000 with Discovery Workbench software (version 3.0.18) (MSD®, Gaitherburg, MD,

166

USA). Three different prototype human assays were run, one for the determination of GLP-1 and leptin (Custom Human Metabolic Duplex - N45ZA-1, Rockville, Maryland), one for the determination of PYY (Human Total PYY - K151MPD-2, Rockville, Maryland), and one for the determination of active ghrelin (Custom Human Active Ghrelin - N45ZA-1, Rockville, Maryland). All assays were performed according to manufacturer’s instructions, in duplicates, and without alterations to the recommended standard curve dilutions. The sensitivity of the assays was 0.3 pg/mL for GLP-1, 56pg/mL for leptin, 13 pg/mL for PYY, and 9 pg/mL for acylated ghrelin. The inter- and intra-assay coefficients of variation were 12% and 9% for GLP- 1, 13% and <5% for leptin, 8% and 6% for PYY and 14% and 9% for acylated ghrelin respectively. All samples for one participant were run on the same plate, so as to limit inter- assay variation within subjects.

8.3.5 Statistical Analyses

Two-factor, repeated measures ANOVAs were used to examine differences between trials over time for appetite sensation, EE, dietary intake, REI, and appetite-regulating peptide hormones. Mauchly’s Test was used to test the assumption of sphericity, and where Mauchly’s test was significant, the Greenhouse-Geisser estimate epsilon was assessed. If epsilon<0.75, then the degrees of freedom were corrected using the Greenhouse-Geisser correction; if epsilon>0.75, then the degrees of freedom were corrected using the Huynh-Feldt correction [41]. Where significant main effects were found, post hoc analysis was performed using the Bonferroni correction for multiple comparisons. Area under the curve (AUC) calculations were made using the trapezoidal method. Pearson product moment correlation coefficients were also used to examine relationships between variables. Data are expressed as mean± standard deviation (SD) unless otherwise stated. Statistical analyses were carried out using the SPSS v. 22.0 software package, and statistical significance was accepted at p<0.05.

167

8.4 Results

Dietary intake (both EI and macronutrient intake) for the two days prior to each chamber day was not significantly different between conditions, nor was EI different from their habitual intake (which was determined through the 3-day weighed food record brought in on visit 3)(for detailed data see [42]). There was also no significant difference in how well the participants slept the night before each trial, as the GSQS score did not differ between any of the conditions and there was also no order effect (p>0.05).

8.4.1 Energy Expenditure (EE)

As expected, the measured EE was significantly lower during the Sedentary trial compared to the other trials, and approximately 4-times greater during the 4 h of standardized military tasks compared to the same 4 h during the Sedentary trial (p<0.05). The estimated daily EE was about 50% greater (p<0.05) during the active trials compared to the Sedentary trial (Table 26). There were no significant differences among the three activity trials for the 4 h EE or for the estimated 24 h EE (p>0.05).

Table 26. Energy expended (EE) during the four hours that the oxygen uptake measurement system was worn and estimates of daily EE on the test day. Condition Energy Expended over 4 h Estimated 24-h Energy (kcal) Expenditure (kcal) Cold 1705±309 3529±518 Hot 1691±268 3514±499 Temperate 1642±320 3465±552 Sedentary 404±68* 2227±327* *denotes significantly different from all other conditions p<0.05 Data are presented as Mean±SD.

8.4.2 Energy Intake (EI) and Relative Energy Intake (REI)

Although 24-h EI was not different between conditions, there was a significant main effect of condition on REI (p<0.05). REI was significantly higher during the Sedentary condition than it was during any of the active conditions (p<0.05). No other differences emerged (Table 27). Only 3 of 18 (17%) participants were in a negative energy balance during the Sedentary trial, whereas 80% of participants were in a negative energy balance during the active trials (Hot, Cold, Temperate). In the Cold trial, only 2 participants were in a positive energy

168

balance, whereas in the Temperate and Hot trials 4 and 5 participants had a positive energy balance respectively (Figure 31).

Table 27: Energy intake and relative energy intake by condition. Condition 24-h Energy Intake (Kcal) Relative Energy Intake (Kcal) Cold 2866±927 -663±959 Hot 3082±909 -432±974 Temperate 2923±802 -542±915 Sedentary 2934±1079 +697±1003* *denotes significantly different from all other conditions p<0.05 Data are presented as Mean±SD.

Figure 31. The 24-h energy balance for each participant recorded at the end of each trial day (Cold: ; Hot: ; Temperate: ; Sedentary: ).

8.4.3 Dietary Consumption

On average, participants consumed approximately 80% of the rations provided to them with ~60% of the consumed energy coming from carbohydrates, ~27% from fat and ~14% from protein. There were no significant effects of condition, or interaction with time for the amount of energy or macronutrients consumed; there was however a significant main effect of time on total energy consumed (p<0.05). More energy was consumed at breakfast (856±393 kcal) and dinner (1009±527 kcal) than at any 30-min interval (p<0.05) (Figure 32). During the activity periods while the portable oxygen uptake measurement device was worn, it was understandably more awkward to eat, and the majority of participants abstained from eating during this time. No food was consumed by any subject during any trial from 0-0.5 h, 0.5-1 h, 1-1.5 h, and 4-4.5 h. Therefore, it is not surprising that EI during these time periods was significantly lower when

169

compared to 2.5-3 h (261±233 kcal), 3-3.5h (163±149 kcal), 3.5-4h (113±93 kcal), and 6.5-7 h (192±166) post start (p<0.05). While in the chamber, participants tended to eat the most about 30 min following each activity period (2.5-3 h post start (261±233 kcal) and 6.5-7 h (192±166) post start); during these times, participants consumed significantly more than they did at 0-0.5 h, 0.5-1 h, 1-1.5 h, 1.5-2 h, 4-4.5 h, 4.5-5 h, 5-5.5 h, and 5.5-6 h following chamber entrance (p<0.05) (Figure 32). The amount of protein, carbohydrate, and fat consumed throughout the day followed a similar trend to the total energy consumed at each time point (Figure 32). The individual variability of EI from trial to trial was fairly large: around 50% of the time individuals consumed within 10% of their median caloric intake regardless of the condition, during the remaining trials, participant’s EIs fluctuated by more than 10% of their median caloric intake (Table 28).

Table 28: Percentage of participants whose energy intake varied by more than 10% of their median energy intake in each condition Condition Participants who consumed Participants who Participants who MORE than 10% of consumed within 10% consumed LESS than 10% median intake (%) of median intake (%) of median intake (%) Cold 11% 56% 33% Hot 39% 44% 17% Temperate 22% 56% 22% Sedentary 28% 44% 28%

170

Figure 32. Average dietary intake over time Average value of total energy (total bar height) and macronutrients (Carbohydrate: ; Fat: ; Protein: ; portions of the total bar) consumed (in kcal) throughout the experimental days. The 2-h ‘activity blocks’ occurred between 0-2 h and between 4-6 h, although during the Sedentary condition the participants were inactive during these blocks. During these time periods, participants wore an oxygen uptake measurement system and had their energy expenditure assessed. Breakfast and dinner were consumed outside of the environmental chamber with all other food consumed throughout the 8-h in the chamber. Dietary consumption is displayed in 30-minute increments throughout the 8 h inside the chamber. * indicates that amount of energy and macronutrients consumed was significantly different from all time points except for Breakfast and Dinner p<0.05. # indicates that that amount of energy and macronutrients consumed was significantly different from 0-0.5 h, 0.5-1 h, 1-1.5 h, and 4-4.5 h p<0.05. Data are presented as means. SEM is shown for total energy intake only.

8.4.4 Visual Analogue Scale for Appetite (VASA)

8.4.4.1 Time

As expected, there was a significant main effect of time on hunger, fullness, satisfaction, and prospective eating, (p<0.05). Fullness and satisfaction scores shared a similar trend over time and participants felt the least full and the least satisfied upon arrival to the lab (following a >10 h overnight fast), and they felt the most full and the most satisfied following breakfast. Hunger and prospective eating scores on the other hand, while being similar to each other, were contrary to fullness and satisfaction scores. The highest hunger and prospective eating scores occurred upon arrival to the lab, and the lowest hunger and prospective eating scores were found following breakfast (Figure 33).

171

Figure 33: Average appetite scores for all participants during all of the conditions, as collected by visual analogue scales for the 4 indices of appetite (hunger: ); fullness: ); satisfaction: ); prospective eating: ). Grey boxes represent the 2-h ‘activity blocks’, although during the Sedentary condition the participants were inactive during these blocks. Fasting and Post-Breakfast data points were collected outside the environmental chamber prior to trial commencement. The 8- h trial began once the participant entered the environmental chamber; this occurred within minutes of the participants completing their breakfast. * indicates that all 4 indices were significantly different from all other time points p<0.05. # indicates that all 4 indices were significantly different from Fasting, 2 h, 3 h, 6 h, 7 h, and 8 h p<0.05. § indicates that all 4 indices were significantly different from Fasting, 5 h, and 6 h p<0.05. ǂ indicates significantly different from Fasting, Post-Breakfast, 1 h, 4 h, and 5 h p<0.05. Data are presented as mean±SD.

8.4.4.2 Condition

There was a significant main effect of condition on hunger, fullness, satisfaction, and prospective eating (p<0.05). Participants were significantly hungrier during the Cold condition than they were during the Hot condition (Cold=4.29±1.35 vs. Hot=3.31±1.35) (p<0.05). They felt significantly fuller and more satisfied during the Hot condition than they did during the Cold condition (Hot: fullness=5.5±1.81, satisfaction=5.62±1.77 vs. Cold: fullness=4.77±1.71, satisfaction=4.75±1.68) (p<0.05). There was also a significant difference in how much participants felt they could eat, they felt that they could eat more during both the Sedentary and Cold trials, as compared to the Hot trial (Sedentary= 4.74±1.83; Cold=4.74±1.68 vs. Hot=3.69±1.77). There were no other significant differences between the conditions. There were also no significant interactions between time and condition for any of the appetite indices (Figure 34).

172

Figure 34: Average appetite scores during each condition (Cold: ; Hot: ; Temperate: ; Sedentary: ), as collected by visual analogue scales for the 4 indices of appetite (hunger, fullness, satisfaction, prospective eating). * indicates a significant difference between conditions p<0.05. Data are presented as mean±SEM.

8.4.4.3 AUC

There was also a significant effect of condition on the AUC0-8h for hunger, fullness, satisfaction, and prospective eating sensation (p<0.05). The AUC0-8h for hunger was significantly lower during the Hot trial than it was during the Cold trial (Hot=24.83±9.19 vs.

Cold=31.93±9.39) (p<0.05). The AUC0-8hfor fullness and satisfaction were significantly higher during the Hot trial than they were during the Cold trial (Hot: fullness=46.55±9.85, satisfaction=48.01±9.30 vs. Cold: fullness=39.72±8.72, satisfaction=39.82±8.59) (p<0.05). The

AUC0-8h for prospective eating was significantly lower during the Hot trial than it was during both the Cold and Sedentary trials (Hot=26.59±9.81 vs. Cold=36.85±8.88; Sedentary=36.75±10.55) (p<0.05). There were no other significant differences between the conditions (Table 29).

Table 29: The area under the curve for the 4 indices of appetite over the trial day for each condition. Condition Hunger Fullness Satisfaction Prospective Eating AUC0-8h AUC0-8h AUC0-8h AUC0-8h Cold 31.93±9.39 39.72±8.72 39.82±8.59 36.85±8.88 Hot 24.83±9.19 * 46.55±9.85 * 48.01±9.30 * 26.59±9.81 * Temperate 29.02±9.05 43.74±8.62 43.55±9.47 31.88±9.51 * Sedentary 31.85±8.88 40.76±11.49 41.73±8.86 36.75±10.55 *denotes significant difference between conditions at p<0.05 Data are presented as Mean±SD.

173

8.4.5 Appetite Hormones

All plasma concentration values were corrected for plasma volume changes [40] and for the dilution occurring from the additives added to the tubes upon sample collection (Figure 35). The vagaries of blood sampling from intravenous catheters in such experiments in an environmental chamber resulted in several missing blood samples from different subjects at different time intervals. In order to have a robust basis for inter-trial comparisons, only the data for the 8 participants for whom a complete set of blood samples are available across all four conditions are presented.

Figure 35. Average plasma volume changes relative to fasting (time 0) during each condition (Cold: ; Hot: ; Temperate: ; Sedentary: ). The grey boxes represent the 2-h ‘activity blocks’, although during the Sedentary condition participants were inactive during these blocks. The ‘0’ time point represents the fasting blood sample collected outside the environmental chamber prior to trial commencement, all other blood samples were collected in the environmental chamber. Data are presented as mean±SEM

8.4.5.1 Time

GLP-1. GLP-1 data were analyzed for 7 participants since two major outliers (>5 standard deviations above the mean) emerged when analyzing the data and were subsequently removed. GLP-1 was the lowest when the participants were fasting, with significantly higher levels at 3.5, 4, and 7 hours (p<0.05) (Figure 36). Additionally, GLP-1 levels at 4 h following chamber entrance (immediately prior to activity block 2), were significantly higher than those found at 6 h following chamber entrance (p<0.05). No other significant differences emerged.

PYY. PYY levels upon arrival to the lab were significantly lower than all time-points other than 3 h, and 4 h post start p<0.05. PYY levels at 2 h were significantly greater than they were at 2.5 h post start p<0.05.

174

Leptin. No significant differences emerged when assessing leptin levels over time when all 8 (7 male, 1 female) participants were included. However, leptin concentrations are known to be different between males and females even after accounting for fat mass differences between the sexes [43]. As a result, when the data for the 7 males was assessed separately, leptin levels were found to be significantly lower 30 min after the first activity block (2.5h) than they were immediately following the first activity block (2h) (p<0.05).

Acylated Ghrelin. No significant differences between time points were found for circulating acylated ghrelin concentrations (p>0.05)(Figure 36).

Figure 36. Average appetite hormone concentrations collected at each time point during all of the conditions. This figure displays complete data for 7 (GLP-1 and Leptin) or 8 (PYY and Acylated Ghrelin) participants. The grey boxes represent the 2-h ‘activity blocks’, although during the Sedentary condition the participants were inactive during these blocks. The ‘0’ time point represents the fasting blood sample collected outside of the environmental chamber prior to trial commencement, all other blood samples were collected in the environmental chamber. * indicates that GLP-1 concentrations were significantly lower than 3.5, 4 and 7 h p<0.05. # indicates that GLP-1 concentrations were significantly higher than, 6 h p<0.05. ǂ indicates that PYY concentrations were significantly lower than 2, 2.5, 3.5, 6, 7, and 8 h p<0.05. § indicates that PYY and leptin concentrations were significantly higher than 2.5 h p<0.05. Data are presented as mean±SEM

8.4.5.2 Condition

Many blood samples were missed throughout the study due to sampling difficulty. As a result, repeated measures ANOVAs were conducted to compare 2 conditions at a time (Condition by Time: 2x9) for each hormone in order to include more of the gathered data.

175

Regardless of whether the results were analyzed using the concentration values attained, or as a percent change from average fasting values (data not shown), the same trends emerged.

GLP-1 and Leptin. No significant differences between conditions emerged for either GLP-1 or Leptin.

PYY. Circulating PYY concentrations were significantly lower during the Sedentary condition (50.1±2.8 pg/mL) as compared to the Cold condition (62.8±6.3 pg/mL) (p<0.05), and PYY levels were also lower during the Sedentary trials than they were during the Temperate trials (49.9±2.5 pg/mL vs. 55.8±3.6 pg/mL respectively) (p<0.05).

Acylated Ghrelin. There was a significant main effect of condition on circulating acylated ghrelin concentrations between all conditions except Hot vs. Temperate. Acylated ghrelin levels were higher during the Sedentary condition than they were during any other condition (Hot: 82.4±18.0 pg/mL vs. Sedentary: 113.4±16.8 pg/mL; Cold: 112.6±20.3 pg/mL vs. Sedentary: 130.6±18.2 pg/mL; Temperate: 82.4±13.1 pg/mL vs. Sedentary: 116.1±17.9 pg/mL) (p<0.05). Acylated ghrelin levels were also higher during the Cold conditions as compared to the Hot and Temperate conditions (Hot: 91.3±24.1 pg/mL vs. Cold: 108.4±25.3 pg/mL; Temperate: 85.3±16.4 pg/mL vs. Cold: 103.8±20.8 pg/mL) p<0.05. Circulating acylated ghrelin concentrations were significantly lower during the Temperate and Hot conditions when compared to Cold and Sedentary with no significant differences between Hot and Temperate.

8.4.5.3 AUC

There was a significant main effect of condition on the AUC0-8h for circulating acylated ghrelin concentration, (p<0.05). Although the omnibus test was significant, no pairwise comparisons were found to be significantly different. No significant differences were found between conditions for the AUC0-8h for any other appetite hormone.

8.4.5.4 Correlations between hormones

When looking at correlations between the hormones of interest, there was a small but significant positive correlation between PYY and GLP-1 (r=0.25, n=571, p<0.05). Similarly, a small positive correlation between acylated ghrelin and leptin (r=0.11, n=577, p<0.05) was found. There were no other significant correlations between the hormones of interest.

176

8.4.6 Correlations between appetite variables

As expected, fasting leptin levels were significantly higher in females (p<0.05) and were significantly correlated to body fat percentage (r=0.75, n=14, p<0.05) but not body mass. No other hormones were correlated with any personal variables.

Several Pearson product-moment correlations were also performed between the four hormones of interest (PYY, acylated ghrelin, GLP-1, and leptin), the four indices of appetite sensation (hunger, fullness, satisfaction, and prospective eating), and the amount of energy (kcal) and nutrients (protein, carbohydrate, and fat) consumed in the 30 min following each sample collection/assessment of subjective appetite.

8.4.6.1 Correlations between hormones and food consumed

Pearson product-moment correlations were determined between the concentrations of the measured hormones and the amount of food (kcal) and nutrients (protein, carbohydrate, and fat) consumed in the 30 min following each sample collection. Table 30 demonstrates that there were only weak, albeit significant, correlations among some of the appetite hormones and the total energy and macronutrients consumed.

Table 30:Pearson product-moment correlations between appetite hormones and dietary consumption 30 minutes following sample collection PYY GLP-1 Leptin Acylated (pg/mL) (pg/mL) (pg/mL) Ghrelin(pg/mL) Amount of energy -0.247* (530) -0.287*(530) 0.041(530) 0.079(519) consumed (kcal) Protein (g) -0.244*(530) -0.275*(530) 0.024(530) 0.089*(519) Carbohydrate (g) -0.231*(530) -0.278*(530) 0.043(530) 0.060(519) Fat (g) -0.251*(530) -0.282*(530) 0.026(530) 0.099*(519) Pearson’s r (n). *signifies significant correlation p<0.05 8.4.6.2 Correlations between hormones and indices of appetite sensation

Similarly, when assessing the correlations between the appetite hormones and the different aspects of appetite sensation, weak correlations were detected between the hormones of interest and the four indices of appetite perception (Table 31).

177

Table 31:Pearson product-moment correlations between appetite hormones and appetite sensation PYY GLP-1 Leptin(pg/mL) Acylated (pg/mL) (pg/mL) Ghrelin(pg/mL) Hunger -0.239*(455) -0.275*(455) -0.046(455) 0.066(447) Fullness 0.235*(455) 0.216*(455) -0.301*(455) -0.133*(447) Satisfaction 0.260*(455) 0.226*(455) -0.252*(455) -0.149*(447) Prospective Eating -0.222*(455) -0.225*(455) 0.156*(455) 0.084(447) p=0.077 Pearson’s r (n). *signifies significant correlation p<0.05

8.4.6.3 Correlations between indices of appetite sensation and food consumed

Moderate positive correlations were found between hunger and the amount energy and nutrients consumed 30 min following appetite sensation assessment, similar correlations were also observed between prospective eating and food consumed. On the other hand, moderate negative correlations were found between fullness and all measures regarding energy and nutrient intake, as well as between satisfaction and the amount of energy and nutrients consumed(p<0.05) (Table 32).

Table 32: Pearson product-moment correlations between appetite sensation and dietary consumption 30 minutes following visual analogue scale completion Hunger Fullness Satisfaction Prospective Eating Amount of energy 0.406* -0.449* -0.471* 0.413* consumed (kcal) Protein (g) 0.403* -0.433* -0.451* 0.407* Carbohydrate (g) 0.388* -0.433* -0.456* 0.404* Fat (g) 0.413* -0.447* -0.470* 0.401* Pearson’s r n=645 for all *signifies significant correlation p<0.05

178

8.5 Discussion

The primary objective of this study was to determine the impact of arduous physical activity in varying ambient temperatures on appetite regulating hormone concentrations, subjective appetite sensations, and dietary intake. Although hormonal responses pointed towards appetite suppression with exercise (with a partial blunting of that response in the cold), and subjective appetite sensation data suggested that appetite was highest in the cold, and lowest in the heat, actual dietary intake was the same regardless of whether the trial was sedentary or active and also regardless of variations in ambient temperature.

Hormonal Response. While neither exercise alone, or in combination with a temperature challenge had any effect on circulating leptin or GLP-1 levels; both exercise, and exercise plus cold exposure increased circulating PYY levels, and suppressed acylated ghrelin concentrations (suggesting appetite suppression). Additionally, acylated ghrelin suppression was further exacerbated in a warmer (either Temperate or Hot) environment resulting in the following hierarchy of acylated ghrelin concentration across the conditions: Sedentary>Cold>Temperate=Hot.

Not surprisingly, leptin levels in the present study were not affected by exercise. Leptin is not typically considered to be involved in short-term appetite regulation with most studies finding no effect of exercise on circulating leptin levels [1, 44]. Interestingly, cold exposure has previously been found to decrease plasma leptin levels [45, 46], although this was not the case in the present study. Since our participants were active for 4 of the 8 h they spent in the -10°C environmental chamber, and they were dressed appropriately for the environmental condition, the significant but small (0.04°C) decrease in average core temperature between the Temperate and Cold conditions (data not shown) was likely not great enough to suppress circulating leptin levels.

Regarding GLP-1, most studies suggest that circulating GLP-1 levels increase following exercise [3, 7, 8, 10], while one study found decreases in GLP-1 following 40 minutes of moderate intensity walking in overweight/obese women [47]. In the present study, circulating GLP-1 was not significantly changed by exercise or ambient temperature. In contrast with other studies, participants in the present study were able to eat at any time; this resulted in large variations in GLP-1 at every time point. As a result, it is likely that any effect of exercise, or

179

exercise in combination with ambient temperature was masked by the substantial inter- individual and inter-sample variations.

In the present study, circulating PYY concentrations were increased similarly in the Temperate and Cold conditions. In contrast, PYY levels were not significantly increased during exercise in the heat. Exercise typically increases circulating PYY levels [2, 3, 7, 8, 10], and this increase is thought to be independent of ambient temperature [2, 16]. In the present study, the lack of effect in the Hot trial, may have been the result of insufficient statistical power, considering that PYY levels during the Temperate trial were significantly higher than they were in the Sedentary trial and the AUC0-8hfor PYY appeared higher in the Hot trial (although not statistically) than it was in the Temperate trial (25.22±8.69 ng/mL vs. 23.57±5.32 ng/mL respectively).

Exercise decreased circulating acylated ghrelin concentrations in the present study, a finding which is consistent with earlier reports [1, 4, 6, 48]. Ambient temperature has previously been reported to impact total ghrelin concentrations by increasing plasma ghrelin levels in the cold (2°C), and decreasing them in the heat (30°C), as compared to levels found at 20°C[49]. However, the interactions of ambient temperature, exercise, and ghrelin are equivocal. While, some studies found no effect of ambient temperature on acylated ghrelin concentration when exercise is performed at temperatures ranging from 10°C to 36°C [2, 18, 48], Crabtree et al.[16] found higher acylated ghrelin levels following brisk walking in the cold (8°C), as compared to a neutral environment (20°C). Similar to Crabtree et al., the present study found that the acylated ghrelin suppression that is seen with exercise is partially blunted in the cold. However, in the current study exercise in 30°C did not result in a suppression of acylated ghrelin beyond the levels observed with exercise at 21°C.

Overall, exercise shifted the hormonal milieu towards appetite suppression with ambient temperature only having minor effects beyond those associated with exercise alone.

Subjective Response. In contrast with previous studies [1, 5, 6], in the current investigation no aspect of subjective appetite was significantly affected by exercise alone (no differences between Sedentary and Temperate conditions). However, unlike most other studies of appetite, the current study’s participants could eat ad libitum throughout the trials; consequently, our participants ate whenever it suited them, thereby, increasing the inter-subject

180

variation among the appetite scores assessed at each time point. Consequently, a larger sample size may have provided the statistical power to detect a significant difference between the Sedentary and Temperate conditions.

When evaluating the impact of ambient temperature, exercise in the heat suppressed subjective appetite more than during the Cold trial; for prospective eating, the suppression was evident when compared to the Sedentary trial as well. This is in agreement with Kojima et al. (2015) who reported that hunger and motivation to eat were suppressed to a greater extent at 36°C and 24°C than they were at 12°C, while Wasse et al. (2013) found suppressed hunger and prospective eating scores at 30°C as compared to 20°C, and lower satisfaction and fullness scores at 10°C as compared to 20°C. Consistent with these studies, the current study found a similar trend towards subjective appetite increasing in the cold, and diminishing in the heat.

Dietary Intake. Interestingly, a substantial amount (4 h) of arduous exercise (expending >1200kcal above resting), even when completed in harsh ambient temperatures (-10°C, and 30°C) did not alter dietary intake (both EI and macronutrient intake).

Although previous studies have also reported no effect of acute exercise on subsequent EI [1, 12], their exercise protocols were of significantly shorter duration (<2 h). In order to stimulate EI above resting levels, some have previously suggested that high volumes of physical activity (≥2 h) [50] and/or arduous exercise [51] are necessary. To our knowledge, the participants in this study underwent the largest exercise stimulus (completing 4 h of physical activity and expending >1200kcal above resting) compared to other published studies of appetite / exercise interactions, and yet measurements of 24-h EI were remarkably similar among conditions (Sedentary intake: 2934±1079kcal vs. Temperate intake: 2923±802 kcal).

Ambient temperature alone is thought to impact EI leading to increased food consumption as ambient temperature decreases and vice versa [52, 53]. The effect of exercise in various temperature conditions on dietary intake has also been thought to follow this trend [16, 48]. Crabtree et al. found that overweight participants ate significantly more if their brisk walk was conducted in a cold, vs. temperate environment, and Wasse et al. (2013) found trends towards both lower EIs in the heat, and higher EIs in the cold. However, EI in the present study was unchanged by ambient temperature.

181

In order to decrease the likelihood of a type II error, EI was divided into 3 categories: 1) typical EI (defined as intake within 10% of a participant’s median intake), 2) low EI (consumption below 10% of a participant’s median intake), and 3) high EI (consumption above 10% of a participant’s median intake); and each trial was categorized accordingly. Even when EI was considered in this manner, no significant differences between conditions emerged; roughly half of the participants consumed within 10% of their median intake in each condition, approximately a quarter had a “low EI” in each condition, and the remaining quarter had a “high EI” in each condition.

In terms of energy balance, our study supports several other reports that acute exercise results in an acute negative energy balance [1, 8, 10, 47] in subjects who have ad libitum access to food after exercise. This was the case in the current study even though our 24 h EE estimates may have underestimated true EE (as the participants may have slept for less than 8 h, and they may have been more active after leaving the chamber than they were in the chamber on the Sedentary day). The negative energy balance detected in the active conditions (Hot, Temperate, and Cold) was likely greater than estimated, since our conservative EE assessment likely underestimated the active conditions the most, since post-exercise elevations in oxygen consumption were not in the estimation, and ‘rest’ in the cold likely would have resulted in EE above that seen in a thermoneutral environment [54].

Notwithstanding these limitations, most of our participants (15 of 18) were in a positive energy balance during the Sedentary trial and a negative energy balance (11 of 18) throughout all three active conditions (Hot, Temperate, and Cold).

It is important to note, that according to our food satisfaction surveys [55], the military rations were deemed to be ‘acceptable’ in a variety of different categories including: ease of preparation/consumption, taste, texture, saltiness, sweetness, density/fullness, digestibility and overall adequacy [42, 55]. Therefore, it is unlikely that the military rations were either over or under consumed, based on their palatability or lack thereof. Additionally, in terms of the total amount of food consumed, neither the food available (being limited to military rations vs. consuming foods usually available to them), nor the surrounding environment (at home vs. in the environmental chamber) had any effect: participants consumed the same amount of food regardless of whether they consumed military rations or their own food at home, and they

182

consumed the same amount of energy and macronutrients two days before the chamber day (at home) as they consumed on each trial day [42, 55].

Correlations. Relationships between the episodic appetite hormones and subjective appetite were found in the hypothesized directions. PYY and GLP-1 were positively correlated with both fullness and satisfaction and negatively correlated with hunger and prospective eating, while acylated ghrelin was negatively correlated with fullness and satisfaction. However, the single tonic appetite hormone leptin, was surprisingly negatively related to fullness and satisfaction, and positively correlated with prospective eating.

While most previous studies have shown no relationship between blood leptin levels and subjective appetite sensations [56, 57], positive correlations between leptin concentrations and satiety have been found following weight loss [58]. Whether previous exercise modulates this relationship is currently unclear considering that Tsofliou et al. [59] found positive correlations between leptin and both fullness and satiety and negative correlations between leptin and both hunger and desire to eat while Vatansever-Ozen et al.[1] reported no correlation between circulating leptin concentrations and any factor of subjective appetite following exercise. It is possible that the conflicting results are due to methodological differences employed in the different studies; in particular, the present study was the only one that did not schedule specific meal times, and perhaps as a result, the relationship between leptin and subjective appetite may have been modified compared with those studies where meals were not ad libitum and were scheduled for specific times.

Although food accessibility during the measurement periods likely affected the timing of meals (due to the oral-nasal mask), variations in: ambient temperature, exercise, hormone concentrations, and subjective appetite sensation had no effect on the amount and timing of food intake, suggesting that none of these factors systematically change eating patterns in any predictable way. Dietary intake in our population was surprisingly similar across conditions. Interestingly, participants were generally unaware of how much they ate; when asked whether they thought they consumed more or less than usual, participants were only correct 35% of the time, 46% of the time they thought they ate less than they did, and 19% of the time they thought they ate more than they did.

183

To the best of our knowledge, this is the first controlled laboratory investigation to simulate a strenuous 8-h work day in varying environmental temperatures and simultaneously evaluate appetite, hormonal concentrations, dietary intake, and EE. It is our belief that investigations such as the current one, with fewer limitations on food intake compared to other studies, have stronger external validity with regard to the hormonal patterns, and demonstrate that: 1) wide variations in physical activity and environmental temperature do impact both PYY and acylated ghrelin concentrations significantly enough that differences between conditions are evident even when participants are consuming food at different times; 2) exercise in the heat and cold result in distinctly different appetite sensations – subjective appetite is significantly lower when exercising in the heat as compared to the cold; 3) irrespective of 1 and 2, dietary intake remains remarkably similar regardless of condition suggesting that variations in hormone concentrations and subjective appetite have minimal effect on dietary intake.

184

8.6 References 1. Vatansever-Ozen, S., et al., The effects of exercise on food intake and hunger: Relationship with acylated ghrelin and leptin. Journal of Sports Science and Medicine, 2011. 10(2): p. 283-91. 2. Kojima, C., et al., The influence of environmental temperature on appetite-related hormonal responses. Journal of physiological anthropology, 2015. 34(1): p. 1. 3. Howe, S.M., et al., No Effect of Exercise Intensity on Appetite in Highly-Trained Endurance Women. Nutrients, 2016. 8(4): p. 223. 4. Broom, D.R., et al., Exercise-induced suppression of acylated ghrelin in humans. J Appl Physiol (1985), 2007. 102(6): p. 2165-71. 5. Broom, D.R., et al., Influence of resistance and aerobic exercise on hunger, circulating levels of acylated ghrelin, and peptide YY in healthy males. Am J Physiol Regul Integr Comp Physiol, 2009. 296(1): p. R29-35. 6. King, J.A., et al., Influence of prolonged treadmill running on appetite, energy intake and circulating concentrations of acylated ghrelin. Appetite, 2010. 54(3): p. 492-8. 7. Ueda, S.Y., et al., Comparable effects of moderate intensity exercise on changes in anorectic gut hormone levels and energy intake to high intensity exercise. J Endocrinol, 2009. 203(3): p. 357-64. 8. Ueda, S.Y., et al., Changes in gut hormone levels and negative energy balance during aerobic exercise in obese young males. J Endocrinol, 2009. 201(1): p. 151-9. 9. Mackelvie, K.J., et al., Regulation of appetite in lean and obese adolescents after exercise: role of acylated and desacyl ghrelin. J Clin Endocrinol Metab, 2007. 92(2): p. 648-54. 10. Martins, C., et al., Effects of exercise on gut peptides, energy intake and appetite. J Endocrinol, 2007. 193(2): p. 251-8. 11. Chanoine, J.P., et al., GLP-1 and appetite responses to a meal in lean and overweight adolescents following exercise. Obesity (Silver Spring), 2008. 16(1): p. 202-4. 12. Balaguera-Cortes, L., et al., Energy intake and appetite-related hormones following acute aerobic and resistance exercise. Applied Physiology, Nutrition, and Metabolism, 2011. 36(6): p. 958-966. 13. Shorten, A.L., K.E. Wallman, and K.J. Guelfi, Acute effect of environmental temperature during exercise on subsequent energy intake in active men. Am J Clin Nutr, 2009. 90(5): p. 1215- 21. 14. Schubert, M.M., et al., Acute exercise and subsequent energy intake. A meta-analysis. Appetite, 2013. 63: p. 92-104. 15. Charlot, K., C. Faure, and S. Antoine-Jonville, Influence of Hot and Cold Environments on the Regulation of Energy Balance Following a Single Exercise Session: A Mini-Review. Nutrients, 2017. 9(6): p. 592. 16. Crabtree, D.R. and A.K. Blannin, Effects of exercise in the cold on Ghrelin, PYY, and food intake in overweight adults. Medicine and science in sports and exercise, 2015. 47(1): p. 49-57. 17. Faure, C., et al., Effect of heat exposure and exercise on food intake regulation: A randomized crossover study in young healthy men. Metabolism, 2016. 65(10): p. 1541-1549. 18. Laursen, T.L., et al., Leptin, adiponectin, and ghrelin responses to endurance exercise in different ambient conditions. Temperature, 2017: p. 1-10. 19. Askew, E.W., et al., Nutrient intakes and work performance of soldiers during seven days of exercise at 7,200 feet Ready-to-Eat ration. 1986, USARIEM: Natick, MA. p. 65. 20. Askew, E.W., et al., Mauna Kea III: Metabolic effects of dietary carbohydrate supplementation during exercise at 4100 M altitude. 1987, USARIEM. 21. Guilland, J.C. and J. Klepping, Nutritional alterations at high altitude in man. Eur J Appl Physiol Occup Physiol, 1985. 54(5): p. 517-23. 185

22. Jacobs, I., et al., Physical performance and carbohydrate consumption in CF commandos during a 5-day field trial. 1989, DCIEM. 23. Tharion, W.J., et al., Energy requirements of military personnel. Appetite, 2005. 44(1): p. 47-65. 24. Fairbrother, B., et al., Nutritional and immunological assessment of soldiers during the Special Forces Assessment and Selection Course. 1995, USARIEM. 25. Hill, N.E., et al., Changes in gut hormones and leptin in military personnel during operational deployment in Afghanistan. Obesity, 2015. 23(3): p. 608-614. 26. Moore, R.J., et al., Changes in soldier nutritional status and immune function during the Ranger training course. 1992, US Army Research Institute of Environmental Medicine: Natick, MA. 27. Shippee, R., et al., Nutritional and immunological assessment of Ranger students with increased caloric intake. 1994, US Army Research Institute of Environmental Medicine: Natick, MA. 28. Baker-Fulco, C.J., et al., Nutrition for health and performance, 2001: nutritional guidance for military operations in temperate and extreme environments:. 2001: USARIEM. 29. Friedl, K.E. and R.W. Hoyt, Development and biomedical testing of military operational rations. Annual review of nutrition, 1997. 17(1): p. 51-75. 30. Marriott, B.M., Not eating enough: Overcoming underconsumption of military operational rations. 1995: National Academies Press. 31. Warburton, D., et al., The 2014 Physical Activity Readiness Questionnaire for Everyone (PAR-Q+) and electronic Physical Activity Readiness Medical Examination (ePARmed-X+). Health & Fitness Journal of Canada, 2014. 7(1): p. 80. 32. Backhaus, J., et al., Test–retest reliability and validity of the Pittsburgh Sleep Quality Index in primary insomnia. Journal of psychosomatic research, 2002. 53(3): p. 737-740. 33. Bruce, R.A., F. Kusumi, and D. Hosmer, Maximal oxygen intake and nomographic assessment of functional aerobic impairment in cardiovascular disease. Am Heart J, 1973. 85(4): p. 546-62. 34. Flint, A., et al., Reproducibility, power and validity of visual analogue scales in assessment of appetite sensations in single test meal studies. International journal of obesity, 2000. 24(1): p. 38. 35. Brondel, L., et al., Acute partial sleep deprivation increases food intake in healthy men. The American journal of clinical nutrition, 2010. 91(6): p. 1550-1559. 36. Meijman, T., et al., The evaluation of the Groningen sleep quality scale. Groningen: Heymans Bulletin (HB 88-13-EX), 1988. 2006. 37. Acheson, K. Theory, assumptions and limitations of calculating energy expenditure and substrate utilization from respiratory exchange data. in Workshop on Methodological Questions on Indirect Calorimetry. 1986. 38. Ainsworth, B.E., et al., 2011 Compendium of Physical Activities: a second update of codes and MET values. Med Sci Sports Exerc, 2011. 43(8): p. 1575-81. 39. Blatnik, M. and C.I. Soderstrom, A practical guide for the stabilization of acylghrelin in human blood collections. Clinical endocrinology, 2011. 74(3): p. 325-331. 40. Dill, D. and D.L. Costill, Calculation of percentage changes in volumes of blood, plasma, and red cells in dehydration. Journal of applied physiology, 1974. 37(2): p. 247-248. 41. Girden, E.R., ANOVA: Repeated measures. 1992: Sage. 42. Ahmed, M., Assessments of Dietary Intakes of Canadian Armed Forces Consuming Field Rations (Doctoral Thesis). 2017, University of Toronto: Toronto, Canada.

186

43. Rosenbaum, M., et al., Effects of gender, body composition, and menopause on plasma concentrations of leptin. The Journal of Clinical Endocrinology & Metabolism, 1996. 81(9): p. 3424-3427. 44. Cheng, M.H., et al., Appetite regulation via exercise prior or subsequent to high-fat meal consumption. Appetite, 2009. 52(1): p. 193-8. 45. Ricci, M.R., S.K. Fried, and K.D. Mittleman, Acute cold exposure decreases plasma leptin in women. Metabolism, 2000. 49(4): p. 421-3. 46. Zeyl, A., et al., Interactions between temperature and human leptin physiology in vivo and in vitro. European journal of applied physiology, 2004. 92(4-5): p. 571-578. 47. Unick, J.L., et al., Acute effect of walking on energy intake in overweight/obese women. Appetite, 2010. 55(3): p. 413-9. 48. Wasse, L.K., et al., Effect of ambient temperature during acute aerobic exercise on short- term appetite, energy intake, and plasma acylated ghrelin in recreationally active males. Applied Physiology, Nutrition, and Metabolism, 2013. 38(8): p. 905-909. 49. Tomasik, P.J., K. Sztefko, and M. Pizon, The effect of short-term cold and hot exposure on total plasma ghrelin concentrations in humans. Horm Metab Res, 2005. 37(3): p. 189-90. 50. Erdmann, J., et al., Plasma ghrelin levels during exercise—effects of intensity and duration. Regulatory peptides, 2007. 143(1): p. 127-135. 51. Pomerleau, M., et al., Effects of exercise intensity on food intake and appetite in women. The American journal of clinical nutrition, 2004. 80(5): p. 1230-1236. 52. Brobeck, J.R., Food intake as a mechanism of temperature regulation. Yale J Biol Med, 1948. 20(6): p. 545-52. 53. Johnson, R.E. and R.M. Kark, Environment and Food Intake in Man. Science, 1947. 105(2728): p. 378-9. 54. Dauncey, M., Influence of mild cold on 24 h energy expenditure, resting metabolism and diet-induced thermogenesis. British Journal of Nutrition, 1981. 45(02): p. 257-267. 55. Jacobs, I., The effects of environmental and physical stress on energy expenditure, energy intake, and appetite. 2016, Defence Research Development Canada - Toronto Research Centre: Toronto, Ontario. 56. Joannic, J.-L., et al., Plasma leptin and hunger ratings in healthy humans. Appetite, 1998. 30(2): p. 129-138. 57. Karhunen, L., et al., Serum leptin and short-term regulation of eating in obese women. Clinical Science, 1997. 92(6): p. 573-578. 58. Heini, A., et al., Association of leptin and hunger-satiety ratings in obese women. International journal of obesity, 1998. 22(11): p. 1084-1087. 59. Tsofliou, F., et al., Moderate physical activity permits acute coupling between serum leptin and appetite–satiety measures in obese women. International journal of obesity, 2003. 27(11): p. 1332-1339.

187

Chapter Nine: General Discussion

9.1 Summary of Findings

This dissertation involves research which was carried out for the purposes of augmenting knowledge that could be used to more accurately assess the amount of nutritional energy that should be provided to CAF infantry members who are deployed for field operations or field training. A secondary objective of this research was to investigate the roles subjective appetite, and selected appetite regulating hormones may play on food intake when EE is acutely elevated for a day in a fashion that simulates a day of common infantry field activities.

In regards to the first objective three different approaches were undertaken:

1) The energy costs of 46 clearly defined infantry tasks were accurately determined thereby providing an improved empirical basis for EE estimations arising from the use of questionnaires or the factorial method.

2) The impact of ambient temperature on EE was determined and found to be minimal (~3%) when the ambient temperature was between -10 and 30°C. This indicates that caloric supplementation on account of the requirement to carry out common tasks in temperature extremes is likely unnecessary during short term operations occurring within these ambient temperatures.

3) A simple algorithm based on accelerometry and heart rate data was developed for use with infantry personnel during field operations to facilitate future assessments of EE.

In regard to the second aim, we determined that ambient temperature influences subjective appetite responses as well as appetite regulating hormone concentrations; however absolute EI nevertheless remains unchanged. Short term EI is unaltered in spite of variations in EE, appetite sensations, appetite-regulating hormone concentrations, eating atmosphere (home vs. laboratory), or even food provided (military rations vs. food typically consumed at home). Consequently, voluntary anorexia in military personnel emerges out of the uncoupling of EE and EI in the short term, with large increases in EE not being followed by increases in EI.

188

9.2 A Physiological Guide to Ensure the Provision of Sufficient Caloric Content during Field Operations

Considering that EE is determined by the nature of the mission and the physical tasks that need to be accomplished, EE can be considered a fixed variable that EI needs to match in order to maintain energy balance. The goal therefore is to adequately determine EE in order to know how much food to provide.

As previously mentioned, there are four components of TEE: BMR, TEF, thermoregulation, and physical activity [46]. Measurements of BMR are strict, requiring participants to refrain from physical activity for 24 h, with measurements occurring in a supine position the morning following an overnight fast [241]. Resting energy expenditure (REE) on the other hand has a more flexible definition and is simply the energy cost of resting while seated or lying down [242] and is often used in place of BMR. REE is relatively stable with minimal intra-individual day-to-day variations although 4 h postprandial REE is on average 5- 6% higher than more typical BMR measurements [243]. Similarly although the energy content of a meal can impact TEF [244], TEF typically only represents 5-15% of total EE [49] and the magnitude of the variation in TEF caused by meal size is minor (following a 1200kcal meal EE was <6 kcal∙h-1 higher as compared to consuming a 600kcal meal) in regard to overall EE. In addition, in thermoneutral environments, thermoregulation also contributes minimally to EE. In contrast, physical activity is the most easily modifiable variable, and can constitute the smallest (during bed-rest) or largest (elite athletes) fraction of EE [46]. Thus, when considering EE in temperate climates, BMR, and TEF can to some degree be treated as fixed variables, and thermoregulatory thermogenesis can be omitted with minimal likely consequence for EE calculations, demonstrating the importance of accurately estimating the fraction of EE that physical activity constitutes. Although physical activity EE can be estimated in a variety of ways, one inexpensive method is the factorial method, a method which involves the accurate documentation of all of the activities engaged in (and their duration) and comparing the completed tasks to reference activities for which EE has previously been measured. When all of the activities have known energy costs, the objective of relatively accurately estimating TEE is attainable. Such a method can be useful in the military population since the physical requirements of training courses and field operations can be assessed ahead of time.

189

9.2.1 A Catalogue of Energy Costs of Infantry Tasks

Although the factorial method (and its ease of use) is attractive for the estimation of EE during field operations and for the EE prediction of upcoming field operations, few clearly described military relevant activities have known energy costs that are published in accessible journals and reports. Considering that the accuracy of the factorial method relies on the accurate matching of the activities completed in a test population with activities for which the energy costs are known, the scarcity of infantry tasks in the published literature severely limits the use of this method in military populations.

The results from the first study (The energy cost of various infantry tasks) clearly demonstrate that previously available data could not be confidently applied for the purposes of accurately estimating EE for the vast majority of common infantry tasks assessed in the current investigation. Only 21 of the 46 tasks (46%) could be confidently matched with either a Compendium or military report activity, while 8 (17%) tasks could not be matched at all. If the factorial method had been used to estimate EE for the 4-h of activity in the Temperate condition in the current investigations, using all of the data that was available through the Compendium or military reports, then EE would have been underestimated by about 12% (Table 33). In comparison, when using the “compilation of infantry tasks” data assembled in Study 1, 4-h estimates of EE were overestimated by only 2% (Table 33).

Table 33: Measured vs. Estimated 4-h Energy Expenditure During the Temperate Trial Condition Measured EE Estimated EE (kcal) over 4-h Estimated EE (kcal) over (kcal) over 4-hA using previously available 4-h using data from sourcesB Study 1C (% difference from measured) (% difference from measured) Temperate 1485±302 1308±264 (12%) 1513±305 (2%) A -1 -1 EE was determined for 17 participants by taking 1 min averages of breath by breath V̇ O2 (ml · min · kg ), dividing V̇ O2 by 3.5 to get METs and then converting METs to EE in kcal: (METs*3.5*body mass (kg))/200. All data for each participant was summed and divided by time (in min) to get EE per minute. Due to missing data, this value (EE per min) was multiplied by 240 in order to get EE over 4 hours. B Completed activities were matched with activities found in the Compendium of Physical Activities [105] as well as relevant military reports [245, 246]. Where reasonable matches could not be made, completed activities were paired with published activities that were regarded to be similar in intensity. C Average METs displayed in Study 1 were inserted for each activity. For the activities missing from the compilation reported in Study 1 (those that were too short in duration to determine an average MET value), a reference value of 10.0 METs was inserted. All recovery time was treated as the same as “sit in one spot” consequently the reference value of 1.53 METs was used. Data displayed are mean±SD with the value in brackets representing the percent difference from the measured value.

190

It was initially hypothesized that where comparable activity matches could be made (between our completed activities and reference activities with published energy costs) that published military reports would accurately estimate the true energy costs of the completed military activities, while the Compendium of Physical Activities was expected to underestimate the energy costs of our military tasks. It was assumed that unless specifically stated, the energy costs published in the Compendium expressed values measured in individuals wearing significantly lighter clothing than would be typical of military personnel. As expected, military reports more accurately estimated the energy costs of our activities especially when clothing weight was within 4kg and walking speed was within 0.5km/h. The Compendium on the other hand, significantly underestimated the energy cost of most military tasks. We propose that this underestimation was due to a combination of three factors: 1) the limited number of military- specific tasks that can be found in the Compendium; 2) the vaguely described Compendium activities which lead to a greater reliance on subjective interpretations of activities and consequently poorer activity-matching (matching completed tasks to Compendium activities); 3) and the fact that military personnel wear heavier clothing than the general population; since the vast majority of Compendium activities provide no information in regards to what clothing was worn during activity completion it is likely that individuals wore typical civilian clothing (which usually weighs <2kg [247]) while completing activities.

With the additional knowledge and planned publication of the 46 well described infantry tasks in the current investigation, EE estimations within relevant military populations can be made more confidently via the factorial method. Our activity descriptions provide a substantial level of detail including information regarding clothing weight, equipment weight, travelling speed, and terrain grade where applicable.

9.2.2 Ambient Temperature

While the fraction of EE that thermoregulation constitutes can generally be ignored under thermoneutral conditions, its contribution increases as the ambient temperature rises [51] or falls[50]. It seems prudent to include thermoregulatory thermogenesis when EE estimations are required in more extreme thermal environments. During exercise in the heat when the level of heat stress is great enough to bring about increases in core temperature, and thermoregulatory processes such as sweating and skin blood flow above that seen during thermoneutral conditions, EE is also reported to be higher [51, 69, 72]. This increase in EE is often attributed

191

to the energy cost of heat dissipation [51]. Similarly during exercise in the cold, EE is elevated when the exercise intensity is unable to maintain the skin and core temperatures above the shivering threshold, [79] thereby increasing the contribution of thermoregulation. Alternatively if the exercise intensity is sufficient to maintain body temperature, shivering will be prevented and EE will mimic values more commonly seen in temperate environments [79, 80].

Both of these scenarios ignore behavioural strategies that individuals use to aid in thermoregulation such as: altering the amount of clothing worn [248], self-selecting lower work- rates in the heat [249, 250], or altering feeding behaviours (e.g. reaching for cool beverages in the heat, or hot meals in the cold) [251]. In many infantry scenarios however, the number of strategies and their thermoregulatory effects are limited. Military personnel cannot retreat from the elements by going indoors, they are limited in their clothing choices since certain clothing items (tactical vest, fragmentation vest, combat boots, and helmet) are mandatory for safety reasons and therefore cannot be removed, and task modifications at an individual level are also largely limited as infantry are required to follow commands. However military personnel can still add and remove discretionary clothing, alter their rate of work to some extent, and adjust their food and water intake to alleviate the thermal strain experienced in more extreme environments. Furthermore the behavioural strategies in the cold, specifically the layering of additional clothing, while decreasing the likelihood of shivering and non-shivering thermogenesis, can significantly increase clothing weight [82] and decrease the efficiency of moving while wearing that clothing [66] resultantly increasing overall EE.

Taken together it was unclear whether the thermoregulatory behaviours would mitigate the effects of ambient temperature on EE or whether a systematic increase in EE would be evident in harsher thermal climates prompting the use of a correction factor when EE estimations are done for hot and cold environments. It was hypothesized that the true impact of ambient temperature on EE during a simulated military exercise would be minimal. In the Hot environment, participants were expected to shed excess clothing and self-select lower work rates to negate the increase in EE, whereas in the Cold environment participants were expected to add layers of clothing to protect themselves from the cold. As a result, it was hypothesized that EE measured during the Cold conditions might be significantly higher due to clothing weight, but the difference (if there was one) was not expected to be large enough to be meaningful.

192

By altering the ambient temperature and allowing participants to respond behaviourally as they would in the field (by adding and taking off layers of discretionary clothing at will, completing half of the military tasks at their own pace, and by consuming food and water ad libitum) the second study (The effect of ambient temperature on the energy cost of infantry activities) was able to record a more realistic effect of ambient temperature on EE. Taken together the results of our research indicates that EE is marginally increased (~3%) when personnel are completing military tasks in hot or cold environments vs. temperate.

In the current study behavioural factors could only go so far to mitigate the thermal strain experienced in the Hot condition. While participants wore the bare minimum in the Hot condition (which remained very high (13.2kg) due to the mandatory clothing items), self- selected lower work-rates (completing 5% fewer repetitions and travelling ~5% slower) during the self-paced tasks, and some even altered their eating behaviours ( 2 out of 18 (11%) participants in the Hot condition stated that they consumed food items in order to cool down) (Table X)), thermal comfort ratings, heart rate, Borg scale ratings, and core temperature were still highest in the Hot condition as compared to the other conditions.

Certain thermoregulatory behaviours were more successful than others, when participants were permitted to self-select their work rate any EE increase that was associated with the heat was eliminated, in contrast other thermoregulatory behaviours (shedding clothing, or adjusting feeding behaviours) were unable to blunt the thermal effect on EE as forced-pace (treadmill) activities had significantly higher EE in the Hot vs. Temperate environment.

Similarly in the cold, participants wore significantly more clothing (wearing on average 1.2kg of more clothing during the Cold vs. Temperate trials) and felt that the ambient temperature affected their eating behaviour with 7 out of 18 (39%) participants stating that they consumed more food in the cold in order to help them stay warm (Table X). Whereas in the heat thermoregulatory behaviours were able to mitigate the increase of EE in certain cases, in the cold these behaviours (namely the increase in clothing weight) were associated with marginal increases in EE. While the energy cost of infantry tasks was marginally higher in the cold, when corrected for clothing weight the EE was not different between Cold and Temperate conditions

Overall, we found that the energy costs of the military activities increased by about 10kcal∙h-1 (~3%) in the Hot (30°C), and Cold (-10°C) environments, as compared to the

193

Temperate (21°C) environment. This increase was small and disappeared when forced-paced activities were excluded suggesting that the added energy cost of the thermal environment only becomes a factor when an activity rate is imposed.

9.2.3 A Military-Specific Algorithm

The use of the factorial method in the military population is substantially improved by having a well described catalogue of infantry tasks and their associated energy costs and by recognizing that ambient temperatures in the vicinity of -10°C and 30°C are likely to increase overall EE by ~3% (Table 34).

Table 34 :Measured vs. Estimated 4-h Energy Expenditure Corrected for Environmental Condition Measured EE Estimated EE (kcal) Estimated EE Estimated EE (kcal) (kcal) over 4- over 4-h using (kcal) using the Condition over 4-h using data hA previously available military-specific from Study 1C sourcesB algorithm Sedentary 405±67 436±87 (8%) 436±87ǂ (8%) 399±87 (1%) Temperate 1485±302 1308±264 (12%) 1513±305 (2%) 1525±336 (3%) Cold 1528±296 ~1347§ (12%) ~1559§ (2%) 1485±322 (3%) Hot 1538±263 ~1347§ (12%) ~1559§ (1%) 1664±449 (8%) A -1 -1 EE was determined for 17 participants by taking 1 min averages of breath by breath V̇ O2 (ml · min · kg ), dividing V̇ O2 by 3.5 to get METs and then converting METs to EE in kcal: (METs*3.5*body mass (kg))/200. All data for each participant was summed and divided by time (in min) to get EE per minute. Due to missing data, this value (EE per min) was multiplied by 240 in order to get EE over 4 hours B Completed activities were matched with activities found in the Compendium of Physical Activities [105] as well as relevant military reports [245, 246]. Where reasonable matches could not be made, completed activities were paired with published activities that were regarded to be similar in intensity. C Average METs displayed in Study 1 were inserted for each activity. For the activities missing from the compilation reported in Study 1 (those that were too short in duration to determine an average MET value), a reference value of 10.0 METs was inserted. All recovery time was treated as the same as “sit in one spot” consequently the reference value of 1.53 METs was used. ǂ To estimate Sedentary 4-h EE the activity “Sitting quietly“(1.3 METs) from the Compendium of Physical Activities [105] was used as “sit in one spot” was not deemed appropriate due to the vastly different clothing that was warn between the Sedentary and Temperate conditions. § To estimate 4-h EE in the Hot and Cold conditions, 4-h EE estimations from the Temperate condition were multiplied by 1.03 to reflect the 3% increase that is expected in such an environment. Data displayed are mean±SD with the value in brackets representing the percent difference from the measured value.

Nevertheless the factorial method is still susceptible to large estimation errors if activity documentation is inaccurate. Even with accurate recording the subjective matching of completed activities with published activities can increase error.

The development of a military specific algorithm on the other hand removes the necessity of subjective documentation altogether. Despite the fact that an algorithm cannot initially be

194

used to predict the energy cost of upcoming field operations, the continued use of such a tool can amass data over time and provide reliable data on which to base food provision decisions on.

The algorithm we developed can estimate group-level EE within 15% of measured EE in thermal environments ranging between -10°C and 30°C, when dressed appropriately for the ambient temperature. The algorithm is easy to use and requires only two measurements (body mass (kg), maximal heart rate (bpm)) on top of the output from the triaxial accelerometer and the heart rate strap.

9.3 An Investigation into the Under Consumption of Military Rations

Unfortunately providing military personnel with the optimal caloric content will not prevent weight loss if the food that is provided is not eaten. Even when calorically sufficient military rations are provided, military personnel consistently consumed an inadequate number of calories during field operations and exercises (Figure 7). As a result, Study 4 (The effects of exercise and ambient temperature on appetite and energy intake) investigated the role that appetite (both subjective and hormonally implied) plays in EI over 24-h in this population.

9.3.1 The Impact of Physical Activity and Ambient Temperature on Energy Intake

The data are pretty clear regarding exercise bouts lasting < 2 h, and their inability to impact absolute EI (Figure 9) in any significant way. However since previous researchers found increases in EI with 120 min of exercise, resulting in partial [186] and complete [124] compensation for the amount of energy expended, it was hypothesized that absolute EI would be higher in the Hot, Temperate, and Cold conditions (considering that 240 min of arduous physical activity was completed) vs. the Sedentary condition. Furthermore, due to previous reports of higher absolute EI in cold environments [229], it was also hypothesized that absolute EI would also be higher in the Cold as compared to the Hot or Temperate conditions. Contrary to both of our initial hypotheses, absolute EI was not different between the four conditions and consequently participants were in a negative energy balance during all of the active conditions (Hot, Temperate, and Cold).

While previous research suggests that the food provided and/or the inconvenience of eating in field scenarios [13-15] underlies the reason behind the under consumption of military rations, our research does not support these claims. In terms of taste and overall acceptability of 195

the military rations provided, participants found the IMPs to be largely acceptable (Appendix U: Figure 38). Some participants (4 of 18 (22%) in the Hot and Temperate and 5 of 18 (28%) in the Cold and Sedentary condition) stated that they thought they ate less because they “didn’t like the taste of the food that was provided” (Appendix T: Table X), with one of those participants further clarifying that the food “didn't taste good. Too salty too much sodium. Too much sugar” (Appendix T: Table X). When only considering the participants who stated that they ate less because they didn’t like the food that was provided, 50% actually ate more than their average usual EI ±10% (average usual EI was determined on visit 3 - when participants consumed their own food at home), 17% consumed within 10% of their average usual EI, and only 33% accurately judged that they did in fact consume less that their average usual EI±10%. Since EI was not different regardless of whether participants ate their own food at home or military rations (Appendix V: Figure 39), the palatability of the military rations is likely not the reason for the under consumption of IMPs in our study [252].

In addition, participants were given almost 4-h on each chamber day during which they were not wearing the metabolic measurement system and were not required to complete tasks. Although we assumed that this would give participants ample time to eat, one participant felt that they did not have enough time to eat during any of the active (Hot, Temperate, and Cold) trials, while an additional participant only felt that they did not have enough time during the Cold trial (Appendix T: Table X). Fortunately the vast majority of participants did not feel rushed during any of the trials. Lastly, our research staff prepared all hot meals on request (since the flameless ration heaters could not be used inside the environmental chamber), ultimately removing the burden of food preparation from the participants. Thereby the usual suspects, the palatability of military food, insufficient time to eat, and inconvenient or lengthy food preparation [13-15] could not be responsible for the a negative energy balance that was reported during the Hot, Temperate, and Cold trials.

9.3.2 The Impact of Physical Activity and Ambient Temperature on Subjective Appetite

One factor that can be considered is subjective appetite, perhaps the participants did not feel hungrier in the active conditions, and therefore did not eat more. Previous research suggests that subjective appetite is highest in the cold and lowest in the heat [226]. As a result it was hypothesized that subjective appetite in Study 4 would be highest in the Cold trial and lowest in the Hot trial and as anticipated, our results were in accordance with our hypothesis. Participants

196

were significantly hungrier and they felt they could eat more during the Cold condition as compared to the Hot condition, and they felt significantly fuller and more satisfied during the Hot condition than they did during the Cold condition (Appendix W: Figure 40). Interestingly, despite all indices of subjective appetite being significantly correlated with the amount of energy consumed 30 min following appetite sensation assessment (a finding that is not always observed [131, 132]); this did not translate to a great caloric intake in the cold, or a reduced energy intake in the heat.

9.3.3 The Impact of Physical Activity and Ambient Temperature on Appetite Regulating Hormones

Various appetite regulating hormones can also potentially impact EI. While there isn’t a consensus on the matter, most previous studies report that exercise transiently decreases plasma acylated ghrelin levels, and increases plasma GLP-1, and PYY levels [203]. When exercise is completed in different thermal environments, plasma leptin concentrations seem to be unaffected, while the effects on plasma ghrelin and PYY concentrations are equivocal. Considering that plasma leptin concentrations has previously been found to decrease following arduous physical activity [212], it was hypothesized that the military activities will lead to lower leptin concentrations during the active conditions vs. Sedentary. Similarly lower acylated ghrelin concentrations and higher GLP-1, and PYY concentrations were expected during the active conditions vs. Sedentary.

Contrary to our hypotheses, the different conditions had no effect on plasma GLP-1 or leptin levels (Appendix X: Figure 41). Plasma PYY concentrations on the other hand were elevated during the Temperate and Cold conditions as compared to the Sedentary condition, suggesting that NPY/AgRP expressing neurons were also more inhibited, which should theoretically inhibit appetite in the Temperate and Cold conditions as compared to Sedentary (Figure 37). While PYY was correlated with every index of appetite sensation and was negatively correlated with the amount of food (kcal) consumed in the 30 min following each sample collection, overall subjective appetite, and EI were not supressed in the Temperate and Cold conditions as compared to Sedentary. PYY may have played a role in the initiation of food intake, rather than meal size.

Acylated ghrelin concentrations were highest in the Sedentary condition followed by the Cold condition, and lowest in the Temperate and Hot conditions (Appendix X: Figure 41). The

197

low acylated ghrelin concentrations in all of the active conditions (Hot, Temperate, and Cold) would lead to less stimulation of the NPY/AgRP expressing neurons were which should also theoretically decrease appetite and EI (Figure 37). Acylated ghrelin concentrations were negatively correlated with fullness and satisfaction scores but were not correlated with the amount of food (kcal) consumed within 30 min of sample collection.

Interestingly neither hormone capable of directly stimulating POMC/CART expressing neurons was significantly impacted by either physical activity or ambient temperature. Although unlikely, it is possible that EI was unchanged because GLP-1 was unchanged.

198

Figure 37: Expected hormonal effects in Study 4. . Glucagon-like peptide-1 (GLP-1) and leptin were unaffected by exercise or ambient temperature in our study and can therefore not be responsible for our results. Plasma peptide YY (PYY) concentrations were elevated during the Temperate and Cold conditions as compared to the Sedentary condition, suggesting that Neuropeptide Y (NPY) and Agouti-related protein (AgRP) neurons (signified as the brown cluster) were more inhibited, in the Temperate and Cold conditions and appetite should have been suppressed. Acylated ghrelin concentrations were found to be Sedentary>Cold>Temperate=Hot. The low acylated ghrelin concentrations in all of the active conditions would lead to less stimulation of the NPY/AgRP expressing neurons were which should also theoretically lead to a suppress appetite. Pro- opiomelanocortin (POMC), and Cocaine- and amphetamine-related transcript (CART) neurons (signified as the orange cluster) did not seem to be affected differently in the different conditions. GHS= Growth hormone secretagogue receptor; Y2R=PYY Y2 receptor; LEPR=Leptin receptor; GLP1R= GLP-1 receptor

EI was not affected by differing perceptions of appetite, or varying concentrations of appetite hormones. This suggests that over the short term, other influences or cognitive factors including “conditioned satiety” [253] (the learned association between a familiar food and its satiating effect which allows an individual to anticipate how much of that food they should eat to feel satiated) are likely to play a larger role in EI than either appetite sensation, or hormonal concentrations do. It is unclear how the different contributors (appetite sensations, hormones,

199

learned associations) impact energy intake, and when these factors are more or less prominent in guiding our eating behaviours.

9.4 Limitations

Since our sample was a convenience sample of 25 CAF members, it should not be considered as representative of the CAF population, or even the infantry population within the CAF. Women represented 24% of our sample, a larger percentage than the 15% reported in the CAF [254]. When compared to the average CAF male (age: 33.4 years, height: 1.77 m, and weight: 88 kg) [45], our male participants (age: 31.5 years, height:1.76 m, and weight: 81 kg) were around the same age and height as the average CAF male, but they were 7 kg lighter. When compared to the average CAF female (age: 34.7 years, height: 1.64 m, and weight: 69 kg) [45], our female participants (age: 35.3 years, height: 1.60 m, and weight: 71 kg) were around the same age and weight, but were 4 cm shorter than the average CAF female. The larger proportion of females in our sample, the minor disparities in anthropometric data, or some other undefined characteristics may have affected our data in an unknown way.

Another limitation is the use of the reference value 3.5 ml·kg·min−1 as the estimate of BMR or 1 MET, that we chose to use in our investigations, instead of expressing 1 MET as each participant’s actual RMR. Using reference values as a substitute to measurements is likely to introduce errors, and this reference value is no exception as it has previously been found to overestimate BMR in heavier, older, and less fit individuals and subsequently to underestimate the energy cost in these populations [255]. Nevertheless the decision to use calculated vs. ratio (Activity EE/BMR) METs in this dissertation was made for two main reasons: 1) in terms of our “compilation of infantry tasks” we wanted our values to be compatible with the Compendium of Physical Activities [105]; and 2) we acknowledge the likelihood that the reference value will be used in the majority of cases where EE estimation is conducted in military personnel, due to the cost and inconvenience of BMR measurements. Although we anticipated the use of the reference value, BMR measurements should have been done in order to determine whether the 3.5 ml·kg·min−1reference value was appropriate in our sample. Without such data, we cannot conclusively say whether the reference value systematically over or underestimates BMR in our sample.

While we can confidently conclude that the energy cost of completing a series of infantry tasks was only minimally increased in our Hot (30°C) and Cold (-10°C) environments, resting 200

EE between the activity circuits was only measured for the first 15 min; as a result, the impact of resting for 2-h in the heat or cold was not clearly characterized. Considering that the heat participants were exposed to during the Hot trial was compensable, and participants reported feeling “fairly comfortable” and their core temperatures decreased during rest, it is unlikely that EE would have been significantly elevated above levels seen during the Temperate condition. In the Cold condition on the other hand, participants felt coldest during rest and core temperatures also dropped during this time. Since feeling cold, can increase EE by ~7% due to non-shivering thermogenesis [77], it is very possible that EE during the rest periods was higher during the Cold vs. Temperate conditions. We did not measure EE during rest and therefore cannot characterize the effect of ambient temperature on resting EE in military personnel in these Hot (30°C) and Cold (-10°C) environments.

It became clear that some participants were ill-prepared for the ambient temperature they endured and either over or under-dressed for it. In the cold in particular, this resulted in many participants over-dressing during the initial 2-h (during the first activity block), which led to an increase in body temperature and many started sweating and their clothing would become damp and cold. Heat loss is significantly accelerated in wet vs. dry clothing [256]. Although some participants changed into dry clothing during the rest periods, many preferred to stay in their sweat soaked clothing and simply covered themselves with blankets and/or a parka. During actual cold-weather field trials, commanders will often recommend what level of clothing is appropriate and in extreme cold trials (such as those completed in the arctic) tasks are modified to prevent sweating whenever possible [257]. Even though participants were advised that they would be exercising during the first two hours and that they may benefit from dressing more lightly, no suggestions were made to participants in regard to what they should wear. As a result, the impact of the Cold trial on some of our participants could have potentially been reduced with the proper guidance, and thus may not be representative of typical cold-weather field trials.

We also experienced a host of complications which resulted in the loss of ~10 min of activity data due to the strenuous nature of our protocol and our equipment’s aversion to the cold. While participants were at times too exhausted to complete an activity, the vast majority of missing data were missed because of the various problems that occurred with the portable oxygen uptake system (the sample line would freeze during the Cold condition, the pump failed, and particulate in the air would block the sample line occasionally). Where appropriate, data

201

have been analyzed “per task” in order to eliminate the impact the missing data could have, however this was not possible in all circumstances and in such cases the potential for the missing data to confound data interpretation has been clearly expressed.

Many aspects of the chamber trial adequately simulated field conditions such as the ambient temperature experienced, the clothing worn, and the tasks completed. However, there were aspects of the simulated environment which differed significantly from important field conditions; one example was that the military rations were heated for participants on request because the participants could not be permitted to do so inside of the chamber. In addition, various potentially stressful aspects of real field operations that may have influenced energy expenditure could not be simulated in the current investigations (e.g. fear, stress of interpersonal conflicts and relationships during intense operations, being away from family, etc.[258]). Another important limitation of the current investigation was the one-day duration of augmented physical activity. There are obvious risks of extrapolating the results of the current investigation to actual field scenarios lasting days and weeks where adaptive responses such as habituation and acclimatization could modify the acute responses described in the current research. In addition, the hormonal effects of prolonged undernutrition (occurring over weeks and months) is likely to be different to the short term (over several hours) hormonal effects described within this work. While these results describe the short-term hormonal effects clearly, the extrapolation of this work to longer term deployments should be greatly cautioned.

Overall our protocol should not be classified as being very stressful considering the excretion of cortisol during the chamber days was similar to the non-chamber days (Appendix Y: Table XIII). Increases in circulating cortisol concentrations as a result of field operations have previously been reported [259, 260]. Although plasma or serum concentrations have typically been collected in military personnel, cortisol concentrations in urine, saliva and blood have been found to be highly correlated [261], suggesting that our lack of effect was an indication that our simulation was less stressful than some of the military field exercises that military engage in. Similarly, while the Hot and Cold trials were found to be fatiguing (Appendix Z: Table XIV), our trials did not negatively impact the participants psychological state (Appendix Z).

202

9.5 Recommendations for Future Research

More activities should be added to the catalogue of infantry tasks

While the infantry tasks that were included in Study 1 are representative they should not be considered comprehensive. The EE associated with several characteristic activities, especially those that are specific to the cold (pulling sleds, snowmobiling etc.) were not assessed. Similarly travelling over different terrains affects EE, and the energy cost of walking through snow, significantly affect EE [83]. Although this catalogue of EE during a variety of infantry tasks is an important addition to existing knowledge, it could and should be expanded to include a much more comprehensive assessment of standard physical activities that are frequently carried out by military personnel in the field.

More extreme thermal environments should be tested

The energy cost of completing a series of infantry tasks was only minimally increased in our hot (30°C) and cold (-10°C) environments. While our results are probably generalizable to less severe heat and cold stress conditions, the effects of more extreme environments should be investigated.

EE measurements during periods of rest in the field should be performed

While the impact of ambient temperature on physical activity EE has been determined, the impact of harsh (Hot: 30°C and Cold: -10°C) environments on sedentary and resting EE should be assessed. These measurements should ideally be made in the field as participants may prepare for an actual field trial differently than they would for a simulation. During field exercises participants may be instructed to change out of wet clothing or be required to wear specific clothing in certain environments which could greatly impact EE.

The military-specific algorithm needs to be validated in the field

While our military-specific algorithm performed well on a different pool of participants engaged in the same activities in the same environments, our algorithm should be validated in a prospectively designed study with a different cadre of participants engaged in field training scenarios.

203

Longer duration trials need to be completed

Study 4 clearly demonstrates that neither physical activity nor ambient temperature impact energy intake within 24 hours, however the impact over a week or a month is unknown. While the data appear to be clear in regard to military personnel under-eating during the vast majority of documented field operations, their eating behaviours, their perceptions of appetite, and/or the hormonal data may change over time. Similarly, adaptive responses such as habituation and acclimatization could modify the acute responses described in the current research. Both longer exposure times and longer assessment periods post-exposure are necessary in order to truly assess the impact of activity and ambient temperature on EE and appetite.

The effects of frequent “easting reminders” on EI and weight loss should be assessed

Our participants were in a negative energy balance during all of the active trials (Hot, Temperate, and Cold) despite having ample time to eat, finding the food to be “acceptable”, and regardless of the variations in hormone concentrations or appetite sensations. It is possible that military personnel incorrectly assume that they are eating enough, or that various learned eating behaviours impact their short-term EI more than physiological variables. If so, establishing frequent eating reminders to encourage ration consumption may effectively limit under consumption; the effects of such a change in standard operating procedures in the field should be investigated.

204

9.6 Significance and Practical Applications

Study 1 – The energy cost of various infantry tasks

This study has accurately measured the energy costs of 46 infantry tasks in military personnel wearing typical military clothing. Previously, many military tasks had unknown energy costs, and similar tasks completed by civilians generally underestimated the actual cost of performing the same activity while wearing ~13kg of clothing. This compilation of tasks will improve the EE estimations garnered through the use of questionnaires and the factorial method.

Study 2 – The effect of ambient temperature on the energy cost of infantry activities

This study is the first of which we are aware to assess the energy cost of the same tasks in three different environments while allowing participants to respond behaviourally as they would in the field (by adding and taking off layers of discretionary clothing at will, completing half of the military tasks at their own pace, and by consuming food and water ad libitum). This allowed us to record a more realistic effect of ambient temperature on EE. Our research indicates that EE is marginally increased (~3%) when personnel are completing military tasks in hot or cold environments vs. temperate. This indicates that caloric supplementation on account of ambient temperature is likely unnecessary during short term operations occurring within ambient temperatures of -10°C and 30°C.

Study 3 – Estimating energy expenditure in military personnel using accelerometry and heart rate

This study generated an easy to use military-specific algorithm that can estimate group EE within 15% of measured EE in ambient temperatures ranging between -10°C and 30°C. To the best of our knowledge, no other algorithm has been tested under various thermal environments or has encompassed a larger number of activities. This tool will facilitate continuing studies of EE data during military operations to further improve military feeding strategies.

Study 4 – The effects of exercise and ambient temperature on appetite and energy intake

This was the first study to assess energy intake, perceptions of appetite, and appetite regulating hormones in the same participants in Hot, Temperate and Cold conditions. It was also the first study to assess the impact of exercise in different ambient temperatures on circulating GLP-1 levels, and the first to employ 4-h of arduous exercise - thereby testing the effects of,

205

arguably one of the most intense exercise stimuli documented to date on appetite. Finally, this is the only appetite study we are aware of that has assessed the impact of exercise on appetite without imposing prescribed and strictly controlled meal times (which can artificially delay or advance meal timing and ultimately impact the amount of food consumed [124]).

We determined that ambient temperature influences subjective appetite responses as well as appetite regulating hormone concentrations. However absolute energy intake nevertheless remains unchanged. Contrary to previous studies where increases in absolute EI were found with 120 min of exercise, resulting in reports of partial [186] and complete [124] compensation for the amount of energy expended, 240 min of exercise in the current study had no effect on absolute EI in our study. Similarly while others reported an increase in EI following exercise in cold vs. temperate conditions [229], the ambient temperature had no effect on the EI in our study. In this study where food was freely available, variations in ambient temperature, exercise vs. rest, appetite hormone concentrations, and subjective appetite sensation had no effect on dietary intake within 24 h of acute, prolonged exercise. While it is still not clear why large increases in EE do not augment EI, voluntary anorexia does not seem to be caused by unpalatable food, lack of time, or lengthy food preparation. Consequently, research should focus on finding ways to shift eating habits during periods of high EE in an effort to limit voluntary anorexia.

206

9.7 Conclusions

In conclusion, this research will improve the estimation of EE in military personnel in three ways: 1) by providing an empirical basis of knowledge of the energy costs of a range of infantry tasks, which allows for anticipatory EE estimations of military field trials and/or the energy costs of previous field trials to be estimated with greater accuracy; 2) by determining that the added energy cost of conducting military operations in hot (30°C) and cold (-10°C) environments vs. temperate (21°C) is minimal (3%), thereby permitting the use of the “compilation of infantry tasks” in field trials conducted in harsh climates with relative confidence; and 3) by providing an easy to use military-specific algorithm for the continued estimation of EE in infantry field operations. With these improvements, the energy requirements of different military operations will become more accurately assessable and consequently the logistical requirements associated with the supply of food to fuel e military field operations can be more confidently determined.

In addition, this dissertation research demonstrated that even in the most favourable scenarios (where hot pre-prepared food is provided to participants on request) military personnel engaged in physically demanding field operations will likely not consume enough food and plunge into a negative energy balance. Based on the results of the current investigations, it does not seem that the frequently identified mediators of voluntary anorexia in military personnel during field operations (provided food and time) are involved in this phenomenon. These results suggest that policy changes (such as establishing a field feeding doctrine to encourage food intake, scheduling eating reminders at set intervals) may be more effective in addressing the under consumption of military field rations than would be expending resources on palatability, or food preparation technologies.

207

References 1. National Institutes of Health (U.S.). Estimates of Funding for Various Research, Condition, and Disease Categories (RCDC), 2017. 2. Mudambo, K.S., C.M. Scrimgeour, and M.J. Rennie, Adequacy of food rations in soldiers during exercise in hot, day-time conditions assessed by doubly labelled water and energy balance methods. Eur J Appl Physiol Occup Physiol, 1997. 76(4): p. 346-51. 3. Fairbrother, B., et al., Nutritional and Immunological Assessment of Soldiers During the Special Forces Assessment and Selection Course. 1995, DTIC Document. 4. Tharion, W.J., et al., Energy requirements of military personnel. Appetite, 2005. 44(1): p. 47-65. 5. Jacobs, I., et al., Physical performance and carbohydrate consumption in CF commandos during a 5-day field trial. 1989, DTIC Document. 6. Edwards, J.S. and D.E. Roberts, The influence of a calorie supplement on the consumption of the meal, ready-to-eat in a cold environment. Mil Med, 1991. 156(9): p. 466-71. 7. Margolis, L.M., et al., Energy balance and body composition during US Army special forces training. Appl Physiol Nutr Metab, 2013. 38(999): p. 396-400. 8. Tharion, W.J., et al., The Effects of 60 Days of Tray Ration Consumption in Marine Combat Engineers While Deployed on Great Inagua Island, Bahamas. 2000, DTIC Document. 9. Nindl, B.C., et al., Physiological consequences of US Army Ranger training. Medicine & Science in Sports & Exercise, 2007. 39(8): p. 1380-1387. 10. Moore, R.J., et al., Changes in soldier nutritional status and immune function during the Ranger training course. 1992, DTIC Document. 11. Askew, E.W., et al., Nutrient Intakes and Work Performance of Soldiers during Seven Days of Exercise at 7,200 Feet Ready-to-Eat Ration. 1986, U.S. Army Research Institute of Environmental Medicine: Natick, MA. p. 65. 12. Askew, E.W., et al., Mauna Kea III: Metabolic effects of dietary carbohydrate supplementation during exercise at 4100 M altitude. 1987, USARIEM. 13. Baker-Fulco, C.J., et al., Nutrition for health and performance, 2001: nutritional guidance for military operations in temperate and extreme environments:. 2001: USARIEM. 14. Friedl, K.E. and R.W. Hoyt, Development and biomedical testing of military operational rations. Annual review of nutrition, 1997. 17(1): p. 51-75. 15. Marriott, B.M., Not eating enough: Overcoming underconsumption of military operational rations. 1995: National Academies Press. 16. Whitlock, G., et al., Body-mass index and cause-specific mortality in 900 000 adults: collaborative analyses of 57 prospective studies. Lancet, 2009. 373(9669): p. 1083-96. 17. Nikolaidis, P.T. and D. Zisimatos, Relationship of body mass index with 1,600 m running, 50 m swimming, and pull-ups performance in army cadets. Saudi Journal of Sports Medicine, 2014. 14(2): p. 144. 18. Mak, K.-K., et al., Health-related physical fitness and weight status in Hong Kong adolescents. BMC public health, 2010. 10(1): p. 88. 19. Goyal, A., K.R. Nimmakayala, and J. Zonszein, Is there a paradox in obesity? Cardiology in review, 2014. 22(4): p. 163. 20. McCrory, M.A., V.M. Suen, and S.B. Roberts, Biobehavioral influences on energy intake and adult weight gain. The Journal of nutrition, 2002. 132(12): p. 3830S-3834S. 21. Harvie, M.N., et al., The effects of intermittent or continuous energy restriction on weight loss and metabolic disease risk markers: a randomized trial in young overweight women. International journal of obesity, 2011. 35(5): p. 714-727. 208

22. Wycherley, T.P., et al., Effects of energy-restricted high-protein, low-fat compared with standard-protein, low-fat diets: a meta-analysis of randomized controlled trials. The American journal of clinical nutrition, 2012: p. ajcn. 044321. 23. Garrow, J. and C. Summerbell, Meta-analysis: effect of exercise, with or without dieting, on the body composition of overweight subjects. European journal of clinical nutrition, 1995. 49(1): p. 1-10. 24. Catenacci, V.A. and H.R. Wyatt, The role of physical activity in producing and maintaining weight loss. Nature Clinical Practice Endocrinology & Metabolism, 2007. 3(7): p. 518-529. 25. Ravussin, E., et al., Reduced rate of energy expenditure as a risk factor for body-weight gain. New England Journal of Medicine, 1988. 318(8): p. 467-472. 26. Consolazio, C.F., et al., The relationship of diet to the performance of the combat soldier. Minimal calorie intake during combat patrols in a hot humid environment (Panama). 1979, Letterman Army Institute of Research. 27. Jones, P.J., et al., Adequacy of food rations in soldiers during an arctic exercise measured by doubly labeled water. J Appl Physiol, 1993. 75(4): p. 1790-7. 28. Crowdy, J., M. Haisman, and H. McGavock, Combat Nutrition: The effects of a restricted diet on the performance of hard and prolonged physical work. APRE Report, 1971. 2: p. 71. 29. Montain, S.J. and A.J. Young, Diet and physical performance. Appetite, 2003. 40(3): p. 255-67. 30. Keys, A., Caloric undernutrition and starvation, with notes on protein deficiency. J Am Med Assoc, 1948. 138(7): p. 500-11. 31. Shippee, R., et al., Nutritional and Immunological Assessment of Ranger Students with Increased Caloric Intake. 1994, DTIC Document. 32. Keys, A., et al., The biology of human starvation.(2 vols). 1950. 33. Johnson, M.J., et al., Loss of muscle mass is poorly reflected in grip strength performance in healthy young men. Med Sci Sports Exerc, 1994. 26(2): p. 235-40. 34. Tharion, W.J., et al., Nutritional challenges for field feeding in a desert environment: use of the UGR and a supplemental carbohydrate beverage. 1997, USARIEM: Natick, MA. p. 190. 35. DeLany, J.P., et al., Field use of D2 18O to measure energy expenditure of soldiers at different energy intakes. J Appl Physiol, 1989. 67(5): p. 1922-1929. 36. Jacobs, I., Field Feeding and Energy Expenditure Research. Final Contract Report, 2015. 37. Fallowfield, J.L., et al., Energy expenditure, nutritional status, body composition and physical fitness of Royal Marines during a 6-month operational deployment in Afghanistan. British Journal of Nutrition, 2014. 112(5): p. 821-829. 38. NATO. Nutrition Science and Food Standards for Military Operations. 2010: Neuilly-sur- Seine, France. 39. Hatton, P., Food and beverage consumption of Canadian Forces soldiers in an operational setting: Is their nutrient intake adequate? 2005: McGill University (Canada). 40. Department of National Defence. Food Services Manual. 2009: Ottawa, Ontario. 41. Institute of Medicine (U.S.). Dietary Reference Intakes for Energy, Carbohydrate, Fiber, Fat, Fatty Acids, Cholesterol, Protein, and Amino Acids (Macronutrients). 2005, Washington, D.C.: The National Academies Press. 42. Institute of Medicine (U.S.). Panel on Macronutrients. Dietary reference intakes for energy, carbohydrate, fiber, fat, fatty acids, cholesterol, protein, and amino acids. 2005: National Academy Press.

209

43. Baker-Fulco, C.J., et al., Military Dietary Reference Intakes: Rationale for Tabled Values. 2001, USAREIM, Natick, MA. 44. Regulation, A., Nutrition and Menu Standards for Human Performance Optimization, in Nutrition Standards and Education. 2017, Departments of the Army, the Navy, and the Air Force: Washington, DC. 45. Keefe, A.A. and H. Angel, 2012 Canadian Forces Anthropometric Survey (CFAS) 2015, Defence Research and Development Canada. 46. Robert-McComb, J.J., Estimating Energy Requirements. The Active Female, 2008: p. 279- 286. 47. Kleiber, M., Body size and metabolic rate. Physiological reviews, 1947. 27(4): p. 511-541. 48. Weinsier, R.L., Y. Schutz, and D. Bracco, Reexamination of the relationship of resting metabolic rate to fat-free mass and to the metabolically active components of fat-free mass in humans. The American journal of clinical nutrition, 1992. 55(4): p. 790-794. 49. Westerterp, K.R., Diet induced thermogenesis. Nutrition & metabolism, 2004. 1(1): p. 5. 50. Jacobs, I., L. Martineau, and A.L. Vallerand, Thermoregulatory thermogenesis in humans during cold stress. Exerc Sport Sci Rev, 1994. 22: p. 221-50. 51. Consolazio, C.F., et al., Energy Requirements of Men in Extreme Heat. The Journal of Nutrition, 1961. 73(2): p. 126-134. 52. Maffels, C., et al., Energy expenditure during walking and running in obese and nonobese prepubertal children. The Journal of pediatrics, 1993. 123(2): p. 193-199. 53. Miller, A. and C.S. Blyth, Influence of body type and body fat content on the metabolic cost of work. Journal of applied physiology, 1955. 8(2): p. 139-141. 54. Bunc, V., Energy cost of treadmill running in non-trained females differing in body fat. Journal of sports medicine and physical fitness, 2000. 40(4): p. 290. 55. Friedl, K.E. and J.W. Ness, Upper Limits of Energy Expenditure for Men and Women in Military Operational Settings: Does Female Fat Metabolism Provide a Performance Advantage?, in NATO HFM symposium - Impacts of Gender Differences on Conducting Operational Activities 2008: Antalya, Turkey. 56. Delaney, J.P., Determination of Total Daily Energy Requirements and Activity Patterns of Service Women. 2001, DTIC Document. 57. Hoyt, R.W., et al., Negative energy balance in male and female rangers: effects of 7 d of sustained exercise and food deprivation. Am J Clin Nutr, 2006. 83(5): p. 1068-75. 58. Tharion, W.J., et al., Adequacy of garrison feeding for Special Forces soldiers during training. Mil Med, 2004. 169(6): p. 483-490. 59. Margolis, L.M., et al., Energy requirements of US army special operation forces during military training. Nutrients, 2014. 6(5): p. 1945-1955. 60. Brouns, F., Heat--sweat--dehydration--rehydration: a praxis oriented approach. J Sports Sci, 1991. 9 Spec No: p. 143-52. 61. Cranston, W., Temperature regulation. British medical journal, 1966. 2(5505): p. 69. 62. Consolazio, C.F., et al., Environmental temperature and energy expenditures. J Appl Physiol, 1963. 18: p. 65-8. 63. Sawka, M.N., et al., Physiological responses to exercise in the heat, in Nutritional Needs in Hot Environments, M. B.M., Editor. 1993, National Academy Press: Washington, DC. 64. Askew, E.W., Nutrition and performance at environmental extremes, in Nutrition in exercise and sport, I. Wolinsky and J.F. Hickson, Editors. 1993, CRC Press, Inc.: Boca Raton, Florida. p. 456-470.

210

65. Cheung, S.S. and T.M. McLellan, Heat acclimation, aerobic fitness, and hydration effects on tolerance during uncompensable heat stress. J Appl Physiol, 1998. 84(5): p. 1731- 1739. 66. Gray, E.L., F.C. Consolazio, and R.M. Kark, Nutritional Requirements for Men at Work in Cold, Temperate and Hot Environments. Journal of Applied Physiology, 1951. 4(4): p. 270-275. 67. Yamashita, Y., et al., Decreased energy expenditure during prolonged sub-maximal exercise in a warm environment. European Journal of Sport Science, 2005. 5(4): p. 153- 158. 68. Rowell, L.B., et al., Human metabolic responses to hyperthermia during mild to maximal exercise. J Appl Physiol, 1969. 26(4): p. 395-402. 69. Consolazio, C.F., et al., Environmental temperature and energy expenditures. Journal of Applied Physiology, 1963. 18(1): p. 65-68. 70. Fink, W., D. Costill, and P. Van Handel, Leg muscle metabolism during exercise in the heat and cold. European journal of applied physiology and occupational physiology, 1975. 34(1): p. 183-190. 71. Al-Haboubi, M.H., Energy expenditure during moderate work at various climates. International Journal of Industrial Ergonomics, 1996. 17: p. 379-388. 72. Sun, Y. and N. Zhu, Experimental research on human metabolism in hot environment. Advanced Materials Research, 2012. 599: p. 241-244. 73. Brouha, L., et al., Physiological reactions of men and women during muscular activity and recovery in various environments. Journal of Applied Physiology, 1961. 16(1): p. 133- 140. 74. Smolander, J., et al., Aerobic and anaerobic responses to incremental exercise in a thermoneutral and a hot dry environment. Acta Physiol Scand, 1986. 128(1): p. 15-21. 75. Williams, C.G., et al., Circulatory and metabolic reactions to work in heat. J Appl Physiol, 1962. 17: p. 625-38. 76. Febbraio, M.A., et al., Muscle metabolism during exercise and heat stress in trained men: effect of acclimation. J Appl Physiol (1985), 1994. 76(2): p. 589-97. 77. Dauncey, M.J., Influence of mild cold on 24 h energy expenditure, resting metabolism and diet-induced thermogenesis. Br J Nutr, 1981. 45(2): p. 257-67. 78. Noakes, T.D., Exercise and the cold. Ergonomics, 2000. 43(10): p. 1461-79. 79. Young, A.J., M.N. Sawka, and K.B. Pandolf, Physiology of cold exposure, in Nutritional needs in cold and in high-altitude environments, B.M. Marriott and S.J. Carlson, Editors. 1996, National Academy Press: Washington, D.C. 80. Doubt, T.J., Physiology of exercise in the cold. Sports Med, 1991. 11(6): p. 367-81. 81. Leblanc, J.A., Effect of environmental temperature on energy expenditure and caloric requirements. J Appl Physiol, 1957. 10(2): p. 281-3. 82. Romet, T., et al., The metabolic cost of exercise in a cold air environment (-20 [degrees]C). Medicine & Science in Sports & Exercise, 1983. 15(2): p. 156. 83. Pandolf, K., M. Haisman, and R. Goldman, Metabolic energy expenditure and terrain coefficients for walking on snow. Ergonomics, 1976. 19(6): p. 683-690. 84. Booth, C., et al., The effect of consumption of Australian combat rations on military personnel after a medium-term field exercise. 2001. 85. Burstein, R., et al., Energy Balance In Military Recruits Performing Intense Physical Efforts Under Extreme Climate Conditions. 1994, DTIC Document. 86. Forbes-Ewan, C.H., et al., Use of doubly labeled water technique in soldiers training for jungle warfare. J Appl Physiol, 1989. 67(1): p. 14-8.

211

87. Hoyt, R.W., et al., Warfighter physiological status monitoring (WPSM): energy balance and thermal status during a 10-day cold weather US Marine Corps Infantry Officer Course field exercise. 2001, DTIC Document. 88. Hoyt, R.W., et al., Doubly labeled water measurement of human energy expenditure during strenuous exercise. J Appl Physiol, 1991. 71(1): p. 16-22. 89. King, N., et al., Nutrition and hydration status of soldiers consuming the 18-Man Arctic Tray Pack Ration Module with either the Meal, Ready-to-Eat or the Long Life Ration Packet during a cold weather field training exercise. 1992, DTIC Document. 90. Frykman, P., et al., Effects of a World Record Unsupported Ski Trek Across Greenland (The G2 Expedition) on Physical Performance and Body Composition. 2001, DTIC Document. 91. Neilson, H.K., et al., Estimating activity energy expenditure: how valid are physical activity questionnaires? Am J Clin Nutr, 2008. 87(2): p. 279-291. 92. Roberts, S.B. and G.E. Dallal, Energy requirements and aging. Public Health Nutrition- Cab International, 2005. 8(7A): p. 1028. 93. Ainslie, P.N., T. Reilly, and K.R. Westerterp, Estimating human energy expenditure: a review of techniques with particular reference to doubly labelled water. Sports Medicine, 2003. 33(9): p. 683-698. 94. Westerterp, K., et al., Use of the doubly labeled water technique in humans during heavy sustained exercise. J Appl Physiol, 1986. 61(6): p. 2162-2167. 95. Hoyt, R.W., et al., Total energy expenditure estimated using foot-ground contact pedometry. Diabetes technology & therapeutics, 2004. 6(1): p. 71-81. 96. Kinnunen, H., et al., Wrist-worn accelerometers in assessment of energy expenditure during intensive training. Physiological Measurement, 2012. 33(11): p. 1841. 97. Starling, R.D., et al., Assessment of physical activity in older individuals: a doubly labeled water study. J Appl Physiol, 1999. 86(6): p. 2090-2096. 98. Li, R., P. Deurenberg, and J.G. Hautvast, A critical evaluation of heart rate monitoring to assess energy expenditure in individuals. Am J Clin Nutr, 1993. 58(5): p. 602-7. 99. Morio, B., et al., Critical evaluation of the factorial and heart-rate recording methods for the determination of energy expenditure of free-living elderly people. British Journal of Nutrition, 1997. 78: p. 709-722. 100. Spurr, G., et al., Energy expenditure from minute-by-minute heart-rate recording: comparison with indirect calorimetry. Am J Clin Nutr, 1988. 48(3): p. 552-559. 101. Schoeller, D.A. and P. Webb, Five-day comparison of the doubly labeled water method with respiratory gas exchange. Am J Clin Nutr, 1984. 40(1): p. 153-8. 102. Warwick, P.M. and J. Baines, Energy expenditure in free-living smokers and nonsmokers: comparison between factorial, intake-balance, and doubly labeled water measures. Am J Clin Nutr, 1996. 63(1): p. 15-21. 103. Seale, J.L., et al., Comparison of doubly labeled water, intake-balance, and direct- and indirect-calorimetry methods for measuring energy expenditure in adult men. Am J Clin Nutr, 1990. 52(1): p. 66-71. 104. Speakman, J.R., The history and theory of the doubly labeled water technique. Am J Clin Nutr, 1998. 68(4): p. 932S-938S. 105. Ainsworth, B.E., et al., 2011 Compendium of Physical Activities: a second update of codes and MET values. Med Sci Sports Exerc, 2011. 43(8): p. 1575-81. 106. Goldman, R., Energy expenditure of soldiers performing combat type activities. Ergonomics, 1965. 8(3): p. 321-327.

212

107. Consolazio, C.F., Energy expenditure studies in military populations using Kofranyi- Michaelis respirometers. The American journal of clinical nutrition, 1971. 24(12): p. 1431-1437. 108. Rennie, K., et al., A combined heart rate and movement sensor: proof of concept and preliminary testing study. European journal of clinical nutrition, 2000. 54(5): p. 409. 109. Brage, S., et al., Branched equation modeling of simultaneous accelerometry and heart rate monitoring improves estimate of directly measured physical activity energy expenditure. J Appl Physiol, 2004. 96(1): p. 343-351. 110. Assah, F.K., et al., Accuracy and validity of a combined heart rate and motion sensor for the measurement of free-living physical activity energy expenditure in adults in Cameroon. Int J Epidemiol, 2011. 40(1): p. 112-120. 111. Villars, C., et al., Validity of combining heart rate and uniaxial acceleration to measure free-living physical activity energy expenditure in young men. J Appl Physiol, 2012. 112. Edwards, J.S., et al., An assessment of the nutritional intake and energy expenditure of unacclimatized US Army soldiers living and working at high altitude. 1991, USARIEM. 113. Friedl, K.E. and R.W. Hoyt, Development and biomedical testing of military operational rations. Annu Rev Nutr, 1997. 17: p. 51-75. 114. Baker-Fulco, C.J., et al., Nutrition for Health and Performance, 2001: Nutritional Guidance for Military Operations in Temperate and Extreme Environments. 2001, DTIC Document. 115. Dictionary, O.E., Oxford English dictionary online. 2007, JSTOR. 116. Parker, B., et al., Relation between food intake and visual analogue scale ratings of appetite and other sensations in healthy older and young subjects. European journal of clinical nutrition, 2004. 58(2): p. 212-218. 117. Holmstrup, M., et al., Satiety, but not total PYY, Is increased with continuous and intermittent exercise. Obesity, 2013. 21(10): p. 2014-2020. 118. Blundell, J., et al., Appetite control: methodological aspects of the evaluation of foods. Obesity reviews, 2010. 11(3): p. 251-270. 119. Vuksan, V., et al., Viscosity of fiber preloads affects food intake in adolescents. Nutrition, Metabolism and Cardiovascular Diseases, 2009. 19(7): p. 498-503. 120. Wansink, B., J.E. Painter, and J. North, Bottomless bowls: why visual cues of portion size may influence intake. Obesity, 2005. 13(1): p. 93-100. 121. Martins, C., et al., Effects of exercise on gut peptides, energy intake and appetite. J Endocrinol, 2007. 193(2): p. 251-8. 122. Wren, A.M., et al., Ghrelin enhances appetite and increases food intake in humans. J Clin Endocrinol Metab, 2001. 86(12): p. 5992. 123. King, J.A., L.K. Wasse, and D.J. Stensel, The acute effects of swimming on appetite, food intake, and plasma acylated ghrelin. Journal of obesity, 2010. 2011. 124. Verger, P., M.T. Lanteaume, and J. Louis-Sylvestre, Human intake and choice of foods at intervals after exercise. Appetite, 1992. 18(2): p. 93-99. 125. Hetherington, M.M., et al., Understanding variety: tasting different foods delays satiation. Physiology & behavior, 2006. 87(2): p. 263-271. 126. Wilson, M.-M.G., et al., Appetite assessment: simple appetite questionnaire predicts weight loss in community-dwelling adults and nursing home residents. The American journal of clinical nutrition, 2005. 82(5): p. 1074-1081. 127. Lou, L., et al., Evaluation of food intake in hemodialysis using a food consumption and appetite questionnaire. Nefrologia: publicacion oficial de la Sociedad Espanola Nefrologia, 2002. 22(5): p. 438-447.

213

128. Stubbs, R.J., et al., The use of visual analogue scales to assess motivation to eat in human subjects: a review of their reliability and validity with an evaluation of new hand-held computerized systems for temporal tracking of appetite ratings. British Journal of Nutrition, 2000. 84(4): p. 405-415. 129. Flint, A., et al., Reproducibility, power and validity of visual analogue scales in assessment of appetite sensations in single test meal studies. International journal of obesity, 2000. 24(1): p. 38. 130. De Castro, J.M. and D.K. Elmore, Subjective hunger relationships with meal patterns in the spontaneous feeding behavior of humans: Evidence for a causal connection. Physiology & behavior, 1988. 43(2): p. 159-165. 131. Douglas, J.A., et al., Acute effects of exercise on appetite, ad libitum energy intake and appetite-regulatory hormones in lean and overweight/obese men and women. International Journal of Obesity, 2017: p. ijo2017181. 132. Hagobian, T.A., et al., Effects of acute exercise on appetite hormones and ad libitum energy intake in men and women. Applied Physiology, Nutrition, and Metabolism, 2012. 38(999): p. 66-72. 133. Holt, G.M., et al., Systematic literature review shows that appetite rating does not predict energy intake. Critical reviews in food science and nutrition, 2017. 57(16): p. 3577-3582. 134. Mattes, R., Hunger ratings are not a valid proxy measure of reported food intake in humans. Appetite, 1990. 15(2): p. 103-113. 135. Williams, G., J.A. Harrold, and D.J. Cutler, The hypothalamus and the regulation of energy homeostasis: lifting the lid on a black box. Proc Nutr Soc, 2000. 59(3): p. 385-96. 136. Zhan, C., et al., Acute and long-term suppression of feeding behavior by POMC neurons in the brainstem and hypothalamus, respectively. J Neurosci, 2013. 33(8): p. 3624-32. 137. Cegla, J., T.M. Tan, and S.R. Bloom, Gut-brain cross-talk in appetite regulation. Curr Opin Clin Nutr Metab Care, 2010. 13(5): p. 588-93. 138. Perry, B. and Y. Wang, Appetite regulation and weight control: the role of gut hormones. Nutr Diabetes, 2012. 2: p. e26. 139. Ostlund, R.E., Jr., et al., Relation between plasma leptin concentration and body fat, gender, diet, age, and metabolic covariates. J Clin Endocrinol Metab, 1996. 81(11): p. 3909-13. 140. Montague, C.T., et al., Congenital leptin deficiency is associated with severe early-onset obesity in humans. Nature, 1997. 387(6636): p. 903-8. 141. Heymsfield, S.B., et al., Recombinant leptin for weight loss in obese and lean adults: a randomized, controlled, dose-escalation trial. JAMA, 1999. 282(16): p. 1568-75. 142. Hukshorn, C.J., M.S. Westerterp-Plantenga, and W.H. Saris, Pegylated human recombinant leptin (PEG-OB) causes additional weight loss in severely energy-restricted, overweight men. Am J Clin Nutr, 2003. 77(4): p. 771-6. 143. Westerterp-Plantenga, M.S., et al., Effects of weekly administration of pegylated recombinant human OB protein on appetite profile and energy metabolism in obese men. Am J Clin Nutr, 2001. 74(4): p. 426-34. 144. Arch, J.R., M.J. Stock, and P. Trayhurn, Leptin resistance in obese humans: does it exist and what does it mean? Int J Obes Relat Metab Disord, 1998. 22(12): p. 1159-63. 145. Mars, M., et al., Fasting leptin and appetite responses induced by a 4-day 65%-energy- restricted diet. Int J Obes (Lond), 2006. 30(1): p. 122-8. 146. Rattarasarn, C., et al., Relationships of body fat distribution, insulin sensitivity and cardiovascular risk factors in lean, healthy non-diabetic Thai men and women. Diabetes Res Clin Pract, 2003. 60(2): p. 87-94.

214

147. Kieffer, T.J. and J.F. Habener, The adipoinsular axis: effects of leptin on pancreatic beta- cells. Am J Physiol Endocrinol Metab, 2000. 278(1): p. E1-E14. 148. Rodin, J., Insulin levels, hunger, and food intake: an example of feedback loops in body weight regulation. Health Psychol, 1985. 4(1): p. 1-24. 149. Woods, S.C., et al., Chronic intracerebroventricular infusion of insulin reduces food intake and body weight of baboons. Nature, 1979. 282(5738): p. 503-5. 150. Chapman, I.M., et al., Effect of intravenous glucose and euglycemic insulin infusions on short-term appetite and food intake. Am J Physiol, 1998. 274(3 Pt 2): p. R596-603. 151. Neary, N.M., C.J. Small, and S.R. Bloom, Gut and mind. Gut, 2003. 52(7): p. 918-21. 152. Druce, M.R., et al., Ghrelin increases food intake in obese as well as lean subjects. Int J Obes (Lond), 2005. 29(9): p. 1130-6. 153. English, P.J., et al., Food fails to suppress ghrelin levels in obese humans. J Clin Endocrinol Metab, 2002. 87(6): p. 2984. 154. Wortley, K.E., et al., Absence of ghrelin protects against early-onset obesity. J Clin Invest, 2005. 115(12): p. 3573-8. 155. Neary, N.M., et al., Acylated ghrelin stimulates food intake in the fed and fasted states but desacylated ghrelin has no effect. Gut, 2006. 55(1): p. 135-135. 156. Inhoff, T., et al., Desacyl ghrelin inhibits the orexigenic effect of peripherally injected ghrelin in rats. Peptides, 2008. 29(12): p. 2159-2168. 157. Asakawa, A., et al., Stomach regulates energy balance via acylated ghrelin and desacyl ghrelin. Gut, 2005. 54(1): p. 18-24. 158. Williams, D.L., Expecting to eat: glucagon-like peptide-1 and the anticipation of meals. Endocrinology, 2010. 151(2): p. 445-7. 159. Secher, A., et al., The arcuate nucleus mediates GLP-1 receptor agonist liraglutide- dependent weight loss. J Clin Invest, 2014. 124(10): p. 4473-88. 160. Gutzwiller, J.P., et al., Glucagon-like peptide-1: a potent regulator of food intake in humans. Gut, 1999. 44(1): p. 81-6. 161. Naslund, E., et al., Energy intake and appetite are suppressed by glucagon-like peptide-1 (GLP-1) in obese men. Int J Obes Relat Metab Disord, 1999. 23(3): p. 304-11. 162. Cohen, M.A., et al., Oxyntomodulin suppresses appetite and reduces food intake in humans. J Clin Endocrinol Metab, 2003. 88(10): p. 4696-701. 163. Druce, M.R. and S.R. Bloom, Oxyntomodulin : a novel potential treatment for obesity. Treat Endocrinol, 2006. 5(5): p. 265-72. 164. Adrian, T.E., et al., Human distribution and release of a putative new gut hormone, peptide YY. Gastroenterology, 1985. 89(5): p. 1070-7. 165. Price, S.L. and S.R. Bloom, Protein PYY and its role in metabolism. Front Horm Res, 2014. 42: p. 147-54. 166. Batterham, R.L., et al., Critical role for peptide YY in protein-mediated satiation and body- weight regulation. Cell Metab, 2006. 4(3): p. 223-33. 167. De Silva, A., et al., The gut hormones PYY 3-36 and GLP-1 7-36 amide reduce food intake and modulate brain activity in appetite centers in humans. Cell Metab, 2011. 14(5): p. 700-6. 168. Batterham, R.L., et al., Gut hormone PYY(3-36) physiologically inhibits food intake. Nature, 2002. 418(6898): p. 650-4. 169. Sainsbury, A., et al., Y4 receptor knockout rescues fertility in ob/ob mice. Genes & development, 2002. 16(9): p. 1077-1088. 170. Kojima, S., et al., A role for pancreatic polypeptide in feeding and body weight regulation. Peptides, 2007. 28(2): p. 459-463.

215

171. Loh, K., H. Herzog, and Y.-C. Shi, Regulation of energy homeostasis by the NPY system. Trends in Endocrinology & Metabolism, 2015. 26(3): p. 125-135. 172. Adrian, T.E., et al., Distribution and release of human pancreatic polypeptide. Gut, 1976. 17(12): p. 940-944. 173. Asakawa, A., et al., Characterization of the effects of pancreatic polypeptide in the regulation of energy balance. Gastroenterology, 2003. 124(5): p. 1325-36. 174. Batterham, R.L., et al., Pancreatic polypeptide reduces appetite and food intake in humans. J Clin Endocrinol Metab, 2003. 88(8): p. 3989-92. 175. Lieverse, R.J., et al., Satiety effects of a physiological dose of cholecystokinin in humans. Gut, 1995. 36(2): p. 176-9. 176. Crawley, J.N. and M.C. Beinfeld, Rapid development of tolerance to the behavioural actions of cholecystokinin. Nature, 1983. 302(5910): p. 703-6. 177. Thorogood, A., et al., Isolated aerobic exercise and weight loss: a systematic review and meta-analysis of randomized controlled trials. The American journal of medicine, 2011. 124(8): p. 747-755. 178. Johns, D.J., et al., Diet or exercise interventions vs combined behavioral weight management programs: a systematic review and meta-analysis of direct comparisons. Journal of the Academy of Nutrition and Dietetics, 2014. 114(10): p. 1557-1568. 179. King, N.A., A. Tremblay, and J.E. Blundell, Effects of exercise on appetite control: implications for energy balance. Med Sci Sports Exerc, 1997. 29(8): p. 1076-89. 180. Ueda, S.Y., et al., Comparable effects of moderate intensity exercise on changes in anorectic gut hormone levels and energy intake to high intensity exercise. J Endocrinol, 2009. 203(3): p. 357-64. 181. Shorten, A.L., K.E. Wallman, and K.J. Guelfi, Acute effect of environmental temperature during exercise on subsequent energy intake in active men. Am J Clin Nutr, 2009. 90(5): p. 1215-21. 182. King, N.A., V.J. Burley, and J.E. Blundell, Exercise-induced suppression of appetite: effects on food intake and implications for energy balance. European journal of clinical nutrition, 1994. 48(10): p. 715-724. 183. Imbeault, P., et al., Acute effects of exercise on energy intake and feeding behaviour. British Journal of Nutrition, 1997. 77(4): p. 511-521. 184. Hubert, P., N. King, and J. Blundell, Uncoupling the effects of energy expenditure and energy intake: appetite response to short-term energy deficit induced by meal omission and physical activity. Appetite, 1998. 31(1): p. 9-19. 185. Pomerleau, M., et al., Effects of exercise intensity on food intake and appetite in women. The American journal of clinical nutrition, 2004. 80(5): p. 1230-1236. 186. Erdmann, J., et al., Plasma ghrelin levels during exercise—effects of intensity and duration. Regulatory peptides, 2007. 143(1): p. 127-135. 187. King, J.A., et al., Influence of prolonged treadmill running on appetite, energy intake and circulating concentrations of acylated ghrelin. Appetite, 2010. 54(3): p. 492-8. 188. Vatansever-Ozen, S., et al., The effects of exercise on food intake and hunger: Relationship with acylated ghrelin and leptin. Journal of Sports Science and Medicine, 2011. 10(2): p. 283-91. 189. Wasse, L.K., et al., Influence of rest and exercise at a simulated altitude of 4,000 m on appetite, energy intake, and plasma concentrations of acylated ghrelin and peptide YY. Journal of Applied Physiology, 2012. 112(4): p. 552-559.

216

190. Deighton, K., et al., Appetite, gut hormone and energy intake responses to low volume sprint interval and traditional endurance exercise. European journal of applied physiology, 2013. 113(5): p. 1147-1156. 191. Balaguera-Cortes, L., et al., Energy intake and appetite-related hormones following acute aerobic and resistance exercise. Applied Physiology, Nutrition, and Metabolism, 2011. 36(6): p. 958-966. 192. Faure, C., et al., Effect of heat exposure and exercise on food intake regulation: A randomized crossover study in young healthy men. Metabolism, 2016. 65(10): p. 1541- 1549. 193. Holliday, A. and A.K. Blannin, Very Low Volume Sprint Interval Exercise Suppresses Subjective Appetite, Lowers Acylated Ghrelin, and Elevates GLP-1 in Overweight Individuals: A Pilot Study. Nutrients, 2017. 9(4): p. 362. 194. Kojima, C., et al., The Effect of a 20 km Run on Appetite Regulation in Long Distance Runners. Nutrients, 2016. 8(11): p. 672. 195. Donnelly, J.E., et al., Does increased exercise or physical activity alter ad-libitum daily energy intake or macronutrient composition in healthy adults? A systematic review. PloS one, 2014. 9(1): p. e83498. 196. Ueda, S.-y., et al., Comparable effects of moderate intensity exercise on changes in anorectic gut hormone levels and energy intake to high intensity exercise. Journal of Endocrinology, 2009. 203(3): p. 357-364. 197. Broom, D.R., et al., Exercise-induced suppression of acylated ghrelin in humans. Journal of applied physiology, 2007. 102(6): p. 2165-2171. 198. Kawano, H., et al., Effects of different modes of exercise on appetite and appetite- regulating hormones. Appetite, 2013. 66: p. 26-33. 199. Larsen, P.S., et al., Effects of Aerobic, Strength or Combined Exercise on Perceived Appetite and Appetite-Related Hormones in Inactive Middle-Aged Men. International journal of sport nutrition and exercise metabolism, 2017. 27(5): p. 389-398. 200. Ueda, S.Y., et al., Changes in gut hormone levels and negative energy balance during aerobic exercise in obese young males. J Endocrinol, 2009. 201(1): p. 151-9. 201. Broom, D.R., et al., Influence of resistance and aerobic exercise on hunger, circulating levels of acylated ghrelin, and peptide YY in healthy males. Am J Physiol Regul Integr Comp Physiol, 2009. 296(1): p. R29-35. 202. Howe, S.M., et al., No Effect of Exercise Intensity on Appetite in Highly-Trained Endurance Women. Nutrients, 2016. 8(4): p. 223. 203. Schubert, M.M., et al., Acute exercise and hormones related to appetite regulation: a meta-analysis. Sports Med, 2014. 44(3): p. 387-403. 204. Broom, D.R., et al., Acute effect of exercise intensity and duration on acylated ghrelin and hunger in men. Journal of Endocrinology, 2017. 232(3): p. 411-422. 205. Chanoine, J.P., et al., GLP-1 and appetite responses to a meal in lean and overweight adolescents following exercise. Obesity (Silver Spring), 2008. 16(1): p. 202-4. 206. Broom, D.R., et al., Exercise-induced suppression of acylated ghrelin in humans. J Appl Physiol (1985), 2007. 102(6): p. 2165-71. 207. Martins, C., et al., Effect of moderate-and high-intensity acute exercise on appetite in obese individuals. Medicine & Science in Sports & Exercise, 2015. 47(1): p. 40-48. 208. Hilsted, J., et al., Gastroenteropancreatic hormonal changes during exercise. American Journal of Physiology-Gastrointestinal and Liver Physiology, 1980. 239(3): p. G136- G140.

217

209. Hagobian, T.A., et al., Effects of exercise on energy-regulating hormones and appetite in men and women. Am J Physiol Regul Integr Comp Physiol, 2009. 296(2): p. R233-42. 210. Cheng, M.H., et al., Appetite regulation via exercise prior or subsequent to high-fat meal consumption. Appetite, 2009. 52(1): p. 193-8. 211. Kraemer, R.R., H. Chu, and V.D. Castracane, Leptin and exercise. Experimental Biology and Medicine, 2002. 227(9): p. 701-708. 212. Leal-Cerro, A., et al., Serum leptin levels in male marathon athletes before and after the marathon run. The Journal of Clinical Endocrinology & Metabolism, 1998. 83(7): p. 2376-2379. 213. Castell, L.M., C. Douglas Thake, and W. Ensign, Biochemical markers of possible immunodepression in military training in harsh environments. Military medicine, 2010. 175(3): p. 158-165. 214. Lavin, J.H., et al., Interaction of insulin, glucagon-like peptide 1, gastric inhibitory polypeptide, and appetite in response to intraduodenal carbohydrate. The American journal of clinical nutrition, 1998. 68(3): p. 591-598. 215. Brobeck, J.R., Food intake as a mechanism of temperature regulation. The Yale journal of biology and medicine, 1948. 20(6): p. 545. 216. Refinetti, R., Effects of food temperature and ambient temperature during a meal on food intake in the rat. Physiology & behavior, 1988. 43(2): p. 245-247. 217. Ingram, D. and K. Legge, Effects of environmental temperature on food intake in growing pigs. Comparative Biochemistry and Physiology Part A: Physiology, 1974. 48(3): p. 573- 581. 218. Johnson, R.E. and R.M. Kark, Environment and food intake in man. Science (Washington), 1947: p. 378-9. 219. Westerterp-Plantenga, M., et al., Energy metabolism in women during short exposure to the thermoneutral zone. Physiology & behavior, 2002. 75(1): p. 227-235. 220. Bernhard, M.C., et al., Warm ambient temperature decreases food intake in a simulated office setting: A pilot randomized controlled trial. Frontiers in nutrition, 2015. 2. 221. Westerterp-Plantenga, M.S., et al., Energy metabolism in humans at a lowered ambient temperature. European journal of clinical nutrition, 2002. 56(4): p. 288. 222. Langeveld, M., et al., Mild cold effects on hunger, food intake, satiety and skin temperature in humans. Endocrine connections, 2016. 5(2): p. 65-73. 223. Pizon, M., et al., Low ambient temperature lowers cholecystokinin and leptin plasma concentrations in adult men. The Open Nutrition Journal, 2009. 3: p. 5-7. 224. Ricci, M.R., S.K. Fried, and K.D. Mittleman, Acute cold exposure decreases plasma leptin in women. Metabolism, 2000. 49(4): p. 421-3. 225. Tomasik, P., K. Sztefko, and M. Pizon, The effect of short-term cold and hot exposure on total plasma ghrelin concentrations in humans. Hormone and metabolic research, 2005. 37(03): p. 189-190. 226. Wasse, L.K., et al., Effect of ambient temperature during acute aerobic exercise on short- term appetite, energy intake, and plasma acylated ghrelin in recreationally active males. Applied Physiology, Nutrition, and Metabolism, 2013. 38(8): p. 905-909. 227. Kojima, C., et al., The influence of environmental temperature on appetite-related hormonal responses. Journal of physiological anthropology, 2015. 34(1): p. 1. 228. Laursen, T.L., et al., Leptin, adiponectin, and ghrelin responses to endurance exercise in different ambient conditions. Temperature, 2017: p. 1-10.

218

229. Crabtree, D.R. and A.K. Blannin, Effects of exercise in the cold on Ghrelin, PYY, and food intake in overweight adults. Medicine and science in sports and exercise, 2015. 47(1): p. 49-57. 230. White, L.J., et al., Increased caloric intake soon after exercise in cold water. International journal of sport nutrition and exercise metabolism, 2005. 15(1): p. 38-47. 231. Dressendorfer, R., 228 Effect Of Internal Body Temperature On Energy Intake Soon After Aerobic Exercise. Medicine & Science in Sports & Exercise, 1993. 25(5): p. S42. 232. Halse, R.E., K.E. Wallman, and K.J. Guelfi, Postexercise water immersion increases short-term food intake in trained men. Medicine & Science in Sports & Exercise, 2011. 43(4): p. 632-638. 233. Shorten, A.L., K.E. Wallman, and K.J. Guelfi, Acute effect of environmental temperature during exercise on subsequent energy intake in active men. The American journal of clinical nutrition, 2009. 90(5): p. 1215-1221. 234. Duggan, A., Energy cost of stepping in protective clothing ensembles. Ergonomics, 1988. 31(1): p. 3-11. 235. King, J.A., et al., Differential Acylated Ghrelin, Peptide YY3–36, Appetite, and Food Intake Responses to Equivalent Energy Deficits Created by Exercise and Food Restriction. Journal of Clinical Endocrinology & Metabolism, 2011. 96(4): p. 1114-1121. 236. Borg, G., Perceived exertion as an indicator of somatic stress. Scandinavian journal of rehabilitation medicine, 1970. 2(2): p. 92-98. 237. Bruce, R.A., F. Kusumi, and D. Hosmer, Maximal oxygen intake and nomographic assessment of functional aerobic impairment in cardiovascular disease. American Heart Journal, 1973. 85(4): p. 546-562. 238. McNair, D.M., Manual profile of mood states. 1971: Educational & Industrial testing service. 239. Girden, E.R., ANOVA: Repeated measures. 1992: Sage. 240. Gunasekara, F.I., et al., Fixed effects analysis of repeated measures data. International journal of epidemiology, 2013. 43(1): p. 264-269. 241. Henry, C., Basal metabolic rate studies in humans: measurement and development of new equations. Public health nutrition, 2005. 8(7a): p. 1133-1152. 242. McMurray, R.G., et al., Examining variations of resting metabolic rate of adults: a public health perspective. Medicine and science in sports and exercise, 2014. 46(7): p. 1352. 243. Haugen, H.A., et al., Variability of measured resting metabolic rate. The American journal of clinical nutrition, 2003. 78(6): p. 1141-1145. 244. Kinabo, J. and J. Durnin, Thermic effect of food in man: effect of meal composition, and energy content. British Journal of Nutrition, 1990. 64(1): p. 37-44. 245. Ricciardi, R., P.A. Deuster, and L.A. Talbot, Metabolic demands of body armor on physical performance in simulated conditions. Military medicine, 2008. 173(9): p. 817- 824. 246. Christie, C.J. and P.A. Scott, Metabolic responses of South African soldiers during simulated marching with 16 combinations of speed and backpack load. Military medicine, 2005. 170(7): p. 619-622. 247. Whigham, L., et al., Effect of clothing weight on body weight. International Journal of Obesity, 2013. 37(1): p. 160-161. 248. Kwon, J. and J. Choi, The relationship between environmental temperature and clothing insulation across a year. International journal of biometeorology, 2012. 56(5): p. 887-893. 249. Periard, J.D., et al., Cardiovascular strain impairs prolonged self-paced exercise in the heat. Exp Physiol, 2011. 96(2): p. 134-44.

219

250. Tucker, R., et al., Impaired exercise performance in the heat is associated with an anticipatory reduction in skeletal muscle recruitment. Pflugers Arch, 2004. 448(4): p. 422- 30. 251. Satinoff, E., Behavioral thermoregulation in the cold. Comprehensive Physiology, 2011. 252. Ahmed, M., Assessments of Dietary Intakes of Canadian Armed Forces Consuming Field Rations (Doctoral Thesis). 2017, University of Toronto: Toronto, Canada. 253. Booth, D., Conditioned satiety in the rat. Journal of comparative and physiological psychology, 1972. 81(3): p. 457. 254. Defence, D.o.N., Women in the Canadian Armed Forces, D.o.N. Defence, Editor. 2016. 255. Kozey, S., et al., Errors in MET Estimates of Physical Activities Using 3.5 ml. kg-1min-1 as the Baseline Oxygen Consumption. Journal of physical activity & health, 2010. 7(4): p. 508. 256. Young, A., et al., Sustaining health and performance in the cold: A pocket guide to environmental medicine aspects of cold-weather operations. 1992, USARIEM, Natick, MA. 257. Hedblom, E., Polar Manual Arctic and Antarctic Living Conditions, Personnel selection, nygiene and sanitation, clothing, nutrition, supplies and equipment, visual disabilities and cold injuries, first aid, safety, and survival. 1965, Naval Medical School, Bethesda MD. 258. Bray, R.M., et al., The effects of stress on job functioning of military men and women. Armed Forces & Society, 2001. 27(3): p. 397-417. 259. Kyröläinen, H., et al., Hormonal responses during a prolonged military field exercise with variable exercise intensity. European journal of applied physiology, 2008. 102(5): p. 539- 546. 260. Opstad, P.K., Circadian rhythm of hormones is extinguished during prolonged physical stress, sleep and energy deficiency in young men. European journal of endocrinology, 1994. 131(1): p. 56-66. 261. Weinstein, D.D., et al., Minor physical anomalies, dermatoglyphic asymmetries, and cortisol levels in adolescents with schizotypal personality disorder. American Journal of Psychiatry, 1999. 156(4): p. 617-623. 262. Jacobs, I., et al., Muscle glycogen in soldiers on different diets during military field manoeuvres. Aviat Space Environ Med, 1983. 54(10): p. 898-900. 263. Hagobian, T.A., et al., Effects of exercise on energy-regulating hormones and appetite in men and women. American Journal of Physiology-Regulatory, Integrative and Comparative Physiology, 2009. 296(2): p. R233-R242. 264. Ueda, S.-y., et al., Changes in gut hormone levels and negative energy balance during aerobic exercise in obese young males. Journal of Endocrinology, 2009. 201(1): p. 151- 159. 265. King, J.A., et al., Influence of prolonged treadmill running on appetite, energy intake and circulating concentrations of acylated ghrelin. Appetite, 2010. 54(3): p. 492-498. 266. Mackelvie, K.J., et al., Regulation of appetite in lean and obese adolescents after exercise: role of acylated and desacyl ghrelin. Journal of Clinical Endocrinology & Metabolism, 2007. 92(2): p. 648-654. 267. Chanoine, J.P., et al., GLP‐1 and Appetite Responses to a Meal in Lean and Overweight Adolescents Following Exercise. Obesity, 2008. 16(1): p. 202-204.

220

Appendices Appendix A – Weight Loss in Military Personnel Table I: Typical weight loss in military personnel when adequate rations are provided (short term) Weight Length Breakdown of Amount Loss Total Energy Rations Provided Population Environment of Trial Macronutrients % Consumed kcal (% of Expenditure (kcal/day) Reference (Kcal/day) (days) (CHO/PRO/FAT) (% Consumed) Initial OR Activities Weight) 4 Zimbabwean Gained Food preparation, cleaning 40°C (29%) 3762 + CHO support 12 52/14/35 NA 1.1kg and maintaining camp [2] (Hot and Dry) drink soldiers (2%) 3344±239 19 Heat 3500 Gained Various military training 3301 ± 380 Acclimated 30°C -34°C 10 ad libitum 64/21/15 0.24kg exercises and jungle [26] (94%) soldiers (0%) manoeuvres 3350±555 Reconnaissance, 8 Soldiers 3880 55/11/35 (86%) Combat activities, marches, -5°C to -25°C 4.5 NA [262] 3721±719 bivouacking, winter war 8 Soldiers 5420 64/10/27 (69%) games. Unitized Group Ration (UGR) 2 2631±498 0.4 Artillery training exercise. 31 US Marines (7°C -31°C A, B, or T Rations 48/16/37 (66%) (0%) Different crews likely had (10% - 55 %) + 1 MRE 12 different tasks: i.e. loader [34] Warm and (4000) vs. radio operator Dry) UGR (4000) + 8% 3050±700 0.3 4115±724 32 US Marines maltodextrin drink 61/12/28 (76%) (0%) ad libitum 2693±596* 20 Very Cold <- (61%) 0.63kg Canadian 10 4350 48/18/34 4317±927 [27] 30°C *(Intake was (1%) Soldiers underreported)

221

Table I: Typical weight loss in military personnel when adequate rations are provided (short term) Amount Weight Total Energy Length of Rations Breakdown of Consumed Loss Expenditure Population Environment Trial Provided Macronutrients % kcal (% of Reference (kcal/day) (days) (Kcal/day) (CHO/PRO/FAT) (% Initial OR Activities Consumed) Weight) IMP 2571 ± 303 1.9kg ad libitum 57/13/26 (71%) (2.4%) 29 (3600) Canadian Armed Temperate IMP 5 >6000 [5] Forces Airborne Climate ad libitum 1.1kg 3217 ± 410 Regiment + mandatory 67/10/20 (1.4%) (71%) 240g of CHO (~4560) 2349 ± 449 at sea level 10 Soldiers (65%) NA Sedentary (SED) 1513 ± 885 at 45/14/41 altitude + (42%) 6 in total → Sugar free juice (0g of 2881 ± 981 at CHO) + hot chocolate (8g sea level 4100 m. -2 days at MRE of CHO) (80%) 13 Soldiers (EX) altitude. sea level ad libitum 2.1kg (3%) [12] 1787 ± 815 at -10°C to 32°C (3600) altitude -4 days at Two h run each (50%) altitude day at 75% of max 2300 ± 360 at heart rate 45/14/41 sea level 6 Soldiers (EX + + (64%) 1.4kg (2%) CHO) Juice (36g of CHO) + hot 2325 ± 527 at chocolate (42g of CHO) altitude (65%) 2195 m. 12 → altitude. -7 days of MRE 2390 ± 724 15 Soldiers 16.1°C exercise at ad libitum 56/22/22 2kg (3%) ~3600 [11] (66%) 78% humidity altitude (3600)

222

Table I: Typical weight loss in military personnel when adequate rations are provided (short term) Weight Total Energy Length of Breakdown of Amount Loss Rations Provided Expenditure Population Environment Trial Macronutrients % Consumed kcal (% of Reference (Kcal/day) (kcal/day) (days) (CHO/PRO/FAT) (% Consumed) Initial OR Activities Weight) 4 MRE 2024 2.2kg 32 Soldiers Group 1 VI/man/day 42/17/42 (42%) (2.8%) (4816) 2810 1.6kg 31 Soldiers Group 2 -42°C to - 3.5 MRE (4571) 47/16/38 (61%) (2%) 2°C 10 3 MRE plus 1 Estimated at 4500 [6] (Very 2841 1.5kg 30 Soldiers Group 3 supplemental pack 45/17/39 Cold) (65%) (1.9%) (4352) 3 MRE plus 1 3562 1.4kg 34 Soldiers Group 4 supplemental pack 50/15/35 (76%) (1.7%) (4658) Males: 2284 ± Males: 2kg Males: 5099 ± 676 1046 (57%) (2.6%) 18 Soldiers (9 males, 9 -21°C to - 5 3 IMPs (4000) 68/12/21 Females: [36] females) 2°C Females: 1911 ± Females: 4643 ± 2.1kg 1194 (48%) 559 (2.7%) AVERAGE ~8 4178kcal 67% ± 14% 1.5% 4452kcal

223

Table II: Typical weight loss in military personnel when adequate rations are provided (long term) Amount Weight Total Energy Length of Breakdown of Consumed Loss Rations Provided Expenditure Population Environment Trial Macronutrients kcal (% of Reference (Kcal/day) (kcal/day) (days) (CHO/PRO/FAT) (% Initial OR Activities Consumed) Weight) 35 Male 2 B rations + 1 MRE 1.7kg 46/18/34 2140 (53%) Soldiers ad libitum (~4060) (2%) 2 B rations + 1 MRE + 32 Male 125g CHO 1.7kg Bolivia 48/18/33 2265 (50%) Soldiers Supplemental Pack ad (2%) 3500-4050 m 15 3549±608 [112] libitum (~4560) elevation 2 B rations + 1 MRE + 13 Female 125g CHO 0.5kg 52/16/31 1168 (26%) Soldiers Supplemental Pack ad (0.7%) libitum (~4560) 7 Special MRE (4020 available 2960 ± 455 1.1kg operations -1°C to 16°C 28 but only took 3600 49/15/36 (82% of what 3400 ± 260 [35] (1.5%) soldiers into the field) was taken) 39 US Marines Two B Rations (1430) 2866 ± 549 (diet analyzed + MRE (1200) = 62/18/20 (3.1%) (71%) for 34) 4060 26°C (70%) 60 3328 ± 637 [8] 21 US Marines Two T Rations (1420) 2572 ± 241 (diet analyzed + MRE (1200) = 61/19/20 (3.2%) (64%) for 17) 4040 Three MRE (3896) per ~3500 ± 750 day and/or food from 4.2kg 28 US soldiers 8°C-34°C 64 63/17/20 (90% of MRE) ~4500 ± 800 [7] the dining facility ad (5%) (30%-96%) libitum

Pre-mid 3.3kg deployment (4%) 249 Royal ~121 ~2750 ± 821 Marines (EE Afghanistan Multi-Climate Ration (average of pre, 66/13/21 ~3626 ± 450 [37] analyzed for 9°C -39°C (4098) mid and post 18) Mid to post deployment) Gained deployment 0.5kg ~86 (0%)

AVERAGE ~62 4170kcal 62% 2.4% 3681kcal

224

Table III: Typical weight loss in military personnel when restricted rations are provided Weight Length Breakdown of Kcal Total Energy Population Loss of Rations Provided Macronutrients Consumed Expenditure (number of Environment (% of Reference Trial (Kcal/day) Consumed (% (kcal/day) subjects) Initial (days) (CHO/PRO/FAT) Consumed) OR Activities Weight) 18 heat acclimated 585 ± 46 2.95kg 600 75/25/1 soldiers (98%) (4%) Various military 18 heat acclimated 948 ±76 3.47kg training exercises 30°C -34°C 10 1000 85/14/1 [26] soldiers (95%) (5%) and jungle 16 heat acclimated 1362 ± 38 3.09kg manoeuvres 1500 84/16/1 soldiers (91%) (4%) Temperate 20 MRE (1300) + 13 US Ranger Forest A rations (2246) + 1 4.9kg 17 ~4150 students 27°C (65%) Survival Meal = (7%) ~3400 Mountains 8 MRE (1300) + 15 A US Ranger 2.1kg 22°C 18 rations (2246) + 3 B ~4500 students (3%) rations (1433) = 2688 Not Measured Not [10] Swamp/jungle 13 MRE (1300) + 7 A Measured US Ranger 2.5kg 27°C (74%) 16 rations (2246) + 2 box ~3550 students (4%) lunches (690) = 2125 15 MRE (1300) + 6 A US Ranger Desert 14 rations (2246) + 4 B 2.6kg ~3800 students 19°C rations (1433) + 1 box (4%) lunch (690) = 2814 Special operations Lightweight Ration 1930 ± 125 4.3kg -1°C to 16°C 28 39/14/47 3400 ± 260 [35] soldiers (7) (1980) (97%) (5.6%) 10 men from the Norwegian 48-525 7.7kg 6353± 478 Half received: Military (97%) (10%) Forest 48 to 544 8/63/30 Academy 15°C -30°C 7 [57] 6 women from the Half received: Norwegian 48-454 5.9kg 5231 ± 478 64/19/17 Military (85%) (10%) (Women) Academy AVERAGE ~16 1823kcal 94% 5.7% 4426kcal

225

Appendix B – Effects of Exercise on Appetite-Regulating Hormones Table IV: Effects of exercise on circulating leptin and appetite Subjects Intervention Leptin Appetite Notes Reference - 7:00 am arrival. -Energy intake in NEUT was - 7:15 am → either: - The decreased energy higher than CON. 11 young, 1) rest (CON) or intake post-exercise in -↓in leptin - Relative energy intake (taking healthy, 2) running on a treadmill for 40 min at 70% the HEAT could be due following into account the energy [233] active O peak in (HEAT) → 36°C, with 30% RH or to increased body 2 breakfast expended during exercise) males 3) (NEUT) at 25°C with 30% RH temperature in the HEAT during HEAT was lower than - 8:30 am buffet-type breakfast was provided for CON. 30 min. Condition 1 Baseline→ no-exercise + subjects in -In men, lower leptin was 18 (9 men, energy balance - Fasting leptin related to decreased - Compared with DEF, men 9 women) Condition 2 (BAL)→ 4 days of exercise (treadmill concentrations desire to eat, decreased had: ↓desire to eat, hunger, and overweigh at 50–65% of V̇ O until 30% of total daily EE were higher in fullness and increased 2peak PFC ratings during BAL. t/ was expended) with added energy to maintain EB. women than hunger after DEF. [263] - Women reported no obese On day 5 a ‘meal tolerance test (MTT) was done’ men. - In women, no hormones difference in any subjective individual Condition 3 (DEF)→ 4 days of exercise (same as - No differences correlated with any of the appetite measures across trials. s above) without added energy to induce ED. across conditions appetite ratings. On day 5 – MTT - ↓ Hunger scores following (1) Meal consumption (M) 12 the test meal in the M and ME - Leptin was not (2) Exercise 2 h after a meal (ME) No differences in healthy, trials but not EM. responsible for lower (3) Exercise 1 h before a meal. (EM) leptin across [210] active, - Exercise lowered hunger hunger ratings in this Exercise was performed at a work rate eliciting trials males ratings prior to meal study. 60% of VO for 50 min. 2max consumption in the EM trial. Halfway through One month of winter field-training exercise at training plasma altitude (2400 – 3650 m). 61 male leptin decreased Fasting blood samples were obtained, 6 times: Leptin, may be a combat by 56%, and (1) baseline; (2) within 24 h of arrival at 7,000-ft NA potential marker of [213] infantry remained base camp; (3) halfway through training;(4) end overtraining. marines suppressed of training; (5) 37 days after completion; and through the end (6) 98 days after completion. training. 08:00 arrival 10 elite - ↓ Hunger scores at 11:00 and 1) rest (control) Leptin was not correlated male No differences in 12:00 and AUC for hunger 2) at 09:00 treadmill exercise 105 min at 50% with hunger or energy [188] soccer leptin across during exercise trial VO2max followed by 15 min at 70% VO2max intake. players trials -No difference in energy intake 12:00 Buffet meal provided

226

Table V: Effects of exercise on circulating ghrelin and appetite Subjects Intervention Ghrelin Appetite Notes Reference - No differences in VASA 7 obese - 8:50 am – breakfast responses as a result of exercise in - No changes in ghrelin were males and 7 -10:00 am – 60 min at 50% VO2max on a cycle either the normal or obese males. found between exercise and None [264] normal ergometer vs. 1 h of rest. - Food consumption ↓ at the test rest conditions weight males - 12:00 pm – buffet style meal provided meal by 146kcal in normal weight men, and by 330kcal in obese men. - AUC values for acylated ghrelin were Exercise: correlated with -8am arrival hunger over the first - AUC for the acylated - 9am →1-h run at 72% of VO2max on a 3 h of the exercise ghrelin concentration was -AUC hunger ratings for the 3 h 9 young, treadmill trial (0800–1100). 38% ↓ over the first 3 h, and before the meal (0800–1100), were healthy, -11am meal -There was a [197] 35% ↓ over the full 9 h of the significantly lower for the exercise active males vs. tendency for higher exercise trial compared with trials vs. the control. 9 h of Rest: hunger ratings over the control trial -8am arrival the last 5 h of the -11am meal exercise trial compared with the control trial. Exercise: 9am - 90 min run at 70% of VO2max on a treadmill followed by 8.5 h of rest. 11:30 am – buffet meal - AUC for acylated ghrelin ↓ Hunger and PFC at 0.5, 1 and Exercise induced a 2:30 pm – buffet meal concentration was 40%↓over 9 healthy 1.5 h during exercise vs. control. brief suppression of 6:00pm – buffet meal the first 2.5 h of the exercise [265] young males ↑ Fullness and satisfaction at 0.5 appetite and plasma vs. trial vs. Rest and was 25% ↓ and 1 h during exercise vs. control. acylated ghrelin. 10 h of Rest: over the entire 10 h. 11:30 am – buffet meal 2:30 pm – buffet meal 6:00pm – buffet meal

227

Table V: Effects of exercise on circulating ghrelin and appetite Subjects Intervention Ghrelin Appetite Notes Reference - In men, the acylated ghrelin response to the MTT did not Condition 1 Baseline→ no-exercise + subjects in vary by condition. energy balance - In women, the acylated Condition 2 (BAL)→ 4 days of exercise - Compared with DEF, men had: ↓ -Women had a 18 (9 men, 9 ghrelin response to the MTT (treadmill at 50–65% of V̇ O until 30% of desire to eat, hunger, and PFC bigger increase in women) 2peak was the highest after DEF, total daily EE was expended) with added energy ratings during BAL. acylated ghrelin after overweight/ then BAL, with baseline [263] to maintain EB. - Women reported no difference in DEF and BAL, than obese being the lowest. On day 5 a ‘meal tolerance test (MTT) was done’ any subjective appetite measures men. individuals - Total ghrelin concentrations Condition 3 (DEF)→ 4 days of exercise (same as across trials. and AUC higher after DEF above) without added energy to induce ED. and BAL compared with On day 5 - MTT baseline.

- Total ghrelin (TG) concentrations in OW were ↓prior to exercise compared with NW at all time points. - TG was unaffected by exercise - Acylated Ghrelin (AG) was ↓in OW compared to NW before exercise at time points - Both AG and DG 17 Normal 120 and 180 min. - 5 consecutive, daily 60 min sessions of aerobic - NW boys had ↑ hunger, desire to correlated positively weight (NW) -The liquid meal ↓AG exercise (walking or jogging on treadmill, eat, and PFC than OW boys before with markers of and 17 concentrations in all. cycling, or step climber) at 65-75% of the and after exercise. appetite. Overweight - Exercise ↑AUC for AG for [266] maximum heart rate reserve. -Exercise led to decreases in - Alternately TG (OW) the first hour. - Liquid Meal (Post) fullness in all boys. could not predict adolescent - Exercise ↑AG changes in appetite. boys concentrations more in fasted

NW compared with fasted OW. - Desacyl Ghrelin (DG) was ↓in OW compared to NW before exercise. - Exercise led to ↓DG concentrations in NW boys, and ↑DG concentrations in OW boys

228

Table V: Effects of exercise on circulating ghrelin and appetite Subjects Intervention Ghrelin Appetite Notes Reference - Resistance exercise suppressed acylated ghrelin at - ↓ Hunger scores following both - Resistance exercise trial. 90 min free weight 0.75 h and 1.5 h compared to resistance and aerobic exercise. session (3 sets of 12 reps of 10 different weight- control. -Hunger was suppressed during lifting exercises at 80% of 12 repetitions 11 young, - Aerobic exercise suppressed and after exercise compared with Acylated ghrelin maximum.) Followed by 6.5 h rest. healthy, acylated ghrelin at 0.75 h the control trial. concentrations were - Aerobic exercise trial. 60 min treadmill [201] active, compared to control. - After the first meal there were no not correlated with running at a speed predicted to elicit 70% of males - AUC for acylated ghrelin for differences between trials hunger. maximum oxygen uptake. Followed by 7 h rest. two h prior before eating, - Aerobic exercise resulted in a - Control trial. Participants rested for 8 h. 9am were lower during the greater suppression of hunger than arrival;11am first meal; 2pm second meal resistance exercise trial than resistance exercise at 0.75h and 1h. the control trial. - Exercise increased fasting - ↓ Hunger scores following the ghrelin in EM group. test meal in the M and ME trials (1) Meal consumption (M) 12 - Following the meal, ghrelin but not EM. (2) Exercise 2 h after a meal (ME) healthy, decreased in all groups. - Exercise lowered hunger ratings None (3) Exercise 1 h before a meal. (EM) [210] active, - Ghrelin levels were higher prior to meal consumption in the Exercise was performed at a work rate eliciting males immediately after exercise in EM trial. 60% of VO for 50 min. 2max the ME trial

- ↓ hunger and PFC scores and ↑fullness and satisfaction scores 09:00 arrival and breakfast bar standardized to during and immediately after body mass was consumed swimming - ↑ hunger and PFC and ↓ fullness Acylated ghrelin 14 -↓ acylated ghrelin during 10:00-11:00 intermittent swimming (6 repetitions and satisfaction from 3.5-8 h in EX concentrations were healthy, swimming and after of 7 min moderate intensity swimming, 3 min trial vs. SED not correlated with [123] active, consumption of the morning rest) followed by 6 h of rest (EX) - no difference in energy or energy intake or any males buffet meal -or rest (SED) macronutrient intake between appetite markers. conditions during either buffet At 12:00 and 16:30 a buffet meal was provided meal. -Relative energy intake lower during EX trial

229

Table V: Effects of exercise on circulating ghrelin and appetite Subjects Intervention Ghrelin Appetite Notes Reference 08:00 arrival Group A = 7 Group A (SED) - resting 30 min (2 men, 5 Group A (LO) - cycle 30 min at 50 W - The AUC for ghrelin during - Exercise intensity had no effect women) Group A (HI) - cycle 30 min at 100 W LO was significantly higher on energy intake or VAS for any healthy 8:45 sandwich consumed until satiated. than SED. appetite measure. No difference

individuals between SED, LO and HI. 08:00 arrival None [186] Group B = 7 Group B (SED) - resting 30 min - No differences between in -Exercising for 120 min (L) (4 men, 3 Group B (S) - cycle 30 min at 50 W ghrelin concentration between resulted in higher energy intake women) Group B (M) - cycle 60 min at 50 W SED, S, M, and L than SED, S, and M healthy Group B (L) - cycle 120 min at 50 W individuals Sandwich provided 15 min following exercise completion (consumed until satiated). - No significant differences in - No significant differences 1) treadmill running at 60% VO2max until 500 15 highly- acylated ghrelin concentration between trials in VAS kcal expended (~46 min) -No sedentary trained between conditions. -↓ Hunger and desire to eat scores [202] 2) treadmill running at 85% VO2max until 500 condition females - ↓ acylated ghrelin following and ↑ satiety and fullness scores kcal expended (~34 min) exercise in both trials following exercise in both trials 1) 60 min of rest Or - No significant

21 healthy 2) exercised on a cycle ergometer at 70% correlations were -- No significant differences males (11) VO2max until they expended 30% of their daily - No significant differences found between [132] in acylated ghrelin and females energy expenditure (~975kcal in ~82min for between trials in VAS energy intake, concentration between sexes (10) men, and ~713kcal in ~84min for women) appetite hormones, or between conditions. 40 min after completion participants were and VAS ratings. presented with a buffet meal 08:45 arrival -↓subjective appetite suppressed Breakfast providing 30% of estimated daily - ↑acylated ghrelin immediately following sprint energy needs concentrations during the exercise but ↑during the afternoon 1) Rest control trial vs. both exercise of the sprint trial. 2) Continuous Endurance Exercise: on a cycle trials (endurance and sprint) - No significant differences ergometer at 65% of VO2max (from 10:45- -↓acylated ghrelin between trials in AUC of VAS 12 healthy Correlations between [190] 11:45) concentrations immediately although trends towards appetite males variables are weak. 3) Sprint Interval Exercise: on a cycle ergometer after exercise and 45 min after suppression were found. six 30-s sprints against 7.5% of body mass with 4 exercise in sprint vs. - No significant differences in min of rest in between sprints were performed endurance trial energy intake between trials (from 11:15-11:45) -↓AUC for acylated ghrelin in -Relative energy intake was lower 12:30 cold buffet lunch provided both exercise trials vs. control in the endurance exercise trial vs. 16:00 hot buffet lunch provided control. 230

Table V: Effects of exercise on circulating ghrelin and appetite Subjects Intervention Ghrelin Appetite Notes Reference 08:00 arrival 1) rest -↓in plasma acylated ghrelin 2) Exercised on a cycle ergometer (60 rpm. 3 sets concentrations during and for -↓ Hunger scores during and for 15 15 healthy of 10 min with 5 min rest in between 30 min following both min following both exercise trials None [198] males ~64% VO2max (EE=288kcal) exercise trials as compared to as compared to control. 3) Rope skipping - 120 skips per min for 10 min. control. 3 sets with 5 min rest in between (EE=295kcal 08:00 arrival -600 kcal breakfast - No effect of condition on 1) Rest (control) hunger and fullness ratings. 2) 09:00 - high-intensity intermittent cycling 12 - ↓ Acylated ghrelin and ↓ - ↓ Desire to eat ratings in the (HIIC) 8s all out sprints with 12 s in between to overweight/o AUC for acylated ghrelin in exercise conditions compared with expend 250kcal (~18min) bese males the MICC and HIIC rest, during the None [207] 3) 09:00 - moderate-intensity (5) and conditions but not S-HIIC exercise period. continuous cycling(MICC) at 70% max HR to females (7) when compared with control. -No difference in absolute energy expend 250kcal (~27min) intake between conditions. 4) 09:00 - short-duration HIIC (S-HIIC) to

expend 125kcal (~9 min) 11:00 buffet style lunch 08:00 arrival 1) Rest (control) - No significant - ↓ Acylated ghrelin in both 2) moderate intensity running - 50% VO2max to correlations were exercise conditions compared 9 healthy expend 600kcal (~55min) - No effect of condition on found between to control males 3) vigorous intensity running - 75% VO2max to Hunger rating changes hunger and - ↓ AUC for acylated ghrelin expend 600kcal (~36min) changes in acylated in vigorous vs. control. At 11:00 a standardized meal was provided (14.3 ghrelin [204] kcal/kg body mass) 08:00 arrival - Hunger ratings were 15%↓in - No significant - ↓ Acylated ghrelin and ↓ 1) Rest (control) EX45and 20%↓ in EX90 vs. correlations were AUC for acylated ghrelin in 9 healthy 2) 45min running -(EX45) 70% VO2max control. found between both exercise conditions males 3) 90 min running (EX90) - 70% -↓ in AUC for hunger by 14% and changes hunger and compared to control At 10:00 and at 14:00 standardized meals were 18% for EX45 and EX90 changes in acylated

provided (11 kcal/kg body mass) respectively vs. control. ghrelin

231

Table V: Effects of exercise on circulating ghrelin and appetite Subjects Intervention Ghrelin Appetite Notes Reference 47 - No significant differences in participants - acylated ghrelin concentration There were no 22 lean (11 09:00 arrival between conditions. - ↓appetite rating (composite score significant male, 11 - Rest (control) - ↓ Desacylated ghrelin during for all 4 indices) in the exercise correlations between female) and - 60 min of treadmill exercise at 60% V̇ O2max the exercise conditions, trial vs. control. appetite ratings, [131] 25 -10:30 standardized breakfast (643kcal for males, compared with control (more -No difference in absolute energy hormone overweight/o 578kcal for females) pronounces in intake between conditions. concentrations or bese (14 -14:00 buffet meal provided overweight/obese energy intake. male, 11 individuals). female) - Baseline acylated ghrelin concentration 08:00 arrival - ↓ Hunger scores at 11:00 and was correlated with 10 elite male 1) rest (control) 12:00 and AUC for hunger during hunger sensation. - ↓AUC for acylated ghrelin soccer 2) at 09:00 treadmill exercise 105 min at 50% exercise trial - Immediately after [188] during exercise vs. control players VO2max followed by 15 min at 70% VO2max exercise a significant 12:00 Buffet meal provided -No difference in energy intake correlation between acylated ghrelin and hunger emerged. - ↓ Hunger, appetite, and prospective eating scores There were no immediately after exercise in 07:00 arrival -↓acylated ghrelin due to significant 23 male exercise vs. control. 1) Rest (control) exercise. correlations between endurance - ↑in satiety scores at all time [194] 2) 20 km run (~78 min) (exercise) -↓acylated ghrelin in exercise hormone runners points following exercise in 30 min after run buffet meal provided vs. control. concentrations and exercise vs. control energy intake. -↓energy intake in exercise vs. control.

232

Table VI: Effects of exercise on circulating GLP-1 and appetite Subjects Intervention GLP-1 Appetite Notes Reference - No differences in - ↑in GLP-1 in normal and VASA responses as a obese males was similar, - 8:50 am - breakfast result of exercise in but EI post exercise was -10:00 am – 60 min at 50% ↑ in mean AUC values for either the normal or lower in obese males 7 obese males and 7 VO2max on a cycle ergometer vs. GLP-1 in the exercise trial obese males. resulting in a negative EB [264] normal weight males 60 min of rest. compared to the resting trial in - Food consumption ↓ - Increased gut hormones as - 12:00 pm – buffet style meal all participants at the test meal by a result of a single bout of provided 146kcal in normal exercise are not completely weight men, and by responsible for differences 330kcal in obese men. in EI - 9:30am arrival -↓ Energy intake after - 9:50am breakfast the high intensity and - 11:00am either cycling exercise moderate intensity at: exercise compared with -There was a significant - Both moderate and high 10 young healthy 1) 75% VO (high intensity rest correlation between GLP-1 2 max intensity exercise led to ↑ males session), - ↓Hunger scores and the amount of energy [196] GLP-1 levels during 2) 50% VO (moderate intensity during and after ingested. 2 max after exercise. session), or exercise 3) Rest (resting session) for 30 min. - No other differences -12:00pm a buffet style test meal in VASA responses was presented. between conditions -Fasting GLP-1 levels are lower in OW compared to NW - NW boys had ↑ - Liquid Meal (Pre) boys hunger, desire to eat, - 5 consecutive, daily 60 min There was no correlation - AUC for GLP-1 and PFC than OW boys 17 Normal weight sessions of aerobic exercise between GLP-1 concentrations was 25% lower before and after (NW) and 17 (walking or jogging on treadmill, concentrations and any of in OW compared to NW exercise. [267] Overweight (OW) cycling, or step climber) at 65-75% the markers of hunger or adolescents. -Exercise led to adolescent boys of the maximum heart rate reserve. satiety. - GLP-1 concentrations rose decreases in fullness in - Liquid Meal (Post) following the test meal all boys.

reaching maximal values at 30 min.

233

Table VI: Effects of exercise on circulating GLP-1 and appetite Subjects Intervention GLP-1 Appetite Notes Reference - No significant 1) treadmill running at 60% - No significant differences in differences between VO2max until 500 kcal expended GLP-1concentration between trials in VAS 15 highly-trained (~46 min) conditions. -↓ Hunger and desire to -No sedentary condition [202] females 2) treadmill running at 85% - ↑ GLP-1 following exercise eat scores and ↑ satiety VO2max until 500 kcal expended in both trials and fullness scores (~34 min) following exercise in both trials 08:00 arrival 1) rest 2) Exercised on a cycle ergometer -↓ Hunger scores

(60 rpm. 3 sets of 10 min with 5 during and for 15 min - No significant differences 15 healthy males min rest in between following both exercise None [198] between trials for GLP-1 ~64% VO2max (EE=288kcal) trials as compared to during or following exercise. 3) Rope skipping - 120 skips per control. min for 10 min. 3 sets with 5 min rest in between (EE=295kcal 08:00 arrival -600 kcal breakfast - No effect of condition 1) Control on hunger and fullness 2) 09:00 - high-intensity ratings.

intermittent cycling (HIIC) 8s all - ↓ Desire to eat ratings - ↑GLP-1 plasma levels during out sprints with 12 s in between to in the exercise 12 overweight/obese the exercise conditions, expend 250kcal (~18min) conditions compared males (5) and compared with control. None [207] 3) 09:00 - moderate-intensity with rest, during the females (7) - No significant differences in continuous cycling(MICC) at 70% exercise period. AUC for GLP-1concentration max HR to expend 250kcal -No difference in between conditions. (~27min) absolute energy intake 4) 09:00 - short-duration HIIC (S- between conditions. HIIC) to expend 125kcal (~9 min) 11:00 buffet style lunch

234

Table VI: Effects of exercise on circulating GLP-1 and appetite Subjects Intervention GLP-1 Appetite Notes Reference 09:00 arrival - ↓appetite rating - Rest (control) (composite score for all 47 participants - 22 - ↑GLP-1 plasma levels during There were no significant - 60 min of treadmill exercise at 4 indices) in the lean (11 male, 11 the exercise conditions, correlations between 60% V̇ O2max exercise trial vs. female) and 25 compared with control (more appetite ratings, hormone [131] -10:30 standardized breakfast control. overweight/obese (14 pronounces in concentrations or energy (643kcal for males, 578kcal for -No difference in male, 11 female) overweight/obese individuals). intake. females) absolute energy intake -14:00 buffet meal provided between conditions.

235

Table VII: Effects of exercise on circulating PYY and appetite Subjects Intervention PYY Appetite Notes Reference - No differences in VASA - ↑in PYY in normal and obese - 8:50 am – breakfast responses as a result of exercise males was similar, but EI post ↑Mean AUC values for PYY in the -10:00 am – 60 min at 50% in either the normal or obese exercise was lower in obese 7 obese males exercise session compared to the VO2max on a cycle males. males resulting in a negative EB and 7 normal resting session in both obese and [264] ergometer vs. 60 min of rest. - Food consumption ↓ at the test - Increased gut hormones as a weight males normal weight males. - 12:00 pm – buffet style meal by 146kcal in normal weight result of a single bout of exercise

meal provided men, and by 330kcal in obese are not completely responsible for men. differences in EI - 9:30am arrival - 9:50am breakfast - 11:00am either cycling - ↑ PYY levels during exercise exercise at: 3–36 -↓ Energy intake after the high -↑ PYY levels were maintained 1) 75% VO (high 3–36 intensity and moderate intensity - No correlations were observed 2 max after the high intensity exercise but 10 young intensity exercise compared with rest between the total increase in PYY not after the moderate exercise healthy males session), - ↓Hunger scores during and after levels and the decrease in the [196] - the mean AUC values observed 2) 50% VO (moderate exercise amount of energy ingested 2 max for PYY depend on exercise intensity session), or 3–36 - No other differences in VASA intensity (↑ intensity = ↑ 3) Rest (resting session) for responses between conditions concentration) 30 min. -12:00pm a buffet style test meal was presented. - 7:00 am arrival. - 7:15 am → either: 1) rest (CON) or - ↑ PYY following breakfast. 2) running on a treadmill for -Energy intake in NEUT was - ↑ PYY, in HEAT compared with higher than CON. 11 young, 40 min at 70% O2peak in CON before the breakfast meal. - Relative energy intake (taking healthy, active (HEAT) → 36°C, with 30% - After the breakfast meal, PYY None [233] into account the energy expended males RH or concentrations in HEAT were during exercise) during HEAT 3) (NEUT) at 25°C with significantly higher than both CON was lower than CON. 30% RH and NEUT. - 8:30 am buffet-type breakfast was provided for 30 min.

236

TableVII: Effects of exercise on circulating PYY and appetite Subjects Intervention PYY Appetite Notes Reference - Resistance exercise trial. 90 min free weight session (3 sets of 12 -↓ Hunger scores following both - ↑ PYY during the aerobic reps of 10 different weight-lifting resistance and aerobic exercise. exercise trial, compared to both exercises at 80% of 12 repetitions -Hunger was suppressed during and the control and resistance 11 young, maximum.) Followed by6.5 h rest. after exercise compared with the exercise trials. These findings healthy, - Aerobic exercise trial. 60 min control trial. PYY concentrations were were confirmed when [201] active, treadmill running at a speed predicted - After the first meal there were no not correlated with hunger. analyzing AUC values. males to elicit 70% of maximum oxygen differences between trials - No differences between the uptake. Followed by 7 h rest. -Aerobic exercise resulted in a control and resistance exercise - Control trial. Participants rested for greater suppression of hunger than trials. 8 h.9am arrival;11am first meal; 2pm resistance exercise at 0.75h and 1h. second meal - ↓ PYY3-36 during the EM trial compared to M and ME trials - ↓ Hunger scores following the test - PYY increased prior to meal consumption. meal in the M and ME trials but not 3-36 (1) Meal consumption (M) following food intake, 12 - ↓ PYY after intake for M EM. (2) Exercise 2 h after a meal (ME) 3-36 reaching a peak within an healthy, compared to the other trials. - Exercise lowered hunger ratings (3) Exercise 1 h before a meal. (EM) hour for all 3 conditions. [210] active, - PYY remained higher than prior to meal consumption in the EM Exercise was performed at a work rate 3-36 - At 7 h all PYY males fasting for up to 3h after food trial. 3-36 eliciting 60% of VO for 50 min. concentrations were 2max intake in the M trial and up to similar. 7h after food intake in the EM trial. - 7:00 am arrival - Nutritional beverage (250kcal) was -No significant differences in hunger provided every 2 h (1500kcal total). or satiety scores over 12 h. 1) rest (SED) or -When assessing VAS by 2 h periods, 11 obese 2) 1 h walking from 7:05-8:05 at 60- - No significant differences in INT hunger scores were lower than participants (8 65% of VO peak (EX) PYY concentration between EX during 13:00-15:00h, and lower None [117] men, 3 2 3) 12 5 min bouts at 60-65% of conditions than EX and SED during 15:00- women) VO2peak performed intermittently 17:00. INT satiety scores were higher every hour (INT) than EX and SED during 13:00-15:00 - Nutritional beverage(250kcal) was and 15:00-17:00 provided every 2 h (1500kcal total)

237

Table VII: Effects of exercise on circulating PYY and appetite Subjects Intervention PYY Appetite Notes Reference - No significant differences in - No significant differences 1) treadmill running at 60% VO2max until PYY concentration between between trials in VAS 15 highly- 500 kcal expended (~46 min) conditions. -↓ Hunger and desire to eat trained -No sedentary condition [202] 2) treadmill running at 85% VO2max until - ↑ PYY following exercise in scores and ↑ satiety and females 500 kcal expended (~34 min) both fullness scores following exercise in both trials -↓subjective appetite 08:45 arrival immediately following sprint Breakfast providing 30% of estimated exercise but ↑during the daily energy needs afternoon of the sprint trial. 1) Rest - No significant differences 2) Continuous Endurance Exercise: on a between trials in AUC of VAS cycle ergometer at 65% of VO2max (from - ↑PYY concentrations 12 healthy although trends towards Correlations between 10:45-11:45) immediately after exercise in the [190] males appetite suppression were variables are weak. 3) Sprint Interval Exercise: on a cycle endurance exercise trial vs. found. ergometer six 30-s sprints against 7.5% of control. - No significant differences in body mass with 4 min of rest in between energy intake between trials sprints were performed (from 11:15-11:45) -Relative energy intake was 12:30 cold buffet lunch provided lower in the endurance 16:00 hot buffet lunch provided exercise trial vs. control. 1) 60 min of rest - In men, PYY was not Or significantly different between 2) exercised on a cycle ergometer at 70% 21 healthy conditions. - No significant correlations VO2max until they expended 30% of their males (11) - No significant differences were found between energy daily energy expenditure (~975kcal in [132] and females - In women, ↑ PYY at the end and between trials in VAS intake, appetite hormones, ~82min for men, and ~713kcal in ~84min (10) 15 min after exercise as compared and VAS ratings. for women) to rest with no difference at 30 40 min after completion participants were min between conditions. presented with a buffet meal 08:00 arrival 1) rest - ↑ PYY concentrations 2) Exercised on a cycle ergometer (60 rpm. -↓ Hunger scores during and immediately following exercise in 15 healthy 3 sets of 10 min with 5 min rest in between for 15 min following both both exercise trials that None [198] males ~64% VO2max (EE=288kcal) exercise trials as compared to disappeared by 30 min post 3) Rope skipping - 120 skips per min for control. exercise. 10 min. 3 sets with 5 min rest in between (EE=295kcal

238

Table VII: Effects of exercise on circulating PYY and appetite Subjects Intervention PYY Appetite Notes Reference 08:00 arrival -600 kcal breakfast 1) Control - No effect of condition on 2) 09:00 - high-intensity hunger and fullness intermittent cycling (HIIC) 8s ratings. all out sprints with 12 s in - ↓ Desire to eat ratings in between to expend 250kcal the exercise conditions 12 overweight/obese males (~18min) - No significant differences in compared with rest, during None [207] (5) and females (7) 3) 09:00 -moderate-intensity PYY concentration between the continuous cycling(MICC) at conditions. exercise period. 70% max HR to expend 250kcal -No difference in absolute (~27min) energy intake between 4) 09:00 - short-duration HIIC conditions. (S-HIIC) to expend 125kcal (~9 min) 11:00 buffet style lunch 09:00 arrival - ↓appetite rating - Rest (control) There were no significant (composite score for all 4 47 participants - 22 lean (11 - 60 min of treadmill exercise at - ↑PYY plasma levels during correlations between appetite indices) in the exercise male, 11 female) and 25 60% V̇ O2max the exercise conditions, ratings, hormone trial vs. control. [131] overweight/obese (14 male, -10:30 standardized breakfast compared with control (more concentrations or energy -No difference in absolute 11 female) (643kcal for males, 578kcal for pronounced in lean intake. energy intake between females) individuals). conditions. -14:00 buffet meal provided - ↓ Hunger, appetite, and prospective eating scores 07:00 arrival immediately after exercise 1) Rest (control) in exercise vs. control. There were no significant 2) 20 km run (~78 min) - No difference in PYY - ↑in satiety scores at all correlations between 23 male endurance runners [194] (exercise) between exercise and control. time points following hormone concentrations and 30 min after run buffet meal exercise in exercise vs. energy intake. provided control -↓energy intake in exercise vs. control.

239

Table VIII: Effects of exercise on circulating PP and appetite Subjects Intervention PP Appetite Notes Reference - - PP↑ Measurements EX→ 3 h at 40% VO2max on a throughout were only 6 healthy cycle ergometer exercise vs. NA made during [208] males vs. rest where the 3 h of REST→ 3 h of rest. levels stayed exercise. at baseline

- 7:00 am arrival. -Energy intake in - 7:15 am → either: NEUT was higher 1) rest (CON) or than CON. 11 2) running on a treadmill for 40 -Following - Relative energy young, min at 70% O peak in breakfast, PP 2 intake (taking into healthy, (HEAT) → 36°C, with 30% RH concentratio None [233] account the energy active or ns increased. expended during males 3) (NEUT) at 25°C with 30% exercise) during RH HEAT was lower - 8:30 am buffet-type breakfast than CON. was provided for 30 min. 06:30 arrival 1) Rest at 22°C (Rest 22) -No significant 2) Rest at Hot: 31°C (Rest 31) differences in energy 3) ↑ PP during 10 intake between Cycling on a cycle ergometer exercise healthy conditions. for 40 min at 60% V̇ O max at conditions None [192] active 2 - ↓ Relative energy 22°C (Ex 22) vs. rest males intake during exercise 4) Same exercise at 31°C (Ex conditions. conditions than rest 31) conditions. -30 min recovery -buffet style meal provided

Appendix C – Recruitment Poster

Appendix D - Recruitment Email Hello, We are doctoral students (Iva Mandic and Mavra Ahmed) from University of Toronto working under the supervision of Dr. Len Goodman (PhD; Defence Scientist, DRDC Toronto Research Centre), Dr. Ira Jacobs (DrMedSc, Dean of Faculty of Kinesiology and Physical Education, University of Toronto) and Dr. Mary L’Abbe (PhD, Chair of Department of Nutritional Sciences, University of Toronto). We are currently seeking volunteers for a physical activity and nutrition “PAN” study to understand the effects of environmental and physical stress on energy expenditure, energy intake and appetite.

Purpose To look at energy requirements of activities and differences in nutritional intake under simulated environmental conditions. Volunteers We are looking for healthy male and/or female Regular or Reserve Canadian Forces members, 18-60 years old, who are accustomed to being physically active and exercise intensities required on duty. Procedures 11sessions ≈ 58 hours in total involving high-intensity activities, rations consumption and recording, energy expenditure testing and blood/urine sampling in hot, cold and temperate environments Risks Minor discomfort/bruising from needle puncture, nausea, dizziness, shortness of breath, dehydration, muscle cramps, numbing and minimal risk of heart attack from exercise stress Compensation Participants will be remunerated (including travelling costs). Location DRDC, Toronto Research Centre, located at 1133 Sheppard Ave. W. Toronto ON Ethics This study has been reviewed and received ethics clearance through Defence Research and Development Canada and University of Toronto Research Ethics Committees.

IF INTERESTED, PLEASE CONTACT: Iva Mandic – 416-978-8563; [email protected] Mavra Ahmed – 416-564-8924; [email protected]

Thank you for your time and support.

Sincerely, Iva Mandic and Mavra Ahmed

242

Appendix E – Consent Form Participant Information Sheet and Volunteer Consent Form

Title: The effects of environmental and physical stress on energy expenditure, energy intake and appetite. Protocol Number: 2013-075 Principal Investigators: Iva Mandic (PhD Candidate; University of Toronto), Mavra Ahmed (PhD Candidate; University of Toronto) Co-Investigators: Ira Jacobs (DrMedSc; University of Toronto), Mary L’Abbe (PhD; University of Toronto), Len Goodman (PhD; DRDC Toronto), Wendy Sullivan-Kwantes (MA; DRDC Toronto). Run Directors: Doug Saunders, Ingrid Smith, Christina Powesland. WBE: Personnel Portfolio/Diagnostics and Health Protection/04kc Nutrition)

I, ______of ______

(name) (address and phone no.)

hereby volunteer to participate as a participant in the DRDC Toronto/University of Toronto experiment “The effects of environmental and physical stress on energy expenditure, energy intake and appetite.” (DRDCHREC Protocol 2013-075). I have had the opportunity to discuss the protocol explained within this consent form with one of the investigators. All of my questions concerning this study have been fully answered to my satisfaction. I may obtain additional information about the research project and have any questions about this study answered by contacting either Iva Mandic (Ph.D. Candidate) by telephone at (416) 978-8563, or by e-mail ([email protected] or [email protected]) or Mavra Ahmed (Ph.D. Candidate) by telephone at (416) 564-8924 or by e-mail ([email protected] or [email protected]).

Purpose of the Research: To (1) understand the energy requirements of demanding activities conducted in harsh environments (e.g. extreme cold/hot) vs. temperate; (2) come up with a feasible and accurate method of estimating energy expenditure in the field; (3) demonstrate novel methods of estimating energy intake; (4) determine whether nutritional intake is different under simulated hot and cold environments in comparison with temperate climates and sedentary behaviour; (5) assess whether physical and environmental stressors affect appetite and food intake.

Voluntary Participation: Your participation in the study is completely voluntary and you may choose to stop participating at any time. Your decision to not volunteer will not influence the nature of the ongoing relationship you may have with the researchers or study staff or the nature of your relationship with DRDC or the University of Toronto either now, or in the future.

What Will Be Asked of You in this Research:

If you are eligible (found to be able to take part in the study as per the screening process) to participate in this study and if you agree to participate, you will make 11 visits to the laboratory during a

2-3-month period. These visits will be used to provide a study overview, perform a VO2max test (to look at your fitness), gather some basic information about your body composition and baseline values of important aspects of your blood and urine, and then complete an actual test trial for each of 4 different conditions being compared in this study. The details for each visit are as described below:

243

Visit 1 –Review and sign the consent form and associated questionnaires

• Duration of this visit: Approximately 60 minutes • Purpose of this visit: The purpose of this visit is to explain the study to you and answer any questions that you may have regarding this study. • What you will be asked to do in this visit: • After the study has been explained to you, and all your questions are answered (feel free to ask any questions at any point throughout the study), you will be asked to read, review and sign this informed consent form and the invasive procedures consent form if you decide to participate. You will then fill out a PAR-Q+ form (The PAR-Q+ is a short questionnaire that will help to determine whether you are healthy enough to perform the physical exercise protocols used in this study), screening form, a socio-demographic questionnaire, a sleep quality questionnaire (PSQI) and a physical activity pattern questionnaire.

Visit 2 - VO2max Exercise Testing and Baseline Data Collection

This visit will take place in accordance with your schedule within 7 days of your visit 1. We will schedule the day/date of this visit once you have agreed to participate in the study and have signed the consent form in visit 1.

• Duration of this visit: Approximately 60 minutes. • Purpose of this visit: In this study, you will complete a maximal exercise test on a treadmill to determine your VO2max, a measure of your cardio-vascular fitness level. • What you will be asked to do in this visit: • For the VO2max test, you will need to bring clothing you are comfortable exercising in. You will be connected to a metabolic cart during the entire test by breathing from and into a mouthpiece. All of the air you will be breathing will be going through the mouthpiece. You will be on a treadmill and the incline and speed of the treadmill will increase over time. The test will end as soon as you can no longer keep up with the treadmill, or once one of the investigators tells you that the test is over. This takes on average 12-15 minutes. • Next, you will get a smartphone, a food scale and a 3-day food record booklet to record your total food/beverage intake for the following three days. One of the Principal Investigators will show you how to use these devices. You will also receive 24-hour urine collection containers to collect your urine for the following two days. One of the investigators will explain to you how to collect your urine, where to store it in your home until your next visit, and how to transport it. Urine samples will be collected for the measurement of urinary free cortisol, urea, potassium and sodium. Detailed instructions on how to collect your urine will be provided to you.

Visit 3 - Blood sample and Bod Pod

This visit will take place three days after visit 2. We will schedule the day/date for this visit during your visit 2.

• Duration of this visit: Approximately 60 minutes • Purpose of this visit: The purpose of this visit to take a blood sample to look at some nutrients as well as do a body composition test.

244

• What you will be asked to do in this visit: • You will be asked to report to the testing facility (DRDC Toronto Centre) having had your last meal at least 10 hours prior to arrival. • One fasting blood sample (approximately 13 ml) will be taken from you with a needle. These samples will be used to measure nutrients. These include cortisol, serum ascorbic acid, urea, potassium, sodium, and lipids. • Your body fat percentage will be measured with the BOD POD TM device. For this test, you will be required to wear tight-fitting clothing and a cap (provided by the researcher). You will sit inside a hollow chamber for approximately two minutes. Female participants will be required to wear a spandex sport bra top (provided by the researcher). The tight-fitting clothing and cap helps to remove as much trapped air as possible, increasing the accuracy of the test. You will be able to choose the gender of the person running the test for your comfort. The provided clothing and cap will be washed after each use. Before the test, one of the investigators will explain the procedure to you. • In this visit you will also be assigned and given either a smartphone or a food scale and a 3-day food record booklet to record your total food/beverage intake for particular days throughout the study as described below.

Visits 4-5 (Condition 1), 6-7 (Condition 2), 8-9 (Condition 3) and 10-11 (Condition 4) –Testing Days

These visits will take place in accordance with your schedule and within a three-month time period. We will schedule the day/date for these visits in visit 3.

Description of the conditions:

• There will be four (4) testing conditions: a. Hot: the chamber will be maintained at either +30 °C with relative humidity of 30% b. Cold: the chamber will be maintained at either -10 °C with relative humidity of 30% c. Temperate: the chamber will be maintained at either +21 °C with relative humidity of 30% d. Rest: the same environmental condition as Temperate, however you will be resting for the entire time.

Assignment of the conditions:

• The above listed conditions will be assigned to you in a randomized order and you will only complete each condition once. Each testing condition consists of two visits, that must be on consecutive days and each of the four testing conditions will be separated by a minimum of 1 week. Visits 4, 6, 8 and 10 will take approximately 10 hours to complete and visits 5, 7, 9 and 11 will take approximately 60 minutes to complete. Depending on the testing condition you will be completing on that day, you will be asked to bring your appropriate military clothing with you (standard combat dress, hot weather clothing etc.) (temperature appropriate military garb).

Preparatory data collected at home:

• For two (2) days prior to visits 4, 6, 8, and 10, you will be given military rations to consume for those two days (2 breakfast, 2 lunch and 2 dinner rations will be provided). • You will be asked to document your food/beverage intake (using the device given to you during visit 3) and consume only the provided military rations. • One (1) day prior to visit 4, 6, 8, and 10 and during visit 4, 6, 8, and 10, you will also collect your urine in 24-hour urine collection containers. Urine samples will be collected for the measurement of urinary free cortisol, urea, potassium and sodium. Detailed instructions on how to collect your urine will be provided to you. 245

• Military rations, food scale and urine containers for your preparatory data collected at home prior to each testing conditions will be provided to you on visit 3, 5, 7 and 9.

Visit 4–Testing Condition 1- A

• Duration of this visit: Approximately ten (10) hours. • Purpose of this visit: The purpose of this visit is to assess your energy output and intake during one of the above listed testing conditions. • What you will be asked to do in this visit: • You will consume a temperature capsule at 6:00am with 250ml of water. • You will be asked to fast prior to your arrival at the lab at 7:45 am. You will begin fasting at 10pm the previous night (an overnight fast) and on the morning of your visit, you will not consume any water or other liquids (other than the 250 ml of water to ingest the temperature capsule) prior to your arrival to the lab. • Your body weight, blood pressure and resting heart rate will be measured. • A catheter will be inserted into your arm and a blood sample will be drawn. The catheter will remain in your arm for 8 hours. You will be asked to provide blood samples at various time points for the 8-hour period for a total of 9 blood samples throughout the day (approximately 73 ml). • You will be outfitted with accelerometers (devices that estimate energy expenditure, similar to those in pedometers, phones apps, and other mobile devices), heart rate monitors and a portable metabolic measurement system (Meta-Max Inc.). The metabolic measurement system measures your energy expenditure by measuring the amount of oxygen and carbon dioxide you breathe out. • You will be asked to change into your temperature appropriate military garb. You will be given 3 military rations (1 breakfast, 1 lunch, and 1 dinner ration). In addition, you will be allowed to drink as much water as you like. We will set up water stations during your testing conditions. You can eat as much or as little of the rations as you would like. You will be asked to record all of your food and beverage intakes using the 3 Day food record with a food scale or the smart-phone app provided to you during your second visit. • You will be required to enter the environmental chamber (carrying all of the food and equipment with you). The chamber will be maintained at either +30 °C with relative humidity of 30% (Hot), +21 °C with relative humidity of 30% (Temperate and Rest conditions), -10 °C with relative humidity of 30% (Cold). • For the rest testing condition, you will rest for 8 hours in the chamber. During the rest condition, you will be provided with movies to watch and reading materials to occupy your time. • For the hot, cold and temperate testing conditions, you will complete two 2-hour activity circuits composed of typical military tasks (covering a range of light, moderate and heavy work rates) dispersed with two bouts of 2-hour rest. During the activity circuits, you will be asked to complete the following:

246

Block Time Activity Sandbag Lift – lift a 20 kg sandbag from the floor above a height of 91.5 cm. Alternates 8:00-8:10 between left and right sandbags separated by 1.25 m. (7 min + 3 min rest) Sandbag passing – Move sandbags from one 91.5cm high table to another –at a pace of 8:10-8:20 4 sandbags per min. (7 min + 3 min rest) Sandbag stacking and tamping – Stack and secure sandbags in an organized manner to 8:20-8:30 build fortifications (7 min + 3 min rest) Walk on hard surface, level ground at 3.6 km/h (1 m/s) with loaded 24.5kg military 8:30-8:40 rucksack. (7 min + 3 min rest) Stretcher carry - Walk on hard surface, level ground at 3.6 km/h (1 m/s) with 40kg 1 8:40-8:50 barbell. (7 min + 3 min rest) (ACTIVE) Simulation of ‘Escape to cover’– Run in the spot as fast as possible for 20 seconds drop 8:50-9:00 to a prone position for 10 seconds. Repeat 10 times. (5 min + 5 min rest) 9:00-9:10 Stand in one spot. (10 min) 9:10-9:15 Stoop/kneel. (5 min) Thumper simulation - Use a 20 kg thumper to hit a designated target (to simulate 9:15-9:25 driving in pickets or CAMOFLET tubes). (7 min + 3 min rest) Walk on hard surface, level ground at 3.6 km/h (1 m/s) with 0, 10, or 20 kg load. 10 min 9:25-9:59 carrying each load in the hands as jerry cans with 2 min of rest in between each load. (34 min) 10:00-11:00 Complete surveys – (60 min) 2(REST) 11:00-12:00 Rest/ Meal preparation/consumption (60 min) Vehicle extrication simulation– Move half of an 86 kg dummy (a mannequin that can be 12:00-12:10 separated into 2 parts) from a table to the other side of the chamber. (7 min + 3 min rest) Walk on hard surface, level ground at 5.6 km/h (1.56 m/s) with a 0, 10, or 20 kg load. 12:10-12:44 10 min carrying each load in the hands as jerry cans with 2 min of rest in between each load. (34 min) Sandbag Drag – Drag four sandbags on the floor from one side of the chamber to the 12:45-12:55 other. (7 min + 3 min rest) 3 12:55- 13:00 Sandbag Pull – Pull on a rope attached to a sandbag. (7 min + 3 min rest) (ACTIVE) Dummy Drag - Drag half of an 86 kg dummy (a mannequin that can be separated into 2 13:00-13:10 parts) from one side of the chamber to the other. (7 min + 3 min rest) 13:10-13:15 Sit in one spot. (5 min) Digging - Using a standard shovel, move stones from one pail to another. (7 min + 3 13:15-13:25 min rest) Walk on hard surface 5% grade at 3.6 km/h (1 m/s) with 0, 10, or 20 kg load. 10 min 13:25-14:00 carrying each load in the hands as jerry cans with 2 min of rest in between each load. (34 min) 14:00-15:00 Complete surveys – (60 min) 4(REST) 15:00-16:00 Rest/ Meal preparation/consumption (60 min)

• During this visit, you will also be provided with a 24h urine container to collect your urine as the visit progresses. You are being asked to collect your urine in order to measure the concentration of some nutrients in your urine for the purposes of accurately detailing your energy intake. For the purposes of this study, we encourage you to do a complete void pre-trial. However, if required, we will provide an enclosed privacy chamber for you to collect your urine during your testing session. • You will be asked how hungry you feel at various points throughout the 8-hour period and we will record the responses on a survey. • You will complete a mood questionnaire. • Once you have completed the allotted time, you will exit the chamber. • You will be provided with a urine sample collection container and collect your urine until the next morning. • You will take all unfinished rations with you and record your food/beverage intake until 10pm that night using the recording devices provided. Do not eat anything other than what is left over 247

from the rations. You will be asked to stop eating after 10pm that night in order to begin your overnight fast.

Visit 5–Testing Condition 1- B

• Duration of this visit: Approximately 60 minutes. • Purpose of this visit: The purpose of this visit is to collect post-testing condition data as discussed below. • What you will be asked to do in this visit: • During this visit you will report to the lab at 8:00am the next day (the day after Visit 4) with your urine collection container, all unfinished food and/or food/beverage waste, and your food records (as documented by the devices given to you). • You will be asked to provide a 13 ml blood sample. • You will complete a knowledge, attitude and behaviour questionnaire regarding the electronic food recording device or the paper food record provided to you as well as complete a food satisfaction questionnaire.

Visits 6-7 (Condition 2), 8-9 (Condition 3) and 10-11 (Condition 4) –Testing Days

These visits will take place in accordance with your schedule and within a three-month time period. We will book the day/date for these visits in visit 3. The testing conditions will be separated by a minimum of 1 week. Visits 6 and 7 → Testing Condition 2 Visits 8 and 9 → Testing Condition 3 Visits 10 and 11 → Testing Condition 4 The procedures for Testing Conditions 2, 3, and 4 will be identical to Testing Condition 1, with the following exception: you will be asked to complete a different trial. If you completed a hot environmental trial in Testing Condition 1, you will complete either a temperate, cold, or resting trial during Testing Condition 2. If you completed a resting trial during Condition 2, you will then complete a temperate or cold trial on Testing Condition 3, and so on. Whichever condition you are not exposed to in Testing Conditions 1, 2, and 3, will be the condition you undergo in Testing Condition 4.

Biological Samples:

Throughout this study, you will provide 41 blood samples in total. During Visits 4, 6, 8 and 10, nine (9) blood samples (73 ml) will be taken from you in order to measure key blood values. These samples will be taken with a catheter. During Visits 3, 5, 7, 9, and 11, one (1) blood sample (13 ml) will be collected using a needle. By signing this consent form, you are agreeing to the collection of a total blood volume of 357ml (~24 tablespoons) over the two months the testing will last. In the event that that you do not complete the testing conditions (hot, cold, temperate and neutral) within three months and choose to complete the study, an additional 13 ml of blood will be collected prior to the final testing conditions in order to confirm that your serum ascorbic acid, urea, sodium, potassium, creatinine, cholesterol and lipid levels have not drifted from those found in the blood sample collected during session 3.

248

You will also be providing 24-hour urine samples in order to measure urinary cortisol, urea, sodium and potassium. The instructions for urine collection will be provided to you by one of the principal investigators.

By volunteering to participate in this study you are indicating that you have been told and agree that blood and urine will be obtained as described above. These samples will be used to measure appetite-regulating hormones and nutrients as well as other factors naturally occurring in your blood and urine. The specific measures include leptin, acylated ghrelin, glucagon-like peptide-1 (GLP-1), polypeptide YY (PYY), urea, creatinine, sodium, potassium, lipids, vitamin C and cortisol. These samples are stored frozen in the lab at Defense Research and Development Canada with your study identification code. Once the samples are analyzed all remaining samples will be destroyed. The analyses of your biological samples will be conducted by members of the research team using kits and the remaining samples will be sent to LifeLabs Medical Laboratory Services with your study identification code.

Health Assessment:

Your body weight, resting heart rate and resting blood pressure will be measured on each visit to the laboratory. During each visit to the lab, you will be asked questions to find out if you have done anything between visits that would prevent you continuing in the study such as: taking prescription drugs, smoking, donating blood, etc. You hereby agree to provide responses to questions that are to the best of your knowledge truthful and complete. You agree to advise the investigators of any health status changes since your initial assessment (including but not limited to minor illnesses, new prescription or 'over-the-counter' medications and new risk of pregnancy). You have been advised that the medical information you reveal will be treated as confidential (‘Protected B’ in accordance with CAF Security Requirements). Moreover, you understand that your experimental data will be protected under the Government Security Policy (GSP) at the appropriate designation (DRDC facility) and not revealed to anyone other than the DRDC-affiliated Investigator(s) or external investigators from the sponsoring agency without your consent except as data unidentified as to source. You are aware of the requirement to sign a separate consent form for invasive medical procedures.

Incidental Medical Findings:

Should an incidental medical finding be detected by qualified personnel, you will be notified of that finding and be advised to seek appropriate medical follow-up.

Time Commitment and Location of the study:

You understand and consent to the time commitment of approximately 58 hours for this study. You will be asked to come to the DRDC Toronto Research Centre laboratory facilitates for all visits and exercise testing.

Restrictions:

Consumption of multivitamins and/or performance enhancers: You are asked to refrain from consumption of any multivitamins, minerals and/or performance enhancers for the purposes of accurately capturing your energy intake requirements from the rations provided to you.

249

Exercise and consumption of alcohol: You are asked to refrain from alcohol consumption and exercise (activities above those of daily living) 24 hours prior to Visits 2-11.

Consumption of caffeine: You are asked to refrain from caffeine consumption 15 hours prior to Visits 2-11.

For Female Participants: You are being informed that this study could be potentially harmful to a fetus and, as such, pregnancy would disqualify you from further participation in this experiment. Therefore, you consent to pregnancy screening and potential counseling by a qualified health care practitioner. The health care practitioner might conclude that further tests are needed so you may be asked to undergo additional testing to determine your pregnancy status. You understand that all discussion pertaining to this matter will be treated as confidential between the health care practitioner and yourself. Also, you are being advised that if you have any concerns regarding a possible pregnancy, you should consult a qualified health care practitioner before undertaking or resuming any phase of this study.

Risks and Discomforts:

VO2maxTesting: You have been told that the principle risks you may experience include dizziness, breathlessness and muscle pain during exercise testing. Muscle stiffness may also be experienced for 2-3 days after the tests if you do not have experience performing high intensity exercise using the upper body & legs. There is a very small risk of death during an exercise test (0.5 per 10,000 tests). All of these risks and side effects will be minimized by having you screened with the use of the PAR-Q+ form and by following standard procedures before and during the test.

A trained and experienced member of the research team who is certified in First Aid/CPR will conduct the VO2max exercise testing. Should an emergency occur during the VO2max exercise testing, laboratory personnel will follow the emergency response plan, which is located in the lab. A phone call to 911 and the on-call physician ((416) 337-5934) will be made. The phone is available within 20 feet of the equipment. An Automated External Defibrillator (AED) will also be present in the laboratory as a safety precaution. A trained and experienced member of the research team will perform and supervise all exercise testing sessions.

Blood Samples: It is possible that you could develop an infection from the blood draw, or experience bruising in that area. These risks are small since standard medical precautions (including: use of sterile needles; properly cleaning the skin area with alcohol prior to drawing the blood; wearing gloves by personnel; following aseptic conditions; no re-using of tubes, needles, or gauze, etc.) will be followed. Blood samples will be drawn by an experienced member of the research team who has been trained to take blood samples.

Blood Loss: It is possible for you to have reduced blood volume over the course of the study. However, the amount of blood taken during the 5-8 week period for the testing visits (visits 3-11) is small (approximately 357ml or about twenty-four tablespoons), and it is unlikely to result in any adverse effects. The collected blood amount is spread over a 5-8 week period and falls under the 450 ml (1 pint) of blood that Canadian Blood Services collects during a blood donation. Considering that 357 ml of blood will be collected during this study, you are asked to not donate blood during, or for 56 days following study completion (the amount of time Canadian Blood Services recommends between blood donations) for their health and safety.

250

Activities in the Chamber: You have been told that you may experience some stiffness in your legs and arms for one or two days after the experiment due to the exercise during conditions hot, cold, altitude and temperate. Chamber activities will be monitored by at least one investigator and one technician at all times during each trial.

Exercise in the Heat: You have been told that the principal risks you may experience during the heat exposure may involve headache, fatigue, weakness, dizziness, nausea, and dehydration, and progress to hyperthermia (elevated body temperature), heat stroke, heat exhaustion or heat cramps. Exercise in the heat is accompanied by sweating. Therefore, you will be asked to drink water during and after the hot session in order to replenish fluid losses. Your core temperature will be monitored throughout the experiment and testing will halt if values reach 39.5ºC. Core temperature will continue to be monitored after the experiment is finished. Once core temperature begins to decrease, this means that the conditions for heat loss exceed those for heat gain and there is no further risk of continued heat storage.

Cold Exposure: You have been told that the principal risks you may experience during the resting periods (between military tasks during the cold condition) in the cold exposure may include decreased temperature of the skin, which may trigger shivering. You will be dressed in the appropriate military clothing meant for cold environments to reduce this risk.

By signing this consent form, you are accepting the above-mentioned risks that are associated with this study. Also, you acknowledge that your participation in this study, or indeed any research, may involve risks that are currently unforeseen by DRDC. You accept these risks.

You have been told that this study is a more than minimal-risk research project and, therefore, several procedures for your safety will be adopted. In the event a First Aid Emergency occurs while you are in the laboratory, a phone call to 911 and/or the on-call physician ((416) 337-5934) will be made, and the phone is available within 20 feet of the equipment. First aid/CPR may be given by one of the trained members of the research team who will be present during all lab visits. In the highly unlikely event that you may become incapacitated during your participation, the emergency medical treatment will be instituted even though you are unable to give your consent at that time. You will go with the Investigator(s) to seek immediate medical attention if either the Investigator(s) or you consider that it is required. Every effort will be made to contact a family member, or the designated person indicated below should that be necessary.

Benefits of the Research and Benefits to Me: Individually, you will learn about various aspects of your fitness levels as well as methods used to assess parameters (e.g heart rate, blood pressure) in exercise physiology labs around the world. You will also learn about your caloric and nutrient intake. You will be provided with information regarding your energy intake, your personal fitness assessment and blood values AFTER you have completed the entire study. If, for any reason, you withdraw in mid-study, the data will be not analyzed and we will not be able to provide you with any information regarding your food/beverage intake, your fitness level or blood values.

Remuneration: As a Canadian Armed Forces member you are entitled to remuneration in the form of stress allowance as follows: $12.72 for Visit 1, $25.44 for Visit 2, $12.72 for Visit 3, $63.37 for Visits 4, 6, 8, and 10, $25.44 for Visits 5, 7, 9, and 11, and $12.72 for each of the 11 days that you will be documenting your food intake and collecting your urine at home. You will be remunerated for any session you complete with the potential to receive a total amount of $546.07 if all sessions are completed. You are entitled to partial remuneration if you do not complete all of the sessions. Stress remuneration is income and is subject to income tax. As a Canadian Forces member, your Service Number (SN) is required for 251

remuneration. Additionally, you will be compensated for any travel costs (TTC fare, gas, etc.) you incur travelling to and from the testing facility.

DRDC Photo/Video Release: The photos and videos are meant to capture the procedures in the study. If you grant us permission to take your photo and/or to videotape you, it will only happen during your time at DRDC Toronto. You can refuse to allow photos and videos to be taken of you at any time during the study without it affecting your participation in the study even if you have given permission.

Release Granting Permission to use digital photo/video images from an Approved Human Research Study in Department of National Defence Publications and Presentations

Initial one:

_____ I hereby grant permission to use digital photo/video images of me from the above study in DND publications and presentations.

_____ I hereby grant permission to use digital photo/video images of me from the above study in DND publications and presentations only if facial features are blurred so that I cannot be recognized in the image.

_____ I hereby do not grant permission to use digital photo/video images of me from the above study in DND publications and presentations.

Secondary Use of Data: I consent/do not consent (circle one) to the use of this study’s experimental data involving me in unidentified form in future related studies, as long as review and approval have been given by both the DRDC Human Research Ethics Committee and Office of Research Ethics at the University of Toronto. If I agree, the results collected from my physiological monitoring and performance tests will be entered into an electronic data base in a form that is anonymous and cannot be linked back to my name. The objective of the database is to facilitate the creation of normative values for the various tests that are used routinely in our research laboratory.

Participant

Signature______Date______

252

Confidentiality: All information you supply during the research will be held in confidence and unless you specifically indicate your consent, your name will not appear in any report or publication of the research. Your name will not be identified or attached in any manner to any publication arising from this study. Moreover, the experimental data may be reviewed by an internal or external audit committee with the understanding that any summary information resulting from such a review will not identify you personally. Data will be collected using several methods, including questionnaires and computers, as well as urine and blood samples. Upon being enrolled in the study, you will receive an individualized study code. Only the study code will be used for identification purposes on any forms or materials collected during the study. Your data will be safely stored in a locked facility at DRDC and only research staff will have access to this information Your experimental data will be protected under the Government Security Policy (GSP) at the appropriate designation and not revealed to anyone other than the DRDC-affiliated Investigator(s) or external investigators from the sponsoring agency without your consent except as data unidentified as to source. Your data files will be stored for at least seven years after the completion of the study after which they will be destroyed by shredding at a commercial document destruction firm. Confidentiality will be provided to the fullest extent possible by law.

Qualifications and Roles of Team Members:

Investigators Iva Mandic, MSc, Doctoral Candidate Faculty of Kinesiology and Physical Education, University of Toronto Tel:416-978-8563 Email: [email protected] [email protected]

Mavra Ahmed, MSc, Doctoral Candidate Dept. of Nutritional Sciences, University of Toronto Tel: 416-564-8924 E-mail: [email protected] [email protected]

Project Managers: Dr. Len Goodman (Phd; Defense Scientist) and Wendy Sullivan-Kwantes (MA; Group Leader) Defense Research and Development Canada 1133 Sheppard Ave. W., Toronto ON Canada M3K 2C9

Principal Investigator: Dr. Ira Jacobs (DrMedSc) Faculty of Kinesiology and Physical Education, University of Toronto

Co-Investigators: Dr. Mary L’Abbe (PhD) Dept. of Nutritional Sciences, University of Toronto

Dr. Len Goodman (PhD) Defence R&D Canada – Toronto Research Centre

Run Directors: Doug Saunders, Christina Powesland, and Ingrid Smith Defense Research and Development Canada 1133 Sheppard Ave. W., Toronto ON Canada M3K 2C9 253

Questions about the research? If you have questions about the research in general or about your role in the study, please feel free to contact Iva Mandic (Ph.D. Candidate) either by telephone at (416) 978- 8563, or by e-mail ([email protected]) or Mavra Ahmed (Ph.D. Candidate) either by telephone at (416) 564-8924 or by e-mail ([email protected]).

If you have any questions about your rights as a research participant, please contact Dr. Don McCreary either by phone at 416-635-2098 or by email ([email protected]).

For CAF members: I acknowledge and understand that the Veterans Review and Appeal Board adjudicate cases of disability or death on an individual basis, and to be eligible for compensation or pension, the injury or harm must arise out of or be directly connected with participation in the study.

I understand that I am considered to be on duty for disciplinary, administrative and Pension Act purposes during my participation in this study and I understand that in the unlikely event that my participation in this study results in a medical condition rendering me unfit for service, I may be released from the CAF and my military benefits apply. This duty status has no effect on my right to withdraw from the study at any time I wish, and I understand that no action will be taken against me for exercising this right.

Withdrawal from the Study: I understand that I am free to refuse to participate and may withdraw my consent without prejudice or hard feelings at any time. Should I withdraw my consent, my participation as a subject will cease immediately, unless the Investigator(s) determine that such action would be dangerous or impossible (in which case my participation will cease as soon as it is safe to do so). I also understand that the Investigator(s), their designate, or the physician(s) responsible for the research project may terminate my participation at any time, regardless of my wishes. In the event I withdraw from the study, all associated data collected will be immediately destroyed.

Legal Rights and Signatures:

I understand that by signing this consent form I have not waived any legal rights I may have as a result of any harm to me occasioned by my participation in this trial beyond all risks I have assumed. I understand that I will be given a copy of this consent form so that I may contact any of the above- mentioned individuals at some time in the future should that be required.

Voluntary Consent

Volunteer’s Signature:______Date: ______

Witness Name: ______Witness Signature: ______Date ______

Principal Investigator: ______

I have verbally briefed the participant of the requirements for this research experiment.

Signature: ______Date: ______254

In case of emergency:

Family Member or Designated Person (Name, Address, Phone number & Relationship)

______

Physician Certification (only as required):

Certified fit to participate in this experiment as outlined in the research protocol with the limitations appended below.

I have read the protocol and certify ______fit to participate in this experiment as outlined in the research protocol. This information is only required if it is deemed necessary by the PAR-Q+.

Physician's Name: ______

Licence # ______Province Issued: ______

Physician's Signature: ______Date: ______

Supervisor Approval: Section Head/Commanding Officer’s Signature (see Notes below) ______

CO’s Unit: ______

Notes:

For Military personnel on permanent strength of CFEME: Approved in principle by Commanding Officer; however, members must still obtain their Section Head’s signature designating approval to participate in this particular research project.

For other military personnel: All other military personnel must obtain their Commanding Officer’s signature designating approval to participate in this research project.

255

Appendix F – Invasive Procedures Consent Form INVASIVE PROCEDURES CONSENT FORM Title: The effects of environmental and physical stress on energy expenditure, energy intake and appetite. Protocol Number: 2013-075 Principal Investigators: Iva Mandic (PhD Candidate; University of Toronto), Mavra Ahmed (PhD Candidate; University of Toronto) Co-Investigators: Ira Jacobs (DrMedSc; University of Toronto), Mary L’Abbe (PhD; University of Toronto), Len Goodman (PhD; DRDC Toronto), Wendy Sullivan-Kwantes (MA; DRDC Toronto). Run Directors: Doug Saunders, Ingrid Smith, Christina Powesland. WBE: Personnel Portfolio/Diagnostics and Health Protection/04kc Nutrition)

1. In conjunction with the experiments in which I am participating as a participant, I have agreed to have 73 ml of blood sampled on the first day of each of the four (4) experimental trials via an indwelling catheter inserted into a forearm vein; and I have agreed to have 13 ml of blood sampled on the second day of each of the four (4) experimental trials via venipuncture (a needle) and on Visit 3. I am aware that possible complications experienced by others include: fainting, nausea, bruising or infection of the punctured site. I am aware that a total of 357 ml (~24 tablespoons) of blood will be taken during visits 3- 11. 2. In the event that I do not complete the testing conditions (hot, cold, temperate and sedentary) within three (3) months, and I still wish to continue my participation in the study, I have agreed to have an additional 13 ml of blood collected from me with a needle prior to the final testing conditions. 3. One of the investigators has discussed with me all details of the procedures involved with the blood sampling and the possible physiological consequences of such sampling. 4. My consent is given voluntarily and under free power if choice. I have been informed that I may revoke my consent to these sampling procedures at any time without prejudice. 5. I have been informed that these procedures will only be carried out by individuals have undergone formal phlebotomy (blood sampling) training and are experienced with the venous blood sampling techniques to be used in this study.

Volunteer’s Signature: ______Date: ______

Witness Name: ______Witness Signature: ______Date: ______

Principal Investigator: I have discussed the risks associated with this procedure with the participant.

Principal Investigator: ______Signature: ______Date: ______

I understand that I shall be given a copy of this consent form.

256

Appendix G – PAR-Q+

258

259

260

Appendix H – Participant Demographics

Participant Demographics

Subject Identifier: ______Visit Date (d/m/y): ____ /____ /____

Age ______Gender □ M □ F

Height (cm): ______Weight (kg): ______BMI:______

Ethnicity: □Caucasian □Asian □African American □Hispanic □Other (specify):______

Marital Status: □Single □ Married/Common-law partner □Separated/Divorced/Widowed

Education: □Less than high school □ High School graduation certificate or equivalent

□Non-university certificate or diploma from a community college, CEGEP □University degree

Tel#: ( ) ______

Email address:______

Investigator: ______

Appendix I – Pittsburgh Sleep Quality Index

262

263

264

265

266

267

Appendix J – Physical Activity Questionnaire Please answer the following questions regarding your physical activity during a typical week over the past 3 months. Keep in mind both work and leisure activity.

1. Approximately how many times in a typical week do you perform vigorous exercise? Vigorous requires that your heart rate is significantly increased and your breathing becomes heavy and laboured. Some examples include intense bicycling, running, heavy lifting, competitive tennis, and basketball.

Times per week (If you answered “0”, move on to Question 3)

2. Approximately how long do you perform vigorous exercise in one of these sessions?

Hours per session

3. Approximately how many times in a typical week do you perform moderate exercise? Moderate exercise requires that your heart rate and breathing rate are above normal, but you are still able to speak. Some examples include gentle swimming, light bicycling, light aerobics, and golf.

Times per week (If you answered “0”, move on to question 5)

4. Approximately how long do you perform moderate exercise in one of these sessions?

Hours per session

5. Approximately how many times in a typical week do you perform light exercise? Light exercise requires that you are able to breathe and speak normally while performing the activity. Some examples include walking, household chores, or gardening.

Times per week

6. Approximately how long do you perform light exercise in one of these sessions?

Hours per session

7. Do you consume any daily and/or pre- or post- exercise nutritional supplements and/or vitamins? Examples may include synthetic protein products, Vitamin A/C/D, multivitamins, fish oils/fatty acids, etc.

Yes No

If yes, please list:______268

Appendix K – Food Record Instructions Created by Ahmed, M.[252] Directions for completing a 3-day weighed food record Please use this form to record all foods and beverages consumed over THREE consecutive days, preferably including TWO days during the weekend ONE weekend day. 1. Follow the description and template below to fill in your food record with specific details regarding your consumption. Meal/Snack Date ____ Food and Drinks description Amount Cooking Time/Place method In this In this In this column, please tell me In this column, In this column, column, everything you eat and drink and in you will specify column, you will please specify as much detail as possible. Please the amount of please specify if its the date the weigh the amounts of foods and/or each item. indicate lunch, meals were use measuring utensils. For Please use the cooking breakfast, recorded, example, whole wheat or white measurement method dinner or a time and the bread. If it is a salad or any other utensils or refer used in snack place. For mixed item you are having, break it to the portion preparing example, 23 down: say you had two slices of sizes sheet as the item. Sep 2013, tomatoes with 5 slices of cucumber. required. For 6pm, Also, note any salt, seasonings and example, restaurant. condiments, butter/margarine you baked, put on your food. That includes salad grilled, dressing, ketchup, mustard. Last but fried, not least, note down the drinks you sautéed. have during the day. For example, tea with how much milk and how much sugar.

2. Utensils needed: • 1 set of standard measuring spoons • 1 set of standard measuring cups • Food Scale (provided) 3. Equivalent measures: • 3 teaspoons (tsp) = 1 tablespoon (tbl) • 2 tablespoons = 1 fluid ounce (oz) • 16 tablespoons = 1 cup (c) 4. Record the date and the exact amount of all foods and liquids you ate during this 3-day period. 5. When possible, list brand names of foods and liquids consumed. 6. Include any recipes in your food record. 7. Mention the amount of water you had during each day. 8. Please be as accurate as possible, and record everything consumed during this time period. 9. Please bring the completed record to the study visit. 10. If you have any questions, please contact: Mavra Ahmed, Study Coordinator, University of Toronto, at 416-564-8924 or email: [email protected]

Meal or Date Place: Food and Drinks Description Amount Cooking Snack? 23 Sep Home/Restaurant Cooking Method (cups, ml, ounces, grams, tsp, #, # Method 2013 (include water, spices, and salt) of shakes, etc.)

Time Breakfast 9:00am Home 2% milk 244 g Breakfast 9:00am Home Omega-3 Egg, Omelette cooked 61 g Fried in olive oil Breakfast 9:00am Home Pumpernickel bread slice ½ inch 32 g Toasted thick Breakfast 9:00am Home Unsalted butter (Becel) 8 g

Lunch 1:00pm Restaurant Home style biscuit-fast food 35 g/ 1 each Fried (Popeye’s) Lunch 1:00pm Restaurant Chicken breast, battered fried- 5 oz Fried fast food Lunch 1:00pm Restaurant Hot sauce 2 tsp Lunch 1:00pm Restaurant Brewed Green Tea 244 g

Dinner 7:00pm Home Spaghetti 1 cup Boiled Dinner 7:00pm Home Pasta Tomato Sauce (Ragu) 70 g Dinner 7:00pm Home Allspice, ground 0.03 g/ 1 shake Dinner 7:00pm Home Salt 0.06 g/3 shakes Dinner 7:00pm Home Mashed potato made in milk 60.5 g Home-made and butter

Snack 10:00pm Home Plum 1 each Snack 10:00pm Home Fruit yogurt (Blueberry) 175 g

All Day : Water 1 litre

270

EXAMPLE TEMPLATE Note: Participants will be provided with a sample empty sheet to record their intakes. Meal or Date Place: Food and Drinks Amount Homemade? Snack? ______Home/Restaurant Description (cups, ml, ounces, grams, tsp, #, # of Cooking Method shakes, etc.) Time (include water, spices, and salt) :

:

:

:

:

:

:

:

:

:

:

:

:

:

:

:

:

:

272

Appendix L – Visual Analogue Scales for Appetite (VASA) Appetite Index

How hungry are you right now? Please place a vertical mark on the line below to indicate the level of hunger you feel.

Not at all hungry Extremely hungry

How full do you feel right now? Please place a vertical mark on the line below to indicate the level of fullness you feel.

Not at all Extremely full full

Page 1 of 2

273

How full satisfied are you right now? Please place a vertical mark on the line below to indicate how satisfied you feel.

I am completely I can’t eat

empty another bite

How much do you think you can eat right now? Please place a vertical mark on the line below to indicate how much you think you can eat.

Nothing at all A huge amount

Page 2 of 2

274

Appendix M - GSQS The Groningen Sleep Quality Score The following questions relate to your sleep quality last night. Please pick only one response for each question – ‘true’ or ‘false’. Please answer all questions. 1. I had a deep sleep last night 11. I didn’t sleep a wink last night True False True False 2. I feel that I slept poorly last night 12. I didn’t have trouble falling asleep True False last night

3. It took me more than half an hour to True False fall asleep last night

13. After I woke up last night, I had True False trouble falling asleep again

4. I woke up several times last night True False

True False 14. I tossed and turned all night last night 5. I felt tired after waking up this morning True False

True False 15. I didn’t get more than 5 hours’ sleep last night 6. I feel that I didn’t get enough sleep last night True False

True False

7. I got up in the middle of the night

True False

8. I felt rested after waking up this morning True False

9. I feel that I only had a couple of hours’ sleep last night True False

10. I feel that I slept well last night

True False

275

Appendix N – Thermal Comfort Scale Thermal Comfort Scale

I am

1. So Cold I am helpless

2. Numb with cold

3. Very Cold

4. Cold

5. Uncomfortably Cool

6. Cool but fairly comfortable

7. Comfortable

8. Warm but fairly comfortable

9. Uncomfortably warm

10. Hot

11. Very hot

12. Almost as hot as I can stand

13. So hot I am sick and nauseated

276

Appendix O – Borg Scale Borg Scale (rating of perceived exertion) 6 7 very very light

8 very light

9 fairly light

12 13 somewhat hard 14 15 hard 16 17 very hard 18 19 very very hard 20 max

277

Appendix P – End of Study Questionnaire End of Study Questionnaire

1) On the long study days, do you think you consumed the same amount of food that you normally would?

Condition YES NO Hot Temperate Cold Sedentary

If ‘YES’ for all conditions skip to question #5

2) If No, do you think you consumed more or less food then you normally would?

Condition More Less Hot Temperate Cold Sedentary

3) Why do you think you ate more food than you normally do, please check all that apply for each condition:

Reason Hot Temperate Cold Sedentary I was hungrier than I normally am I was eating out of boredom I liked the food that was provided I used it to help keep me warm/cool me down I felt I needed more food because I was more active than I normally am Other (please list all reasons for why you think you may have eaten more than you normally do during any of the trials): ______

278

4) Why do you think you ate less food than you normally do, please check all that apply: Reason Hot Temperate Cold Sedentary I didn’t feel as hungry as I normally do There wasn’t enough time to eat I didn’t like the taste of the food that was provided It was too warm or too cold to eat I didn’t feel that I needed as much food as I normally would eat

Other (please list all reasons for why you think you may have eaten less than you normally do during any of the trials): ______

5) Do you typically eat when you’re hungry and stop when you are satisfied? (check one) ⎕ Yes ⎕ No

6) Do you think you typically eat too much, too little or just the right amount? ⎕ Too much ⎕ Too little ⎕ Just enough

7) In general, how satisfied are you with your current body weight/shape/size? (check one)

⎕ Not at all satisfied ⎕ Slightly satisfied ⎕ Fairly satisfied ⎕ Quite satisfied ⎕ Very satisfied

8) Are you actively trying to change your body weight/shape/size? (check one)

⎕ No ⎕ Yes

If ‘NO’ thank you for taking the time to complete the questionnaire

279

9) If yes, what would you like to change about your body weight/shape/size? (check one) ⎕ Lose more than 5 lbs ⎕ Lose 1-5 lbs ⎕ Gain 1-5 lbs ⎕ Gain more than 5 lbs ⎕ Other (please specify)______

10) What are you doing in order to achieve this (check all that apply)?

⎕ Restricting overall food consumption ⎕ Exercising more ⎕ Increasing overall food consumption ⎕ Drinking more water ⎕ Choosing healthier food options ⎕ Eating more frequently ⎕ Eating more fruits and vegetables ⎕ Skipping meals ⎕ Eating more protein ⎕ Counting calories ⎕ Limiting the intake of processed foods ⎕ Limiting sugar intake ⎕ Using natural health products

Other (please specify):

______

Thank you for completing the questionnaire

280

Appendix Q – Food Satisfaction Survey

Food Satisfaction Survey Date: ______Time: ______Breakfast (Only evaluate the food you ate) Main Questionnaire Rating Scale: 1: completely unacceptable 2: largely unacceptable 3: borderline 4: largely acceptable 5: completely

acceptable QUESTION/ISSUE RATING COMMENTS

1. Ease of preparation 1 2 3 4 5

2. Ease of consumption 1 2 3 4 5

3. Food temperature 1 2 3 4 5

4. Food preparation time 1 2 3 4 5

1 2 3 4 5 5. Variety 1 2 3 4 5

6. Quantity 1 2 3 4 5

7. Taste 1 2 3 4 5

8. Texture 1 2 3 4 5

9. Saltiness 1 2 3 4 5

10. Sweetness 1 2 3 4 5

1 2 3 4 5 11. Density and Fullness 1 2 3 4 5

1 2 3 4 5 12. Digestibility 1 2 3 4 5

1 2 3 4 5 13. Overall adequacy 1 2 3 4 5

Please note any major issues/concerns you have about the ration pack: ______****Your responses to this questionnaire are completely anonymous and confidential; your name/rank obtained for consent purposes will not be linked or associated with this questionnaire. It will not be shared to your commanding officer, and the questionnaires will be destroyed after data analysis****

281

Food Satisfaction Survey Date: ______Time: ______Lunch (Only evaluate the food you ate) Main Questionnaire Rating Scale: 1: completely unacceptable 2: largely unacceptable 3: borderline 4: largely acceptable 5: completely

acceptable QUESTION/ISSUE RATING COMMENTS

1. Ease of preparation 1 2 3 4 5

2. Ease of consumption 1 2 3 4 5

3. Food temperature 1 2 3 4 5

4. Food preparation time 1 2 3 4 5

1 2 3 4 5 5. Variety 1 2 3 4 5

6. Quantity 1 2 3 4 5

7. Taste 1 2 3 4 5

8. Texture 1 2 3 4 5

9. Saltiness 1 2 3 4 5

10. Sweetness 1 2 3 4 5

1 2 3 4 5 11. Density and Fullness 1 2 3 4 5

1 2 3 4 5 12. Digestibility 1 2 3 4 5

1 2 3 4 5 13. Overall adequacy 1 2 3 4 5

Please note any major issues/concerns you have about the ration pack: ______****Your responses to this questionnaire are completely anonymous and confidential; your name/rank obtained for consent purposes will not be linked or associated with this questionnaire. It will not be shared to your commanding officer, and the questionnaires will be destroyed after data analysis****

282

Food Satisfaction Survey Date: ______Time: ______Dinner (Only evaluate the food you ate) Main Questionnaire Rating Scale: 1: completely unacceptable 2: largely unacceptable 3: borderline 4: largely acceptable 5: completely

acceptable QUESTION/ISSUE RATING COMMENTS

1. Ease of preparation 1 2 3 4 5

2. Ease of consumption 1 2 3 4 5

3. Food temperature 1 2 3 4 5

4. Food preparation time 1 2 3 4 5

1 2 3 4 5 5. Variety 1 2 3 4 5

6. Quantity 1 2 3 4 5

7. Taste 1 2 3 4 5

8. Texture 1 2 3 4 5

9. Saltiness 1 2 3 4 5

10. Sweetness 1 2 3 4 5

1 2 3 4 5 11. Density and Fullness 1 2 3 4 5

1 2 3 4 5 12. Digestibility 1 2 3 4 5

1 2 3 4 5 13. Overall adequacy 1 2 3 4 5

Please note any major issues/concerns you have about the ration pack: ______****Your responses to this questionnaire are completely anonymous and confidential; your name/rank obtained for consent purposes will not be linked or associated with this questionnaire. It will not be shared to your commanding officer, and the questionnaires will be destroyed after data analysis****

283

Appendix R – University of Toronto Ethics Approval

284

Appendix S – DRDC Ethics Approval

285

286

Appendix T – Perceptions of Food Intake As part of the “End of Study Questionnaire”, participants were asked about their general outlook on their eating behaviours, as well as their thoughts regarding food intake within each trial of the study.

General Eating Behaviour Regarding general attitudes towards food intake, 17 out of 18 participants agreed with the following statement: “Do you typically eat when you’re hungry and stop when you are satisfied?”

When asked: “Do you think you typically eat too much, too little or just the right amount?” Most (10 out of18) stated that they eat “just enough”, 7 stated that they eat “too much” and 1 thought that they ate “too little”.

Food Intake during the Study In regard to their perceptions of their own food intake during each trial, this is how participants answered the following question: “Do you think you ate more/less/or as much as you normally would?”

Table IX: Percentage of participants who thought they ate more, less, or the same amount they normally would during each trial More Same Less Don’t Know* Hot 11% 22% 56% 11% Temperate 6% 50% 28% 17% Cold 22% 22% 44% 11% Sedentary 11% 33% 33% 22% If more than one box was checked off, the participant’s response was listed as “Don’t Know”

When comparing actual intake during each trial with how much participants thought they ate it became clear that participants were generally unaware of how much they ate. When asked whether they thought they consumed more or less than usual, participants were only correct 35% of the time, 46% of the time they thought they ate less than they did, and 19% of the time they thought they ate more than they did.

Participants also provided reasons for why they thought ate more in particular conditions (Table X). Interestingly 7 out of 18 (39%) participants thought they consumed more in the cold in order to help them stay warm.

287

Table X: Why participants thought they ate more in each condition Hot Temperate Cold Sedentary I was hungrier than I normally am 28% 33% 22% 28% I was eating out of boredom 11% 11% 0% 0% I liked the food that was provided 0% 0% 0% 0% I used it to help keep me warm/cool me down 11% 6% 39% 0% I felt I needed more food because I was more active 22% 33% 22% 0% than I normally am Participants could check off any number of statements that they agreed with. As a result, the data represents the percentage of all participants (18) that agreed with each of the following statements for each condition. Additional space was provided for participants to add reasons not mentioned above. One participant also included the following reason: “I think I drank more of the sport drink than I would have”

On the other end of the spectrum, in the hot condition 7 out of 18 (39%) participants, and in the Sedentary condition 6 out of 18 (33%) participants thought they ate less due to not feeling as hungry as they normally would. In addition 7 out of 18 (39%) participants thought they didn’t need as much food during the Sedentary condition as they normally would (Table XI).

Table XI: Why participants thought they ate less in each condition Hot Temperate Cold Sedentary I didn’t feel as hungry as I normally do 39% 11% 17% 33% There wasn’t enough time to eat 6% 6% 11% 0% I didn’t like the taste of the food that was provided 22% 22% 28% 28% It was too warm or too cold to eat 17% 0% 28% 0% I didn’t feel that I needed as much food as I normally would eat 11% 6% 6% 39% Participants could check off any number of statements that they agreed with. As a result, the data represents the percentage of all participants (18) that agreed with each of the following statements for each condition. Additional space was provided for participants to add reasons not mentioned above. Participants also included the following reasons: “I do not eat much for marches and exercises” (n=1); “Under exertion especially in the heat if you are too full, your food starts coming up if you eat too much. I.e. you feel like puking” (n=1); “Insufficient food in ration” (n=1); “It didn't taste good. Too salty too much sodium. Too much sugar” (n=1).

288

Appendix U - Food Satisfaction

The food that was provided was rated as “acceptable” overall.

Figure 38: Average ratings of the military food that was provided. Participants rated each component of food adequacy on a 1-5 scale (1=completely unacceptable; 2= largely unacceptable; 3=borderline; 4=largely acceptable; 5=completely acceptable.) Data are adapted from [252] and are displayed as mean±SD

289

Appendix V- Energy Intake at Home vs. in the Environmental Chamber Energy intakes were not different between conditions. Regardless of whether participants consumed their own food at home, consumed military rations (IMPs) at home, or completed any of the trial days, caloric intake was not different (Table XII) [252].

Table XII: Energy Intakes in different scenarios Average Average Energy Energy Energy Energy Energy Energy intake on intake intake on intake intake when intake when Sedentary on Cold Temperate on Hot Participant consuming consuming trial day trial trial day trial own food at IMPs at (kcal) day (kcal) day home* (kcal) home# (kcal) (kcal) (kcal) 1 2701 1891 2799 2736 1856 2713 2 2296 2563 3148 3573 3730 4164 3 1677 2175 2037 1958 2968 2041 4 3102 2513 2692 2610 2660 3117 5 1836 2497 1806 2155 3564 2079 6 3356 3705 6016 5846 4435 4405 7 2783 3064 2928 2688 2582 2592 8 2767 3137 2114 2964 2636 2923 9 2883 3061 3486 2803 2063 3618 10 2558 2627 2123 3060 3476 1993 11 3917 4062 4228 3336 3651 4944 12 2695 2789 2514 2577 2438 3020 13 3447 2454 1892 2699 2190 3473 14 2634 2843 3013 3162 3579 2679 15 2360 1869 3055 2007 2252 2869 16 2443 3236 3146 3050 3752 3790 17 2626 1908 4227 3073 3313 3579 18 1751 2007 1600 1284 1474 1474 *Average energy intake over three days while participant was consuming their own food at home # Average of 8 day of consuming IMPs at home Cells are coloured green when the energy intake on the experimental day was more than 1 SD above the mean of the 8 days of IMP consumption at home Cells are coloured red when the energy intake on the experimental day was more than 1 SD below the mean of the 8 days of IMP consumption at home Data adapted from [252]

290

Figure 39: Average energy intake when consuming personal food vs. IMPs at home. Each colour depicts a different participant. Data are adapted from [252].

291

Appendix W – Subjective Appetite

Figure 40: Average scores for each index of subjective appetite during each condition(Cold: blue lines; Hot: red lines; Temperate: green lines; Sedentary: purple lines), as collected by visual analogue scales for the 4 indices of appetite. Grey boxes represent the 2-h ‘activity blocks’, although during the Sedentary condition the participants were inactive during these blocks. Fasting and Post-Breakfast data points were collected outside the environmental chamber prior to trial commencement. The 8-h trial began once the participant entered the environmental chamber; this occurred within minutes of the participants completing their breakfast. Data are mean±SEM.

292

Appendix X - Appetite Hormone Concentration

Figure 41: Appetite regulating hormone concentrations during each condition (Cold: blue lines; Hot: red lines; Temperate: green lines; Sedentary: purple lines). This figure displays complete data for 7 (GLP-1 and Leptin) or 8 (PYY and Acylated Ghrelin) participants. The grey boxes represent the 2-h ‘activity blocks’, although during the Sedentary condition the participants were inactive during these blocks. The ‘0’ time point represents the fasting blood sample collected outside of the environmental chamber prior to trial commencement, all other blood samples were collected in the environmental chamber. Data are mean±SEM.

293

Appendix Y – Cortisol Results There was no effect of day, condition, or day by condition on 24-h urinary cortisol.

Table XIII: 24-h Urinary Free Cortisol Before and After Each Trial 24-h Urinary Cortisol 24-h Urinary Cortisol Collected the Collected on Trial Day Day Before the Trial (nmol∙day-1) (nmol∙day-1) Baseline* 147.6±94.5 170.9±85.2 Sedentary 125.7±61.9 149.5±89.9 Cold 130.5±76.6 131.7±95.6 Temperate 138.7±95.6 132.5±47.3 Hot 142.4±63.5 120.5±53.9 *Baseline urinary free cortisol was collected for 2 days prior to Visit 3 Data are displayed as Mean±SD

294

Appendix Z – Profile of Mood States (POMS) Results Total Mood Disturbance (TMD) Score

There was no main effect of condition, on the TMD score, nor was there a main effect of time of day on TMD. There was also no interaction between condition and time of day for TMD.

Tension and Anxiety (TA) Score

A significant main effect of condition on the TA score was suggested by the omnibus test F(3, 51) =3.10, p<0.05; however no pairwise comparisons were able to reach statistical significance. There was no main effect of time of day, on the TA score. There was also no interaction between condition and time of day for TA.

Depression and Dejection (DD) Score

There was a significant main effect of time of day, on the DD score F(1,17) =8.43, p<0.05. DD scores were significantly higher in the morning (Morning DD score=2.08±3.48) than they were in the afternoon (Afternoon DD score=1.65±3.27) p<0.05. There was no main effect of condition, or interaction between condition and time of day for DD.

Anger and Hostility AH) Score

There was a significant main effect of time of day, on the AH score F(1,17) =4.58, p<0.05. AH scores were significantly higher in the morning (Morning AH score=3.44±3.27) than they were in the afternoon (Afternoon AH score=2.86±3.01) p<0.05. There was no main effect of condition, or interaction between condition and time of day for AH.

Vigor and Activity (VA) Score

There was no main effect of condition, on the VA score, nor was there a main effect of time of day on VA. There was also no interaction between condition and time of day for VA.

Confusion and Bewilderment (CB) Score

There was no main effect of condition, on the CB score, nor was there a main effect of time of day on CB. There was also no interaction between condition and time of day for CB.

295

Fatigue Score

The assumption of sphericity was violated for the main effect of condition as indicated by 2 Mauchly’s test χ (5) = 11.27, p<0.05. The assumption of sphericity was also violated for the 2 interaction between condition and time of day as indicated by Mauchly’s test χ (5) = 26.66, p<0.05. As a result, the degrees of freedom were corrected for both the main effect of condition, and for the interaction between condition and time of day using the Greenhouse-Geisser correction.

There was a significant main effect of condition on fatigue F(2.15,36.46) =4.52, p<0.05. Participants reported a significantly higher fatigue score on the hot days than they did on the sedentary and temperate days (Hot fatigue score = 6.97±3.43 vs. Sedentary fatigue score = 4.03±3.61; Temperate fatigue score = 5.33±3.55). There was a significant main effect of time of day, on the fatigue score F(1,17) =12.53, p<0.05. As expected, fatigue scores were significantly higher after the 8 hours in the chamber than they were first thing in the morning (Morning fatigue Score = 4.33±3.32 vs. Afternoon fatigue score = 6.46±3.09). There was a significant interaction effect between condition and time of day on fatigue score F(1.53,25.93) =4.04, p<0.05. During the Hot and Cold trials fatigue increased from morning (pre-trial) to afternoon (post-trial) significantly more than during the sedentary condition; during which participants reported lower fatigue scores in the afternoon (post-trial), compared to the morning (pre-trial) (Table XIV).

Table XIV: POMS Fatigue scores before and after each trial Cold Hot Temperate Sedentary Morning 3.67±4.63 4.56±3.03 4.72±6.12 4.39±4.33 (pre-trial) * * Afternoon 6.83±3.94 9.39±5.26 5.94±4.02 3.67±3.11 (post-trial) *Denotes that the change from pre to post is significantly greater than was reported during the Sedentary condition

296

Appendix AA – Substrate Oxidization Carbohydrate Oxidation The assumption of sphericity was violated for the main effect of condition as indicated by 2 Mauchly’s test χ (2) = 121.93, p<0.05. As a result, the degrees of freedom were corrected using the Huynh-Feldt correction. When investigating carbohydrate oxidization in the three active conditions, a significant main effect of condition emerged F(1.74,1277.62) =35.07, p<0.05. In the Hot condition participants utilized significantly less carbohydrate (268±167 kcal∙h-1), than they did in the Cold (298±166kcal∙h-1) or Temperate (294±163 kcal∙h-1) conditions p<0.05. There were no other significant differences.

Fat Oxidation The assumption of sphericity was violated for the main effect of condition as indicated by 2 Mauchly’s test χ (2) = 223.43, p<0.05. As a result, the degrees of freedom were corrected using the Huynh-Feldt correction. When assessing fat oxidation, a significant main effect of condition was also determined F(1.59,1166.52) =73.61, p<0.05. In the Hot condition participants oxidized more fat (133±97 kcal∙h-1) than they did in either the Cold (104±97 kcal∙h-1) or Temperate (97±80 kcal∙h-1) conditions p<0.05. There was also a trend towards more fat oxidation in the Cold as compared with the Temperate condition p=0.078.

Protein Oxidation The assumption of sphericity was violated for the main effect of condition as indicated by 2 Mauchly’s test χ (2) = 52.96, p<0.05. As a result, the degrees of freedom were corrected using the Huynh-Feldt correction. There was also a significant main effect of condition on protein oxidation F(1.87,1377.71) =11.78, p<0.05. Less protein was oxidized during the Temperate (40±16 kcal∙h-1) condition than during either the Hot (41±16 kcal∙h-1) or Cold (41±17 kcal∙h-1) conditions. No other significant differences emerged.

Table XV: Substrate Oxidation (kcal∙h-1) During Each Trial Cold Hot Temperate Carbohydrate (kcal∙h-1) 298±166 268±167* 294±163 Fat (kcal∙h-1) 104±97 133±91* 97±80 Protein (kcal∙h-1) 41±17 41±16 40±16* Data are displayed as Mean±SD *Significantly different from other conditions

297