{"http:\/\/dx.doi.org\/10.14288\/1.0398500":{"http:\/\/vivoweb.org\/ontology\/core#departmentOrSchool":[{"value":"Health and Social Development, Faculty of (Okanagan)","type":"literal","lang":"en"},{"value":"Health and Exercise Sciences, School of (Okanagan)","type":"literal","lang":"en"}],"http:\/\/www.europeana.eu\/schemas\/edm\/dataProvider":[{"value":"DSpace","type":"literal","lang":"en"}],"https:\/\/open.library.ubc.ca\/terms#degreeCampus":[{"value":"UBCO","type":"literal","lang":"en"}],"http:\/\/purl.org\/dc\/terms\/creator":[{"value":"Gelinas, Jinelle Crystal Marie","type":"literal","lang":"en"}],"http:\/\/purl.org\/dc\/terms\/issued":[{"value":"2021-06-23T21:47:20Z","type":"literal","lang":"en"},{"value":"2021","type":"literal","lang":"en"}],"http:\/\/vivoweb.org\/ontology\/core#relatedDegree":[{"value":"Doctor of Philosophy - PhD","type":"literal","lang":"en"}],"https:\/\/open.library.ubc.ca\/terms#degreeGrantor":[{"value":"University of British Columbia","type":"literal","lang":"en"}],"http:\/\/purl.org\/dc\/terms\/description":[{"value":"Introduction: Phenotyping patients with COPD according to their specific exercise limitation could be important to enhancing our understanding of the integrative physiological responses that limit exercise and how to optimize aerobic exercise prescription for pulmonary rehabilitation. Three studies were performed to identify distinct phenotypes of exercise limitation in COPD based on the ventilatory, cardiovascular and metabolic responses during incremental exercise. Differences in the physiological factors governing submaximal exercise tolerance in each phenotype were also assessed to accurately predict exercise tolerance at two novel indices of sustainable intensity; the highest workload that could be sustained for 30-minutes (critical power 30 [CP30]) and maximal dyspnea steady-state (MDSS). \r\nMethodology: Study 1: Ninety-five COPD patients performed an incremental cardiopulmonary exercise test to assess the physiological exercise responses associated with three different phenotypes of exercise limitation. Study 2: Thirty patients with a ventilatory, cardiovascular or combined (ventilatory and cardiovascular) phenotype performed three constant load exercise tests (CLTs) to determine the individual power-duration relationships and calculate critical power and CP30. An exercise session was performed at CP30. Study 3: In twenty-three patients with different exercise limitation phenotypes, the workload associate with MDSS was calculated from the individual dyspnea responses during the CLTs. A verification trial was performed at MDSS. \r\nResults and Conclusion: Study 1 demonstrated that patients with a ventilatory, cardiovascular or combined exercise limitation phenotype have distinct physiological responses to incremental exercise, that are not solely dependent on FEV\u2081. Study 2 found that patients with a cardiovascular or combined phenotype could sustain continuous exercise for 30-minutes at a higher workload and blood lactate concentration while maintaining a lower end-expiratory lung volume compared to the ventilatory phenotype. Study 3 confirmed that MDSS can be successfully predicted in COPD but found that MDSS was more applicable to the ventilatory phenotype, supporting that dyspnea may not be the primary limiting symptom in all patients with a cardiovascular or combined phenotype. These results demonstrate the importance of classifying patients with COPD according to their exercise limitation phenotype. These phenotypes may be clinically relevant and can be used to develop innovative methods to optimize exercise prescription for rehabilitative purposes.","type":"literal","lang":"en"}],"http:\/\/www.europeana.eu\/schemas\/edm\/aggregatedCHO":[{"value":"https:\/\/circle.library.ubc.ca\/rest\/handle\/2429\/78770?expand=metadata","type":"literal","lang":"en"}],"http:\/\/www.w3.org\/2009\/08\/skos-reference\/skos.html#note":[{"value":"i  PHENOTYPING EXERCISE LIMITATIONS IN CHRONIC OBSTRUCTIVE PULMONARY DISEASE TO ENHANCE EXERCISE PRESCRIPTION  by Jinelle Crystal Marie Gelinas  B.H.Kin., University of British Columbia, 2011 M.Sc., University of British Columbia, 2013  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE COLLEGE OF GRADUATE STUDIES (Kinesiology)  THE UNIVERSITY OF BRITISH COLUMBIA (Okanagan) June 2021  \u00a9 Jinelle Crystal Marie Gelinas, 2021 i   The following individuals certify that they have read, and recommend to the College of Graduate Studies for acceptance, a thesis\/dissertation entitled: PHENOTYPING EXERCISE LIMITATIONS IN CHRONIC OBSTRUCTIVE PULMONARY DISEASE TO ENHANCE EXERCISE PRESCRIPTION submitted by Jinelle Gelinas in partial fulfillment of the requirements of the degree of Doctor of Philosophy .   Dr. Neil Eves, School of Health and Exercise Science Supervisor Dr. Jordan Guenette, Department of Physical Therapy Supervisory Committee Member Dr. Gregory duManoir, School of Health and Exercise Science Supervisory Committee Member Dr. Jeremy Road, Faculty of Medicine University Examiner Dr. Francois Maltais, Faculty of Medicine, Laval University  External Examiner  ii Abstract Introduction: Phenotyping patients with COPD according to their specific exercise limitation could be important to enhancing our understanding of the integrative physiological responses that limit exercise and how to optimize aerobic exercise prescription for pulmonary rehabilitation. Three studies were performed to identify distinct phenotypes of exercise limitation in COPD based on the ventilatory, cardiovascular and metabolic responses during incremental exercise. Differences in the physiological factors governing submaximal exercise tolerance in each phenotype were also assessed to accurately predict exercise tolerance at two novel indices of sustainable intensity; the highest workload that could be sustained for 30-minutes (critical power 30 [CP30]) and maximal dyspnea steady-state (MDSS).  Methodology: Study 1: Ninety-five COPD patients performed an incremental cardiopulmonary exercise test to assess the physiological exercise responses associated with three different phenotypes of exercise limitation. Study 2: Thirty patients with a ventilatory, cardiovascular or combined (ventilatory and cardiovascular) phenotype performed three constant load exercise tests (CLTs) to determine the individual power-duration relationships and calculate critical power and CP30. An exercise session was performed at CP30. Study 3: In twenty-three patients with different exercise limitation phenotypes, the workload associate with MDSS was calculated from the individual dyspnea responses during the CLTs. A verification trial was performed at MDSS.  Results and Conclusion: Study 1 demonstrated that patients with a ventilatory, cardiovascular or combined exercise limitation phenotype have distinct physiological responses to incremental exercise, that are not solely dependent on FEV1. Study 2 found that patients with a cardiovascular or combined phenotype could sustain continuous exercise for 30-minutes at a higher workload and blood lactate concentration while maintaining a lower end-expiratory lung volume compared to the ventilatory phenotype. Study 3 confirmed that MDSS can be successfully predicted in COPD but found that MDSS was more applicable to the ventilatory phenotype, supporting that dyspnea may not be the primary limiting symptom in all patients with a cardiovascular or combined iii phenotype. These results demonstrate the importance of classifying patients with COPD according to their exercise limitation phenotype. These phenotypes may be clinically relevant and can be used to develop innovative methods to optimize exercise prescription for rehabilitative purposes.     iv Lay Summary Although chronic obstructive pulmonary disease is considered a lung condition, the contribution of the pulmonary and\/or cardiovascular systems to exercise limitation varies between patients. The physiological responses underlying these differences may explain help why certain patients can exercise at a higher intensity and\/or for a longer duration during aerobic exercise. The purpose of this dissertation was to identify groups of patients with different physiological exercise responses (phenotypes) and to determine how exercise tolerance differs between phenotypes. It was found that the integrative exercise responses to incremental exercise differ between phenotypes, and that patients with a greater cardiovascular contribution to exercise limitation can sustain continuous aerobic exercise at a higher intensity for the same duration compared to patients with a classic ventilatory limitation. Additionally, two novel indices of sustainable exercise intensity were identified for each phenotype that may help to better individualized exercise prescription for patients in pulmonary rehabilitation.        v Preface This dissertation contains the work of the candidate, Jinelle C.M. Gelinas, under the supervision of Dr. Neil Eves. Ms. Gelinas and Dr. Eves conceptualized all studies presented in this document. Drs. Guenette and duManoir (PhD supervisory committee members) contributed to the development of the study designs and specific methodology. Data collection, analysis, interpretation, and document preparation are primarily the work of the candidate. All chapters of the dissertation have been edited by Dr. Eves.  All experiments presented were provided ethical approval. Study #1 (Chapter 2) received ethical approval from the University of British Columbia Clinical Research Ethics Board (H16-01777), Interior Health Authority Research Ethics Board (2016-17-019-H) and the Conjoint Health Research Ethics Board at the University of Calgary (REB18-0185_REN1). Studies #2 and #3 (Chapters 3 and 4, respectively) received ethical approval from the University of British Columbia Clinical Research Ethics Board (H16-02989) and the Interior Health Authority Research Ethics Board (2016-17-044-H). Prospective data for Study #1 and all data presented in Studies #2 & #3 were collected at the Kelowna General Hospital, located in Kelowna, British Columbia, Canada.  Chapter 1: No aspect of this chapter is currently under review for publication.  Chapter 2: A version of this chapter has been written and will be submitted to a journal as of April 2021: Jinelle CM Gelinas, Megan I Harper, John P Sasso, Stephen P Wright, Bernie Melzer, Gloria Agar, Jordan A Guenette, Gregory R duManoir, Michael Roman, J Douglass Rolf, Neil D Eves. Phenotyping Cardiopulmonary Exercise Limitations in Chronic Obstructive Pulmonary Disease. JCMG and NDE were responsible for study conception and design. Data acquisition and analysis were performed by JCMG, MIH, JPS, SPW, BM, GA and NDE. Interpretation of data as well as the drafting and revising of the manuscript was performed by JCMG, MIH, JPS, SPW, BM, GA, JAG, GRD, MR, JDR, and NDE. All authors approved the final version to be published. vi Chapter 3: A version of this chapter has been written and will be submitted to a journal following publication of Study #1: Jinelle CM Gelinas, Stephen P Wright, Megan I Harper, John P Sasso, Gloria Agar, Bernie Melzer, Gregory R duManoir, Jordan A Guenette, J Douglass Rolf, Neil D Eves. Exercise Limitation Phenotype Alters the Power-Duration Relationship in COPD. JCMG and NDE were responsible for study conception and design. Data acquisition and analysis were performed by JCMG, SPW, MIH, JPS, GA, BM and NDE. Interpretation of data as well as the drafting and revising of the manuscript was performed by JCMG, SPW, MIH, JPS, GA, BM, GRD, JAG, JDR, and NDE. All authors approve the final version to be published.  Chapter 4: No aspect of this chapter is currently under review.  Chapter 5: No aspect of this chapter is currently under review.    vii Table of Contents Abstract....................................................................................................................................... ii Lay Summary ............................................................................................................................. iv Preface ....................................................................................................................................... v Table of Contents\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026..vii List of Tables .............................................................................................................................. x List of Figures ............................................................................................................................ xi Abbreviations ........................................................................................................................... xiii Acknowledgements .................................................................................................................. xvi Dedication ............................................................................................................................... xvii Chapter 1: Introduction ............................................................................................................... 1 1.1 Overview of COPD ........................................................................................................... 1 1.1.1 Prevalence of Disease ............................................................................................... 1 1.1.2 Pathophysiology ......................................................................................................... 2 1.1.3 Clinical Diagnosis ....................................................................................................... 3 1.1.4 Treatment and Management ...................................................................................... 6 1.1.5 Pulmonary Rehabilitation ........................................................................................... 7 1.2 Criteria to Determine Exercise Limitation .......................................................................... 9 1.3 Exercise Limitation in Health ............................................................................................10 1.3.1 Exercise Limitation During Incremental CPET in Young Adults .................................12 1.3.2 Exercise Limitation During Incremental CPET in Normal Aging .................................14 1.4 Exercise Limitation in COPD ............................................................................................16 1.4.1 Ventilatory Contributions to Exercise Limitation .........................................................16 1.4.2 Skeletal Muscle Contributions to Exercise Limitation .................................................20 1.4.3 Cardiovascular Contributions to Exercise Limitation ..................................................25 1.4.4 Integrated Exercise Response in COPD ...................................................................28 1.5 Exercise Prescription in COPD ........................................................................................30 1.6 The Power-Duration Relationship ....................................................................................31 1.6.1 Determining the Power-Duration Relationship in Health ............................................33 1.6.2 Methodological Considerations for the Calculation of CP and W\u2019 ..............................35 1.6.3 Determining the Power-Duration Relationship in COPD ............................................37 1.6.4 Using the Power-Duration Relationship to Accurately Predict Exercise Tolerance in COPD ................................................................................................................................39 1.7 Additional Indices of Sustainability ...................................................................................40 1.7.1 Maximal Dyspnea Steady-State ................................................................................42 1.8 Purpose, Aims and Hypothesis ........................................................................................44 viii Chapter 2: Phenotyping Cardiopulmonary Exercise Limitations in Patients with Chronic Obstructive Pulmonary Disease ................................................................................................47 2.1 Background .....................................................................................................................47 2.2 Methodology ....................................................................................................................48 2.2.1 Participants and Study Design ..................................................................................48 2.2.2 Statistical Analysis ....................................................................................................49 2.3 Results ............................................................................................................................50 2.3.1 Participant Characteristics and Pulmonary Function .................................................50 2.3.2 Peak Cardiopulmonary Exercise Responses .............................................................54 2.3.3 Submaximal Cardiopulmonary Exercise Responses .................................................61 2.4 Discussion .......................................................................................................................61 2.4.1 Lung Volume Responses to Incremental Exercise ....................................................62 2.4.2 Cardiovascular Responses to Exercise .....................................................................63 2.4.3 Metabolic Responses to Exercise .............................................................................64 2.4.4 Skeletal Muscle Contributions to Exercise Limitation .................................................65 2.4.5 Clinical Relevance .....................................................................................................66 2.4.6 Study Considerations ................................................................................................67 2.4.7 Conclusion ................................................................................................................67 Chapter 3: Exercise Limitation Phenotype Alters the Power-Duration Relationship in COPD ....68 3.1 Background .....................................................................................................................68 3.2 Methods ..........................................................................................................................69 3.2.1 Participants and Study Design ..................................................................................69 3.2.2 Pulmonary Function Testing and Incremental CPET .................................................70 3.2.3. Determination of Critical Power and W\u2019 ....................................................................71 3.2.4 CP30 Exercise Session .............................................................................................72 3.2.5 Statistical Analysis ....................................................................................................73 3.3 Results ............................................................................................................................73 3.3.1 Phenotype Characteristics and Incremental CPET Responses .................................73 3.3.2 Physiological Responses during the CLTs ................................................................76 3.3.3 CP and W\u2019 .................................................................................................................80 3.3.4 Responses to Exercise at CP30 ................................................................................84 3.3.5 Secondary Analysis: Lung Volume Responses Within Phenotype .............................87 3.4 Discussion .......................................................................................................................89 3.4.1 Differences in CP and W\u2019 between Phenotypes ........................................................89 3.4.2 Exercise Responses to CP30 ....................................................................................91 3.4.3 Practical Application and Clinical Relevance .............................................................92 3.4.4 Study Considerations ................................................................................................93 ix 3.4.5 Conclusion ................................................................................................................93 Chapter 4: Predicting Maximal Dyspnea Steady-State in Different Phenotypes of Exercise Limitation in COPD ...................................................................................................................94 4.1 Background .....................................................................................................................94 4.2 Methodology ....................................................................................................................96 4.2.1 Participants and Study Design ..................................................................................96 4.2.2 Pulmonary Function, Incremental CPET and CLTs ...................................................96 4.2.3 Determination of MDSS .............................................................................................97 4.2.4 MDSS Verification Trial .............................................................................................99 4.2.5 Statistical Analysis ....................................................................................................99 4.3 Results ............................................................................................................................99 4.3.1 Characteristics and Incremental CPET Responses ...................................................99 4.3.2 Physiological Responses during the CLTs .............................................................. 102 4.3.3 Exercise Responses during the MDSS Verification Trial ......................................... 105 4.3.4 Secondary Analysis: Patients in Whom MDSS Could not be Calculated ................. 110 4.5 Discussion ..................................................................................................................... 111 4.5.1 Physiological Responses at the Predicted MDSS .................................................... 111 4.5.2 Subgroup Analysis .................................................................................................. 113 4.5.3 Practical Application of MDSS ................................................................................. 114 4.5.4 Study Considerations .............................................................................................. 115 4.5.6 Conclusion .............................................................................................................. 116 Chapter 5: General Discussion and Conclusion ...................................................................... 117 5.1 Overall Summary ........................................................................................................... 117 5.2 Clinical Implications of the Exercise Limitation Phenotypes ........................................... 118 5.3 Application of the Exercise Limitation Phenotypes to Exercise Prescription ................... 124 5.4 Strengths and Limitations .............................................................................................. 131 5.5 Future Directions ........................................................................................................... 132 5.6 Overall Conclusion ........................................................................................................ 133 Bibliography ............................................................................................................................ 135 Appendix ................................................................................................................................. 158         x List of Tables Table 2.1. Phenotypic Characteristics and Pulmonary Function ................................................53 Table 2.2. Incremental CPET Responses Between Phenotypes ...............................................55 Table 3.1. Phenotype Characteristics and Incremental CPET Responses ................................75 Table 3.2. Constant Load Exercise Responses Between Phenotypes ......................................77 Table 3.3. Exercise Intensity Domains in Different Phenotypes of COPD ..................................82 Table 3.4. End Exercise Responses at CP30 Between Phenotypes .........................................85 Table 4.1. Patient Characteristics and Incremental CPET Responses .................................... 101 Table 4.2. Constant Load Exercise Responses Between Phenotypes .................................... 103 Table 4.3. End Exercise Responses in Patients who Achieved MDSS .................................... 107 Table A.1. MDSS Subgroup Analysis: Characteristics and Incremental CPET Responses ..... 158 Table A.2. MDSS Subgroup Analysis: Constant Load Trial Exercise Responses .................... 159   xi List of Figures Figure 1.1. An illustrative example depicting the power-duration relationship as determined from three high-intensity constant load trials performed at varying workloads ...................................33 Figure 1.2. An illustrative example of the slope coefficient analysis to determine the workload at which RPE will remain at a steady-state during constant load exercise. ....................................41 Figure 2.1. Study participant flow ..............................................................................................51  Figure 2.2. The distribution of airflow limitation severity within each exercise limitation phenotype\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026..52 Figure 2.3. Phenotypic responses in (A) absolute ventilation, (B) relative ventilation (expressed as percentage of estimated maximal ventilatory capacity [MVC]), (C) tidal volume, and (D) breathing frequency during an incremental CPET. ....................................................................56 Figure 2.4. Changes in global lung volumes in the (A) ventilatory, (B) combined and (C) cardiovascular phenotypes, and (D) end-expiratory lung volume, (E) end-inspiratory lung volume, and (F) inspiratory reserve volume during an incremental CPET ..............................................57 Figure 2.5. Phenotypic responses in (A) inspiratory capacity and (B) tidal volume to inspiratory capacity ratio during an incremental CPET ...............................................................................58 Figure 2.6. Changes in (A) absolute heart rate, (B) relative heart rate (expressed as a percentage of age-predicted heart rate maximum), and (C) O2pulse during an incremental CPET between phenotypes of exercise limitation in COPD ...............................................................................59 Figure 2.7. The change in (A) exertional dyspnea, and the interrelationships between exertional dyspnea and (B) inspiratory reserve volume and (C) tidal volume to inspiratory capacity ratio during an incremental CPET between phenotypes of exercise limitation in COPD ....................60 Figure 3.1. Study participant flow ..............................................................................................74 Figure 3.2. Phenotypic responses in (A, B & C) relative minute ventilation (expresses as a percentage of estimated maximal ventilatory capacity [MVC], the dashed line represents 85% of MVC), (D, E & F) relative heart rate (expressed as a percentage of age-predicted maximum heart rate, the dashed line represents 90% predicted), (G, H & I) end-inspiratory lung volume (EILV) and end-expiratory lung volume (EELV), and (J, K & L) VO2 .....................................................79 Figure 3.3. The individual power-duration relationships depicted as the (A) hyperbolic relationship, (B) linear relationship (absolute CP) and (C) linear relationship relative to percentage maximum workload (relative CP) between different phenotypes of exercise limitation in COPD. ...............81 Figure 3.4. Different exercise intensity domains in the (A) ventilatory phenotype, (B) combined phenotype, and (C) cardiovascular phenotype.. ........................................................................83 Figure 3.5. The change in (A) relative minute ventilation (expresses as a percentage of estimated maximal ventilatory capacity [MVC]), (B) relative heart rate (expressed as a percentage of age-predicted maximum heart rate), (C) VO2, (D) capillary blood lactate concentration, (E) end-xii inspiratory lung volume (EILV) and end-expiratory lung volume (EELV), and (F) dyspnea during the CP30 exercise session ........................................................................................................86 Figure 3.6. Within phenotype responses during CLT-1, CLT-2, CLT-3 and CP30 in (A,B,C) relative end-inspiratory lung volume (EILV) and end-expiratory lung volume (EELV), (D,E,F) the absolute change in tidal volume from rest to end-exercise, (G,H,I) the absolute change in end-inspiratory lung volume from rest to end-exercise, and (J,K,L) the absolute change in end-expiratory lung volume from rest to end-exercise ..............................................................................................88 Figure 4.1. An illustrative example of the dyspnea-time slope analysis used to calculate MDSS depicted using the average dyspnea-time ratings during each constant load trial in the (A) ventilatory phenotype, (B) combined phenotype and (C) cardiovascular phenotype. Panel (D) illustrates the slope-workload relationships generated from each dyspnea-time slopes calculated during the CLTs in each phenotype. ..........................................................................................98 Figure 4.2. Study participant flow ............................................................................................ 100 Figure 4.3. Phenotypic responses in (A) dyspnea, (B) the change in dyspnea from 10 to 30 minutes, (C) end-inspiratory lung volume (EILV) and end-expiratory lung volume (EELV), and (D) the change in tidal volume from rest to 30 minutes ................................................................. 108 Figure 4.4. The change in (A) relative minute ventilation (expressed as a percentage of estimate maximal ventilatory capacity [MVC]), (B) relative heart rate (expressed as a percentage of age-predicted maximum heart rate), (C) capillary blood lactate concentration, and (D) leg fatigue during the MDSS verification trial ............................................................................................ 109 Figure 5.1.  A theoretically model of the exercise limitation continuum in patients with COPD.  ............................................................................................................................................... 120 Figure 5.2. The relative proportion of phenotypes associated with each VO2peak quartile ......... 124 Figure 5.3. The different exercise intensity domains in the (A) ventilatory phenotype, (B) combined phenotype, and (C) cardiovascular phenotype expressed relative to the maximum workload achieved during the incremental CPET ................................................................................... 128 Figure 5.4.  A theoretically decision tree algorithm for aerobic exercise prescription based on a four-zone model in COPD patients with different exercise limitation phenotypes ..................... 130    xiii Abbreviations  COPD \u2013 chronic obstructive pulmonary disease  EFL \u2013 expiratory flow limitation  FEV1 \u2013 forced expiratory volume in 1 second  FVC \u2013 forced vital capacity  LLN \u2013 lower limit of normal  GLI \u2013 Global Lung Function Initiative network GOLD \u2013 Global Initiative for Chronic Obstructive Lung Disease  mMRC \u2013 modified Medical Research Council  FRC \u2013 functional residual capacity  TLC \u2013 total lung capacity  RV \u2013 residual volume  IC \u2013 inspiratory capacity  DLCO \u2013 diffusion capacity for carbon monoxide DLCO\/VA, Kco \u2013 diffusion coefficient for carbon monoxide  VA\/Q \u2013 alveolar ventilation to perfusion ratio  VD\/VT \u2013 physiological dead space to tidal volume ratio  H+ \u2013 hydrogen ions CO2 \u2013 carbon dioxide  VO2peak \u2013 peak aerobic capacity  CPET \u2013 cardiopulmonary exercise test  HRpeak \u2013 peak heart rate  HRmax \u2013 age-predicted maximum heart rate  VEpeak \u2013 peak minute ventilation  MVV \u2013 maximum voluntary ventilation  O2 \u2013 oxygen  xiv VO2max \u2013 maximum aerobic capacity  RER \u2013 respiratory exchange ratio  [BLa] \u2013 blood lactate concentration  MVC \u2013 maximum ventilatory capacity  ATS \u2013 American Thoracic Society  ACCP \u2013 American College of Chest Physicians  Q \u2013 cardiac output  CaO2 \u2013 arterial content of oxygen  SV \u2013 stroke volume  HR \u2013 heart rate  PaO2 \u2013 partial pressure of arterial oxygen  Hb \u2013 hemoglobin  VE \u2013 minute ventilation  VT \u2013 tidal volume  EELV \u2013 end-expiratory lung volume  EILV \u2013 end-inspiratory lung volume  VA \u2013 alveolar ventilation  VC \u2013 vital capacity  AT \u2013 anaerobic threshold  RCP \u2013 respiratory compensation point  SpO2 \u2013 oxyhemoglobin saturation  Wmax \u2013 maximum workload  PaCO2 \u2013 partial pressure of arterial carbon dioxide  VE\/VCO2 \u2013 ventilatory equivalent for VCO2 IRV \u2013 inspiratory reserve volume  Pes\/PImax - ratio of tidal esophageal pressure to maximum inspiratory pressure xv EMGdi\/EMGdimax \u2013 ratio of diaphragmatic EMG amplitude to maximal diaphragmatic EMG amplitude  Pi \u2013 inorganic phosphate  ATP \u2013 adenosine triphosphate  O2ER \u2013 systemic oxygen extraction ratio PVR \u2013 pulmonary vascular resistance  ERS \u2013 European Respiratory Society  CP \u2013 critical power  PCr \u2013 phosphocreatine  ADP \u2013 adenosine diphosphate  K \u2013 potassium  CLT \u2013 constant load exercise trial  Tlim \u2013 time at the limit of tolerance during a constant load exercise trial  MLSS \u2013 maximal lactate steady-state Ti\/Ttot \u2013 ratio of inspiratory time to the total time of one breath  RPE \u2013 rating of perceived exertion  MDSS \u2013 maximal dyspnea steady-state  xvi Acknowledgements I would like to express my sincere gratitude to each of the following: My supervisor Dr. Neil Eves for his mentorship, guidance, and imparting the skills and knowledge that allowed me to become an independent and competent researcher and exercise physiologist. I feel extremely fortunate to have worked together for nearly a decade. You have helped shape me into the academic and person that I am today.   Drs. Guenette and duManoir, for serving on my PhD committee and for their advice and support throughout my doctoral studies, and Drs. Maltais and Road for agreeing to participate on my examination committee. Past and current lab members, specifically, Megan Harper, John Sasso and Steve Wright. Your collegiality and friendship made this PhD possible, and I\u2019m forever grateful for each of you.  Dr. Rolf and the personnel of the Respiratory Services department and Central Okanagan Pulmonary Rehabilitation program at the Kelowna General Hospital for their continued support of our research endeavours. I would also like to thank Wendy Wainwright and the electrophysiology technicians for their assistance. Gloria Agar and Bernie Melzer, these studies would not have been possible without your expertise and unwavering enthusiasm. Thank you for trusting me, and for being a dream team and making anything possible! Dr. Michael Roman, Andrew Kingscote, and Carolynn Meyer at the Rockyview General Hospital (located in Calgary, AB) for their collaboration.  The School of Health and Exercise Sciences at UBC Okanagan, as well as the Canadian Lung Association and the Canadian Respiratory Research Network for their financial support.  My family and friends, my parents and Ivan, for always believing in me and encouraging me to preserver.  Finally, thank you to all the patients who dedicated their time and efforts to these studies and prior ones. I am grateful for the many relationships that I have cultivated throughout my research activities.    xvii Dedication For Bernie, Your passion and dedication to your career, kindness towards others,  and exuberance for life are qualities I aspire to.    1 Chapter 1: Introduction 1.1 Overview of COPD 1.1.1 Prevalence of Disease Chronic obstructive pulmonary disease (COPD) is a progressive, yet treatable respiratory disease characterized by partially reversible, expiratory airflow obstruction and persistent respiratory symptoms.1 Although COPD is typically considered a smoking-related disease, long term exposure to noxious particulate matter such as air pollution, dust, chemical fumes and\/or pulmonary irritants can also trigger a proinflammatory response within the lungs that is linked with the pathogenesis of the disease.2 It is now acknowledged that several environmental, genetic and lifestyle risk factors that may be encountered at various stages of life (e.g., exposure to second-hand smoke in utero) can also contribute to the development of expiratory airflow limitation overtime.3  COPD is an increasing global health concern due to continued risk exposure and the increasing aging population. Currently, COPD is the 3rd leading cause of death worldwide for non-communicable diseases only surpassed by cardiovascular and neoplastic diseases.4 Within Canada, it is estimated that between 10-16% of the Canadian population meets the criteria for at least mild airflow obstruction.5 In the province of British Columbia alone, the prevalence of COPD is projected to increase from ~98,300 in 2010 to ~250,000 by 2030 representing a 155% increase with an annual growth rate of 4.8%.6 COPD accounts for the highest rate of hospital admissions among major chronic illnesses in Canada due to recurrent exacerbations, which is the largest contributor to the economic burden of COPD on the Canadian healthcare system.7 As such, interventions focused on disease management and slowing disease progression will be crucial in reducing the burden of COPD in the foreseeable future.  2 1.1.2 Pathophysiology  The term COPD encompasses emphysema and chronic bronchitis, both of which contribute to chronic airflow limitation.8 However, the disease specific characteristic associated with each pathology may progress at different rates over time leading to an ever-evolving, individual disease phenotype. In individuals who have a susceptible lung, the chronic inhalation of pulmonary irritants triggers an augmented inflammatory response within the lungs and pulmonary vasculature that can persist for many years even after removal of the irritant (e.g.,  smoking cessation).9,10 This chronic inflammatory response results in remodeling and narrowing of the smaller airways and\/or the breakdown of connective tissue components and initiation of repair mechanisms.11 Overtime, small airway obstruction and the repeated injury-repair cycle leads to the permanent structural remodeling and functional changes associated with the pathophysiology of COPD.11  Emphysema is characterized by destruction of the lung parenchyma that can be visibly assessed with computed tomography.12 Emphysema occurs due to the protease-mediated destruction of elastin as a consequence of the chronic inflammatory process.13 The degree and patterning of parenchymal destruction can vary from centriacinar (limited to the central part of the lobule while peripheral alveolar ducts and alveoli remain intact) to panacinar (destruction to the whole lobule).14 An enlargement of alveolar spaces and damage to the surrounding capillary bed reduces the surface for gas exchange. Additionally, the small airways and to a lesser extent larger airways may be narrowed due to thin and atrophied walls resulting in a reduction in airways generation.15,16 The reduction in alveolar attachments to the airways leads to a loss of radial traction of the alveoli, a reduction in airway patency, and reduced elastic recoil of the lung accompanied by an increase in compliance.17 A reduction in static recoil pressure results in lower flow rates relative to lung volume which can lead to the generation of an equal pressure point earlier during expiration.17 This predisposes the airways 3 to premature dynamic compression and\/or airway collapse before complete expiration resulting in pulmonary gas trapping.  Alpha-1-antitrypsin deficiency is the most common hereditary risk factor associated with the development of emphysema. It is caused by a genetic mutation that results in a reduction of alpha-1-antitrypsin, which is a primary serum and lung inhibitor of serine proteinase that protects the lung tissue from proteolytic destruction.18 Although the clinical phenotypes of this condition are highly variable, it generally presents at a younger age and disease progression is typically considered to be more rapid than \u2018usual\u2019 COPD.18 Augmentation therapy is recommended for certain patients as it has been has been demonstrated to reduce the progression of emphysema, particularly in patients with severe deficiency.18 Chronic bronchitis is defined as a persistent cough and sputum production for at least 3 months, in each of two consecutive years.19 Excess mucous production occurs due to an increased number of goblet cells and enlarged submucosal glands in the large bronchi.20 This mucous hypersecretion can lead to mucociliary dysfunction and mucous impactions causing occlusion of the smaller bronchi while increasing the predisposition to respiratory tract infections and exacerbations.21,22 Hypertrophy of the bronchial smooth muscle, increased cholinergic tone, cellular infiltration, edema, and the granulation of tissue in the airway walls due to chronic inflammation leads to further non-uniform narrowing of the airway lumen and increases turbulent airflow throughout the bronchial tree.11 Overtime, irreversible airway fibrosis and remodeling occurs leading to a permanent increased in airway resistance.11   1.1.3 Clinical Diagnosis  Fixed airway obstruction is determined based on the presence of resting expiratory airflow limitation (EFL) as assessed by spirometry, and together with smoking history (or chronic exposure to noxious particulate) and other diagnostic assessments (i.e., computed 4 tomography, bronchoalveolar lavage, sputum culture) confirms a diagnosis of COPD.1 Expiratory flow limitation is determined from the ratio of the volume of air exhaled during the first second of a maximal expiration (FEV1; forced expiratory volume in 1 second) to the volume of air forcibly exhaled from the point of maximal inspiration (FVC; force vital capacity). A post-bronchodilator FEV1\/FVC ratio of <0.7 is indicative of persistent EFL, however the fixed ratio value should also be compared to the lower limit of normal (LLN) defined as either the lowest 5% or a z score of -1.64 (the 5th percentile) from a normal distribution of a comparable healthy population.23 This increases diagnostic accuracy, especially in those with mild disease by distinguishing between a pathological obstruction and the normal age-associated decline in FEV1\/FVC.23 In recent years, the Global Lung Function Initiative (GLI) Network has produced the largest comprehensive database of measures of respiratory function world-wide. The GLI network reference equations are now considered the gold-standard when standardizing and reporting spirometry.24\u201326 The classification of COPD severity is based on the degree of post-bronchodilator FEV1 percent predicted and determined as: mild \u226580%, moderate 50-79%, severe 30-49%, and very severe <30%.1 In addition to spirometry, the Global Initiative for Chronic Obstructive Lung Disease (GOLD) has incorporated an additional ABCD assessment based on the modified Medical Research Council (mMRC) dyspnea scale and the COPD Assessment Test to provide additional information regarding symptom burden and exacerbation risk to help guide therapy.1   The pathophysiological changes associated with COPD manifest as an increase in airway resistance and lung compliance. However, the degree to which each of these characteristics occurs depends upon the clinical phenotype and extent of emphysema versus chronic bronchitis. The increase  in lung compliance associated with emphysema results in reduced lung elastic recoil pressure for any given lung volume, thus reducing driving pressure during expiration.17 Therefore, mid-low lung volume expiratory flow rates (which are effort independent) are often reduced. Reduced expiratory flow rates also occurs in chronic 5 bronchitis due to reduced airway diameter and increased airway resistance. Functionally this is characterized by a \u201cscooped\u201d appearance on the expiratory portion of the maximum flow volume loop.27,28   An increase in functional residual capacity (FRC; volume of air in the lungs at the end of a passive expiration), total lung capacity (TLC), residual volume (RV) and\/or the RV\/TLC ratio are indicative of static lung hyperinflation and although generally increase with worsening disease severity are also dependent on the severity of emphysema and\/or chronic bronchitis.29 Resting inspiratory capacity (IC; the maximum volume of air that can be inhaled from FRC during a normal tidal breath), is an indirect measure of lung hyperinflation that progressively declines as FRC shifts closer to TLC due to more severe expiratory flow limitation.29 The diffusion capacity for carbon monoxide (DLCO) and\/or the diffusion coefficient for carbon monoxide (DLCO\/VA, KCO) may also be reduced particularly in patients with a more emphysematous phenotype indicating reduced surface area for gas exchange.30 Additionally, significant alveolar ventilation-to-perfusion (VA\/Q) mismatching due to increased physiological dead space and a higher dead space to tidal volume ratio (VD\/VT) has been demonstrated across the range of disease severities in COPD.31,32   Although COPD is primarily considered a pulmonary disease, it is associated with multiple systemic manifestations that are thought to occur as a result of a \u2018spill-over\u2019 effect of inflammatory mediators from the lungs into the systemic circulation.33 Chronic low-grade levels of systemic inflammation are thought to initiate or worsen underlying comorbidities such as atherosclerosis, diabetes, osteoporosis, cancer and\/or anemia. It is estimated that up to two-thirds of patients with COPD die of non-pulmonary causes, mainly attributed to cardiovascular diseases and lung cancer.34 In particular cardiovascular disease is of major concern, as patients with COPD have a three to fourfold increase in mortality due to cardiovascular diseases independent of age, sex and smoking history.35 As such, the 6 physiological impairments associated with COPD are far reaching beyond the lungs and have major implications for the treatment and management of this disease.   1.1.4 Treatment and Management  COPD is associated with a variety of clinical manifestations including exertional dyspnea, exercise intolerance and reduced quality of life.36 Exertional dyspnea is not only a distressing acute symptom, but also plays a key role in mediating the long-term reduction in physical activity and functional capacity reported in COPD.37 The unpleasant sensation of exertional dyspnea can lead to a conscious and\/or subconscious reduction in physical activity levels resulting in further deconditioning and, thus greater dyspnea upon exertion. Overtime, this results in a self-perpetuating downwards dyspnea-spiral, whereby patients become more dyspneic during activities of daily living.38 Severe deconditioning due to activity avoidance in conjunction with disease progression, exacerbations, and co-morbidities can lead to the elimination of instrumental activities of daily living (e.g., shopping, household chores, cooking, leisure activities and maintenance of personal hygiene), and an eventual loss of functional independence. Therefore, the ultimately goal of treatment and management for patients with COPD is to reduce symptom burden, improve health related quality of life and extend the ability to live independently.   Smoking cessation (together with avoidance of exposure risk factors) and pharmacological therapy are the primary treatment and management goals for COPD. While smoking cessation has the greatest potential to reduce the rate of decline in FEV1,39 pharmacotherapy has been shown to improve lung function, exercise tolerance and health status as well as reduce the frequency and severity of exacerbations.40\u201342 A pharmacological treatment regimen should be individualized, however it is recommended that a short-acting \u03b22-angonist or short-acting muscarinic antagonist be prescribed for all patients as a rescue inhaler for immediate symptom relief.43 The addition of a long-acting \u03b22-angonsit, long-acting 7 muscarinic antagonist, inhaled corticosteroid or a combination of therapies is determined based on disease severity, the frequency and severity of exacerbations, symptoms and blood eosinophil counts (used for inhaled corticosteroid therapy).43 Additional pharmacotherapies such as phosphodiesterase-4 inhibitors, antibiotics (i.e., azithromycin, erythromycin), mucolytics, and statins may also provide added benefits in certain scenarios and are considered on a case-by-case basis.1   1.1.5 Pulmonary Rehabilitation  In conjunction with optimal pharmacotherapy, pulmonary rehabilitation is considered an integral component in the routine treatment and management of COPD. Pulmonary rehabilitation is a comprehensive intervention consisting of exercise training, education, and behavior change modification that is tailored to the specific needs of each individual patient.44 Substantial evidence supports that improvements in important clinical outcomes such as exertional dyspnea, exercise capacity and tolerance, heath-related quality of life, and reduced health care utilization are associated with pulmonary rehabilitation, despite no direct effect on lung function.45,46 Although these improvements are attributed to a combination of factors (e.g., disease specific education, smoking cessation, adherence to pharmacotherapy, and recognition and treatment of exacerbations), it is the physiological adaptations gained from exercise training that ultimately translate into an improvement in exertional symptoms and exercise capacity and tolerance. The primary adaptation thought to mediate these improvements is enhanced oxidative capacity of the skeletal muscle following aerobic exercise training. Increased oxidative capacity delays the onset of hydrogen ion (H+) and carbon dioxide (CO2) production from anaerobic glycolysis, thus reducing the ventilatory requirement at a given submaximal workload.47\u201349 This reduced ventilatory demand aids in 8 mitigating exertional symptoms (dyspnea and\/or leg fatigue), therefore increasing exercise tolerance by enabling patients to exercise at a higher intensity and\/or for a longer duration.  While many patients achieve physiological adaptations that translate into improvements in clinical outcomes, there is considerable heterogeneity in the magnitude of improvements reported between patients.50\u201354 In general, exercise training programs that achieve a greater training volume (longer durations and higher frequencies and\/or intensities) frequently report superior physiological training effects.47,55,56 A recent meta-analysis of 112 randomized controlled trials comparing aerobic training to usual care reported that a larger improvement in peak aerobic capacity (VO2peak) was associated with a greater training volume in patients with COPD.57 However, a novel inverse relationship between FEV1 and the change in VO2peak was also reported, even when controlling for baseline VO2peak. The authors interpreted this finding to suggest that in patients with more severe disease, a pathological ventilatory limitation may prevent the higher training intensities and\/or longer durations required for physiological adaptations compared to patients with mild to moderate disease.57 Evidence to support that training volume may be related to the degree of ventilatory limitation is provided by a retrospective analysis of 290 non-hypoxemic patients with COPD who were stratified according to their primary exercise limitation determined from an incremental cardiopulmonary exercise test (CPET).58 It was found that patients who obtained a cardiovascular limitation (i.e, peak heart rate (HRpeak) \u226580% of age-predicted maximum heart rate (HRmax)) achieved a greater improvement in VO2peak compared to patients with a primary ventilatory limitation (i.e., HRpeak <80%HRmax, and peak minute ventilation (VEpeak)\/ maximum voluntary ventilation (MVV) \u226580%). The authors proposed that patients with a primary ventilatory limitation may have not been able to exercise at a high enough intensity or for a long enough duration thus reducing the total training volume achieve and physiological adaptations gained.58 9 In further support of the constraining role of a ventilatory limitation, a previous rehabilitation study that randomized COPD patients to breathe either room air or helium-hyperoxia during exercise demonstrated that ventilatory limited patients who received helium-hyperoxia achieved larger improvements in exercise tolerance compared to room air,59 which were of similar magnitude to those reported in patients with a cardiovascular limitation.60 This was due to more optimal changes in lung volumes and an associated reduction in dyspnea during exercise training that enabled patients to exercise at higher absolute workloads for longer durations and thus accumulate a greater training volume.59,60 As important pulmonary rehabilitation outcomes appear to be related to the volume of exercise achieved, and the ability to achieve a greater training volume appears to be related to the primary exercise limitation in COPD, it is important to understand what factors contribute to exercise limitation in health, and how these variables may be altered due to the pathophysiological consequences of COPD.   1.2 Criteria to Determine Exercise Limitation Exercise limitation during whole-body exercise is traditionally determined by examining the physiological responses during an incremental CPET. As oxygen (O2) transport and utilization requires the coordinated integration of multiple physiological systems (namely the pulmonary, cardiovascular, musculoskeletal and nervous systems), exercise limitation cannot be reduced to a single factor. Instead, exercise limitation is based on the comprehensive analysis and interpretation of all exercise responses to determine the integrative contribution of each system to the attainment of maximum aerobic capacity (VO2max).   The determination of exercise limitation is dependent upon reaching a true maximal end point. In health, VO2max is considered to be attained once a plateau in VO2 occurs (\u2264150ml\/min increase despite an increase in workload) or secondary criteria is reached including a respiratory exchange ratio (RER) >1.15, HRpeak \u00b110 beats per minute of HRmax, 10 and\/or blood lactate concentration ([BLa]) \u22658mmol\/L.61 If necessary, VO2max can also be verified by performing a supramaximal exercise trial shortly after the initial test. In clinical populations, VO2max is often significantly reduced due to disease specific pathology and exercise may be terminated before the aforementioned criteria is reached.  However, at the point of exercise cessation, an atypical reduction in reserve (i.e., VEpeak \u226585% of maximum ventilatory capacity (MVC)) is often observed and accompanied by intolerable exertional symptoms (i.e., dyspnea and\/or leg fatigue) indicating that a maximal effort was given. Therefore, the term VO2peak is used instead but is still considered to be indicative of a true maximal end point for patients. In this scenario, assessment of exertional symptoms together with the integrated submaximal and peak exercise responses becomes important when determining factors that contribute to exercise limitation and thus VO2peak.   In a joint statement by the American Thoracic Society (ATS) and American College of Chest Physicians (ACCP), suggested guideline criteria have been proposed to assess the normality of incremental CPET responses.62 In addition to helping determine if the exercise responses observed are normal or abnormal, these criteria also provide reference cutoffs to indicate if a physiological system has reached or is close to reaching its maximum capacity. If a minimal reserve exists in a physiological system (i.e., HRpeak >90% HRmax), it is likely that system contributes to exercise limitation.   1.3 Exercise Limitation in Health  In healthy individuals undergoing whole-body exercise at sea level, VO2max is primarily limited by cardiac output (Q) as at maximal exercise, arterial oxygen content (CaO2) remains unchanged from resting levels, arterial-venous oxygen extraction increases to ~15-18ml\/dl, and a significant skeletal muscle metabolic reserve exists. Initial studies observed that the increase in Q at the start of exercise is accomplished by an increase in both stroke volume (SV) and the augmentation of heart rate (HR).63\u201367 However, at moderate-high intensities 11 (~50% of VO2max), Q increases almost exclusively due to an increase in HR as SV reaches a plateau.63\u201367 As such the attainment of HRmax is indicative of maximal convective O2 delivery. Further evidence to support that Q and convective O2 delivery limit VO2max comes from investigations that have observed that whole-body VO2max does not improve when arm crank exercise is added to two-legged cycling at close to maximal exercise, suggesting that O2 delivery is finite and Q has reached its maximum capacity at VO2max.68\u201370 Additionally, in a seminal study by Andersen et al., it was observed that VO2max measured in the isolated quadricep muscle group during single-leg knee extensor exercise was approximately 2-3 times higher than that measured in the same muscle group during whole-body exercise, indicating that when limitations to convective O2 delivery are removed mitochondrial O2 uptake increases.71 Collectively, these findings demonstrate that the skeletal muscle has a substantial metabolic reserve that far exceeds the convective O2 delivery capacity of the heart, and therefore VO2max in heath is limited by O2 supply to the mitochondria rather than the obtainment of the maximal rate of mitochondrial oxidative phosphorylation.  It is also important to acknowledge that the diffusive capacity of O2 transport also contributes to exercise limitation. In non-trained healthy individuals, diffusive O2 transport between the muscle capillary and mitochondria rather than the alveoli-capillary interface, represents an important site of limitation. There is a direct link between convective and diffusive O2 delivery as capillary O2 is determined by convective O2 transport, and O2 diffusion from the plasma into the interstitial space and sarcolemma is dependent upon the partial pressure gradient of arterial O2 (PaO2). Therefore, any change in PaO2 and thus capillary O2 has the potential to affect O2 diffusion. However, diffusive O2 transport is also dependent upon a number of factors including muscle capillarity (the area of apposition between capillary and muscle fiber and mean distance between capillary and mitochondria), capillary hematocrit, gas solubility, temperature, muscle myoglobin concentration and mitochondrial density.72\u201376 12 Out of these factors, it has been proposed that the distance between hemoglobin (Hb) and the muscle sarcolemma is considered to be the main site of diffusive O2 limitation in healthy.77   1.3.1 Exercise Limitation During Incremental CPET in Young Adults  During incremental exercise, Q increases linearly relative to VO2 to increase O2 delivery and meet the metabolic demands of the skeletal muscle. Earlier studies observed that the initial increase in Q is accomplished by an increase in both SV due to an increase in contractility and preload, and the augmentation of HR through parasympathetic withdrawal and increased sympathetic activation.63\u201367 However, at moderate-high intensities (~50% of VO2max) the increase in Q is almost exclusively due to an increase in HR as SV reaches a plateau due to a reduction in diastolic filling time.63\u201367 In elite athletes, SV may continue to rise throughout exercise due to beneficial training induced adaptations in cardiac volumes and geometry,78,79 however this response is rarely observed even in healthy, well-trained individuals.80\u201382 Measuring Q and SV accurately during exercise is challenging as although there are numerous techniques and methodologies that can be used, they are either technically difficult to perform (e.g., pulmonary artery catherization, inert gas techniques, echocardiography) or less accurate (e.g., cardioimpedance).  As such, Q and SV are not commonly measured directly during clinical incremental CPET. However, since the relationship between HR and VO2 is linear during incremental exercise and SV reaches a plateau in the majority of individuals, it is well accepted that when HRmax is attained Q and convective O2 delivery are maximal and thus VO2max has been or is close to being reached.63\u201367,80 Traditionally, a ventilatory limitation to exercise is determined by the relationship between ventilatory demand (i.e., VEpeak) to ventilatory capacity (typically determined either by performing a 12-second MVV maneuver or estimating MVC based on the calculation FEV1 * 35 or 4083). In healthy adults of average fitness, the pulmonary system does not contribute significantly to exercise limitation as the capacity of the pulmonary system is larger than 13 ventilatory demand and PaO2 remains tightly regulated at close to resting values even at VO2max.84 Therefore, despite a 10-15-fold increase in minute ventilation (VE) due to reflex-mediated hypernea, VEpeak typically reaches ~70% of MVC at VO2max representing a ventilatory reserve of >15% (calculated as the VEpeak\/MVC ratio).62  Studies performed in healthy participants that have challenged the pulmonary system by using added dead space or a small amount of inhaled CO2 during maximal exercise have observed an increase in VEpeak with no further increase in VO2max, providing further evidence that a ventilatory reserve does exists in healthy individuals.85\u201388 The rise in minute ventilation (VE) during an incremental CPET occurs due to both an increase in breathing frequency and tidal volume (VT). However, at low to moderate intensities VE increases primarily due to VT expansion at the expense of a reduction in end-expiratory lung volume (EELV) below resting values and an increase in end-inspiratory lung volume (EILV).89,90 These changes in lung volumes maintain tidal breathing on the linear portion of the respiratory system\u2019s sigmoidal pressure-volume relationship where compliance is greatest.90 This results in an efficient breathing pattern such that appropriate alveolar ventilation (VA) is maintained while the oxygen cost of breathing is minimized.90  At ~50-60% of vital capacity (VC), VT plateaus and breathing frequency increases at a steeper rate to continue to increase VE to meet metabolic demand until exercise is ceased.90 Alveolar ventilation and pulmonary gas exchange are sufficient to maintain a normal PaO2 during moderate to severe exercise (i.e., beyond the anaerobic threshold (AT) and respiratory compensation point (RCP)).84 Therefore, CaO2 and oxyhemoglobin saturation (SpO2) remain relatively unchanged up to VO2max demonstrating normal O2 delivery and the maintenance of blood gas homeostasis throughout incremental exercise.84  14 1.3.2 Exercise Limitation During Incremental CPET in Normal Aging  Aging is associated with intrinsic changes to the cardiovascular and pulmonary systems that affect incremental CPET responses. Although age does not significantly alter the Q-VO2 relationship, peak Q and HRpeak decrease with age.80 HRmax declines with age due to decreased sinoatrial node sensitivity to \u03b2-adrenergic stimulation.91 However, the relative contribution of the age-related reduction in HRmax to the decline in peak Q varies widely with studies reporting between 40-100%.92\u201394 The contribution of SV to the decline in Q is less clear, as although maximal SV has been shown to be reduced in sedentary individuals, this reduction may be offset with long term endurance training.92\u201394 However, despite these age-related changes, the HR-VO2 and SV-VO2 relationships during incremental CPET remain preserved such that SV plateaus at ~50%VO2max and HR continues to rise linearly until VO2max is reached.65,92\u201394 Therefore, similar to young adults, the ability to reach HRmax indicates that Q and convective O2 delivery contribute to exercise limitation in normal aging.    Ventilatory capacity is also reduced in aging due to age-related reductions in the elastic recoil of the lungs, chest wall compliance, and respiratory muscle strength.95 These alterations result in EFL (particularly at lower lung volumes closer to RV96), and the generation of an equal pressure point at lower expiratory rates. As such, when expiratory time is reduced such as during exercise, the normal lung volume response may be altered. End-expiratory lung volume may initially decrease to expand VT, but then rise to resting FRC or beyond due to pulmonary gas trapping.97,98 When EELV rises beyond resting FRC, dynamic hyperinflation is considered to occur. Minute ventilation is also increased for a given workload due to an age-related increase in physiological dead space, such that while the VD\/VT ratio decreases with exercise its nadir value remains higher compared to younger adults.99 Additionally, DLCO\/VA is reduced with aging due to a lower alveolar-capillary density to alveolar diameter ratio and decreased pulmonary capillary blood volume.100 While the combined effect of these 15 changes is an increase in VA\/Q mismatch, gas exchange is relatively still preserved as exercise induced hypoxemia is rarely observed in non-trained older adults.99   Age-associated changes in skeletal muscle properties also contribute to increased ventilatory demand during exercise. Reduced peripheral muscle elasticity, increased muscle fibrosis, and atrophy of type ll fibers accompanied by type IIa fiber denervation have been documented, particularly in populations \u226570 years of age.101,102 Additionally, reduced mitochondrial content of type I and II fibers, and reduced mitochondrial oxidative capacity that is not fully reversible with training has also been demonstrated with aging.103,104 Together, these age-related changes in peripheral muscle structure and function contribute to reduced muscular efficiency and an increase in ventilatory demand for a given submaximal workload. However, an accompanying reduction in maximal workload (Wmax) and VO2max due to these changes ultimately results in a reduced aboslute VEpeak at maximal exercise compared to younger adults. In sedentary individuals VO2max naturally declines at a rate of ~10% per decade after 30 years of age.105 Since the typical lifelong decline in FEV1 is roughly proportional,96 MVC is therefore also reduced. As such, in normal healthy aging a ventilatory reserve typically exists at maximal exercise but absolute VEpeak is lower due to the reduced oxidative capacity of the skeletal muscle.  As such, the ability to reach HRmax in the presence of a ventilatory reserve further supports convective O2 delivery significantly contributes to exercise limitation in normal aging. However, it is important to recognize that in older women and well-trained older adults, the age-related increase in EFL and changes in airway and  lung mechanics may play a larger role in exercise limitation due to dysanapsis and\/or increased ventilatory demand due to a greater oxidative capacity of the skeletal muscle.106\u2013109 In these individuals, a mechanical constraint to ventilation may contribute to a greater degree to overall exercise limitation.  16 1.4 Exercise Limitation in COPD Due to the heterogeneity in the pathophysiology and symptomology of COPD, a range of exercise responses have been observed during incremental CPET in these patients. As COPD is primarily a pulmonary disease, gas exchange abnormalities and altered lung mechanics during whole-body exercise have been the focus of many investigations. In general, increased disease severity has been associated with reduced exercise capacity and intolerance as disease progression is associated with increased static hyperinflation and a progressive reduction in IC that results in altered ventilatory mechanics and intolerable exertional symptoms at a lower VO2peak, Wmax and\/or VEpeak.54,110\u2013112 However, FEV1 remains poorly correlated with VO2peak indicating that contributions to exercise limitation extend beyond the severity of airflow obstruction in COPD. As such, while it may seem intuitive that patients with COPD would be predominantly limited by the pulmonary system during exercise, a large body of evidence now supports a contributory role for skeletal muscle function. Additional studies also exist providing evidence for the role of the cardiovascular system in exercise limitations in this patient population, however they are much less numerous.   1.4.1 Ventilatory Contributions to Exercise Limitation In COPD, ventilatory capacity is decreased due to EFL as a consequence of reduced lung elastic recoil pressure and increased airway resistance. However, ventilatory demand is increased for a given workload due to the interaction between various chemical, mechanical and neural factors during exercise. The imbalance between increased ventilatory demand and reduced capacity manifests as a reduction in ventilatory reserve such that VEpeak\/MVC is \u226585% and\/or MVC - VEpeak is \u226411L\/min at peak exercise while a significant cardiac reserve (i.e., HRpeak is well below HRmax) is usually maintained.62 In moderate-to-severe COPD, the inability to continue to efficiently increase VE as exercise intensity progresses results in a lower absolute VEpeak and contributes to the reduced Wmax and VO2peak compared to healthy 17 controls; although, these differences occur to less of a degree in mild COPD.113\u2013120 As such, ventilatory limitation is generally considered the primary contributor to exercise limitation during whole-body aerobic exercise in non-hypoxemic COPD. Interestingly, a greater ventilatory reserve (i.e., VE\/MVC <80%) has been reported in a subset of patients with COPD,121 and as such, the incorporation of additional physiological measures such as lung volumes, intrathoracic pressures and diaphragmatic electromyography (EMGdi) during incremental CPET have provided a more detailed assessment of the origins of ventilatory constraint during exercise.    During exercise, multiple integrated factors contribute to the increased ventilatory demand in COPD. Due to an increase in physiological dead space, VD\/VT is higher at rest and the magnitude of reduction in VD\/VT during exercise is less compared to healthy age-matched controls.118,122,123 As such, ventilation must increase to maintain an appropriate VA and blood gas homeostasis. However, as VE increases during exercise, patients often adopt a rapid and shallow breathing pattern due to rising lung volumes and VT constraint.110 This breathing pattern further increases anatomical dead space ventilation, which may in turn lead to an abnormally widened alveolar-to-arterial oxygen gradient, hypoxemia and\/or hypercapnia.124,125  A reduction in PaO2 and\/or increase in the partial pressure of arterial CO2 (PaCO2) will contribute to an increased drive to breathe through central and peripheral chemoreceptor stimulation.   The utilization of anaerobic glycolysis as a greater contributor to energy production commonly occurs earlier during incremental exercise in patients with COPD due to alterations in skeletal muscle structure and function and\/or a low Q.114,126\u2013129 This results in a greater production of metabolic by-products, accumulation of H+, and CO2 produced (VCO2) for a given workload, which further contributes to an increased drive to breathe from chemoreceptor stimulation.127,130\u2013132 Consequently, changes in the nadir, intercept, and slope of the ventilatory equivalent for CO2 (VE\/VCO2) have been reported in COPD.133 The VE\/VCO2 18 nadir reflects the onset of frank compensatory hyperventilation for metabolic acidosis with a concomitant reduction in PaCO2 (criteria often used to indicate the RCP134). In COPD, the VE\/VCO2 nadir often occurs at a lower workload but is increased due to a high VD\/VT.133 However, in patients with more severe disease exercise may be terminated before a true nadir is reached. The VE-VCO2 intercept increases with disease severity while the VE-VCO2 slope is increased in mild COPD and progressively decreases as disease severity progresses.133 In severely ventilatory limited patients who experience significant mechanical constraint to ventilation due to severe static and\/or dynamic hyperinflation, the VE-VCO2 slope is often reduced demonstrating an inability to continue to increase VE despite an increase in VCO2.133  Increased type III & IV afferents also contribute to the increased drive to breathe during exercise in COPD.  Type III & IV afferent receptors are activated by mechanical distortion (i.e., mechanoreceptors) and metabolic chemical stimuli (i.e., metaboreceptors) arising from the respiratory and skeletal muscles during exercise.135 Increased type III & IV afferent signaling is directly relayed via the dorsal horn of the spinal cord to the nucleus tractus solitarius and other cardiorespiratory controller neurons of the ventral lateral medulla, and stimulates a reflexive increase in ventilation, blood pressure, and HR (i.e., the metabo\/mechanoreflex) to ensure adequate blood flow and O2 delivery to the exercising muscles.136 In patients with moderate to severe COPD, when type III & IV afferents were attenuated during constant load exercise by using a spinal blockade of intrathecal injection of fentanyl, constant load exercise time was significantly increased.131 This was associated with an attenuated VE responses throughout exercise despite similar VO2, VCO2, and lactate kinetics compared to exercise without the spinal blockade.131 The attenuation in VE was due to a reduction in breathing frequency and was associated with a larger IC and reduction in iso-time leg fatigue and dyspnea. However, at exercise cessation, VEpeak, inspiratory reserve volume (IRV) and exertional symptoms were similar between trials.131 These findings 19 demonstrated a direct contribution of type III & IV afferents to the increased drive to breathe in COPD that plays an important role in constant load exercise tolerance in these patients.  Altered lung mechanics during exercise contributes to both a reduced ventilatory capacity and increased ventilatory demand in COPD. Due to EFL, reduced expiratory time during exercise commonly leads to dynamic hyperinflation. In patients with mild to moderate disease, the initial rise in EELV during low-moderate intensities can aid in increasing airway conductance and expiratory flow rates. However, as exercise progresses and ventilatory demand increases, expiratory time is reduced and EELV continues to rise above resting FRC. The rate of dynamic hyperinflation can be measured indirectly as the reduction in IC,137 which also provides an indication of the operating position of VT relative to TLC.   As a result of the rise in EELV, EILV is also pushed closer to TLC reducing IRV in an effort to maintain VT expansion.110,118,120,138,139 As a result of these dynamic changes in lung volumes, VT becomes positioned on the upper alinear portion of the respiratory system\u2019s pressure-volume relationship and respiratory system compliance is greatly reduced.140 At this point, VT is considered to be mechanical constrained and an inflection in breathing frequency occurs to try to continue increase VE to meet metabolic demand. 110,118,120,138,139 An increase in breathing frequency further exacerbates dynamic hyperinflation ultimately reducing the capacity to continue to increase VE any further. 110,118,120,138,139 Dynamic hyperinflation occurs due to the initiation of inspiration before complete expiration. As such, alveolar volume remains elevated at the initiation of inspiration resulting in an increase in intrinsic positive end-expiratory pressure (iPEEP) which must be overcome before inspiratory flow can start. As such, higher levels of iPEEP greatly increases the inspiratory work required by the respiratory muscles.141 Additionally, as dynamic hyperinflation worsens the diaphragm becomes flattened and shortened which alters the length-tension relationship and mechanically disadvantages the diaphragm to produce force.142 This results in a greater level of neural activation required to generate the inspiratory pressures and flows 20 needed to produce adequate VA.120,143 Due to the flattened diaphragm\u2019s reduced ability to generate adequate pressure and flow, demand on the accessory muscles of inspiration is increased and eventually they too become shortened and mechanically disadvantaged.144,145 As such, significant swings in intrathoracic pressure (e.g., ~-20 cmH2O to  ~+20cmH2O during exercise) and an associated increase in the oxygen cost of breathing for a given absolute VE and\/or workload is often observed during exercise in COPD.138,139,143,146\u2013149 The imbalance between increased ventilatory demand and reduced ventilatory capacity (especially with constraint to VT expansion), ultimately results in intolerable dyspnea and exercise cessation at a reduced Wmax and VO2peak. It has been documented across the spectrum of COPD disease severities that the point at which the VT\/VE relationship reaches a plateau, coincides with when VT\/IC reaches ~70%, EILV \u226590% of TLC, and IRV reaches a critical minimum volume ~500-600ml.110,138,139 At this point, an inflection in the IRV-dyspnea relationship is observed and represents the point beyond which there is a widening disparity between the increased neural drive to breathe (from afferent sensory stimuli) and the mechanical\/muscular response of the respiratory system (the efferent activity of the respiratory muscles).117,138,139 Respiratory effort (determined as the ratio of tidal esophageal pressure to maximum inspiratory pressure [Pes\/PImax]) continues to rise as VT expansion (expressed as the VT\/VC ratio) remains stable.117,138,139 Therefore, the ratio of inspiratory effort to VT displacement (i.e., Pes\/PImax: VT\/VC) increases as VE continues to rise throughout exercise, indicating neuromechanical uncoupling of the respiratory system.117,138,139 Similar observations have been documented with the use of diaphragmatic EMG (expressed as EMGdi\/EMGdimax) as a surrogate for respiratory neural drive.120,150\u2013152  1.4.2 Skeletal Muscle Contributions to Exercise Limitation Alterations in lower limb skeletal muscle structure and function have been documented across the COPD disease spectrum and are associated with clinical and prognostic outcomes.153\u2013155 21 A contributing role for altered skeletal muscle function to exercise intolerance in COPD has been demonstrated by the proportion of patients who report leg fatigue as the main exertional symptom at peak exercise,116 the persistence of exercise intolerance following double lung transplant,156 and the lack of improvement in constant load exercise time when quadriceps fatigue is present despite bronchodilator induced improvements in lung function.157 The underlying mechanisms of skeletal muscle dysfunction in COPD are complex with numerous etiologies including systemic inflammation, redox balance, nutritional status, exacerbations, hypoxemia\/hypercapnia, corticosteroid use, comorbidities and chronic physical inactivity.158 Debate exists as to whether skeletal muscle dysfunction is largely the product of deconditioning due to reduced physical activity levels or if a specific myopathy intrinsic to COPD exists. Regardless, unfavorable morphological and functional changes have been documented in this patient population including atrophy of both type-I and type-II fibers, a reduction in the proportion of type-I fibers accompanied by a shift in fiber type distribution towards glycolytic type-IIx fibers, reduced oxidative capacity, decreased capillarization and surface area, and reduced mitochondrial density and function.114,159\u2013162 Although there is large heterogeneity in the degree to which these changes occur between patients, the collective result is a reduction in muscle strength and increased fatiguability which contribute to exercise intolerance and a reduced VO2peak.  Quadriceps muscle strength measured by volitional (i.e., isometric maximal voluntary contraction) and non-volition (i.e., magnetic stimulation of the femoral nerve) methods has been demonstrated to be reduced by ~20-30% in COPD.163\u2013166 However, when normalized to cross sectional area or mass, the difference in muscle strength is minimized between COPD and controls suggesting that the reduction in muscle strength is related to skeletal muscle atrophy.163\u2013166 Interestingly, large interindividual heterogeneity in muscle weakness has been reported when categorizing patients based on FEV1, such that ~50% of patients with severe disease in whom muscle weakness was expected did not show reduced quadricep 22 strength.167 As such, the concept of a limb muscle dysfunction phenotype in COPD has been proposed.   The shift in skeletal muscle fiber-type distribution away from fatigue resistance slow-twitch oxidative type-I fibers, and towards fatigue susceptible fast-twitch glycolytic type-IIx fibers is an important contributor to the increased leg fatiguability and reduced exercise tolerance in COPD.159 The reported reduction in oxidative enzymes (e.g., citrate synthase, 3-hydroxyacl-CoA-dehydrogenase) and increased activity of glycolytic enzymes (e.g., phosphofructokinase, lactate dehydrogenase) is consistent with this fiber shift.114,127,168\u2013170 As such, oxidative metabolic pathways such as the Krebs cycle and \u03b2-oxidation are downregulated. Earlier reliance on anaerobic glycolysis results in a greater accumulation of metabolic by-products at lower workloads. The accumulation of H+ contributes to the susceptibility of fatigue by interfering with excitation-contraction coupling and ventilatory demand through chemoreceptor stimulation, and together with other addition metabolic by-products (e.g., inorganic phosphate (Pi))  increases type III & IV afferents.130,131,170 Additionally, reductions in mitochondrial density, enzyme activity and dysfunction further contribute to decreased muscle oxidative metabolism and have been associated with increased oxidative stress, which can lead to protein catabolism and apoptosis.161,162,171  While it is generally agreed that mitochondrial density and function is altered in COPD, the underlying reason for this and the severity of dysfunction between patients is less clear. Previous studies have suggested that an underlying myopathy occurs in COPD as reduced citrate synthase activity, lower mitochondrial respiration, and reduced adenosine triphosphate (ATP) production per unit of fiber weight has been observed from muscle biopsies of patients compared to healthy controls.161,162,172 However, in other studies that have reported mitochondrial respiration per unit of citrate synthase and thus assessed respiratory function per mitochondria, no differences in cellular respiration were found between patients and healthy controls.173,174 This finding was interpreted to suggest that the reduced oxidative 23 capacity observed may be due to a lower mitochondrial volume mediated through a reduction in physical activity levels rather than due to intrinsic abnormalities. However, as more severe alterations in skeletal muscle function have been associated with an increase in secondary systemic complications (e.g., low body-mass index, hypoxemia, chronic low dose steroid use) that occur with disease progression, an intrinsic myopathy may exist in patients with more advanced disease.172,175,176 Additional evidence to support that intrinsic skeletal muscle function may be preserved in some individuals with COPD is provided by a study by Richardson et al., that compared exercise responses during whole-body cycling and single-knee extensor exercise while breathing room air, hyperoxia or heliox in patients with severe COPD.177 The authors observed that a greater muscle mass-specific power was achieve with knee extensor exercise and that exercise in hyperoxia resulted in a higher Wmax achieved during both whole-body cycling and knee extensor exercise.177 These findings suggested that O2 supply, and thus convective O2 delivery plays an important role in limiting maximal work capacity in COPD. The authors also reported that the increase in Wmax during knee extensor exercise was augmented in hyperoxia compared to normoxia, and that breathing heliox increased Wmax during whole-body cycling but not during knee extensor exercise.177 These findings were attributed to an improvement in blood flow distribution during whole-body exercise as a result of a reduced work of breathing due to heliox. However, as pulmonary ventilation was not constrained during knee extensor exercise heliox did not have the same affect during smaller muscle mass exercise. Collectively, these findings are consistent with those previously observed in health individuals and demonstrate that a skeletal muscle metabolic reserve exists in COPD.71  While the aforementioned findings suggest that whole-body exercise in COPD is limited by convective O2 delivery and not skeletal muscle function per se, it is likely that the negative consequences of reduced oxidative capacity are further aggravated by a primary 24 reduction in convective O2 delivery. In a study by Oelberg et al., the systemic oxygen extraction ratio (O2ER, measured as [arterial-mixed venous O2 content different\/CaO2]) was reduced during maximal whole-body cycling in patients with very severe COPD (25\u00b11%pred FEV1) compared to both controls and a small group of patients with heart failure.128 In the COPD group, the lactate threshold occurred at a lower workload and was correlated with the reduced O2ER at VO2peak.128 As such, the authors suggested that the early onset of H+ production was related to abnormal O2 extraction. However, as peak Q was also reduced compared to controls (~11L\/min vs. ~17L\/min), reduced convective O2 delivery likely also contributed to this response. In a subsequent study performed by the same group, incremental CPET responses in very severe COPD (19\u00b11%pred FEV1) were compared while breathing room air or heliox.178 Mixed venous O2 remained elevated and peak Q remained severely reduced (~5.6L\/min in normoxia versus ~5.9L\/min in heliox) regardless of an increase in VEpeak and VO2peak with heliox.178 These findings demonstrated that despite respiratory muscle unloading with heliox, abnormalities of O2 transport and utilization still contributed to the abnormally low VO2peak in severe COPD. While the mechanisms for reduced systemic O2 extraction could not be deduced from this study, the authors acknowledged that it may have occurred due to a mismatch between perfusion and skeletal muscle metabolism, an inherent defect of skeletal muscle oxidative capacity or both. In a more recent investigation performed by Broxterman et al., using single-knee extensor exercise in severe COPD, both convective and diffusive O2 delivery were diminished resulting in a reduction in leg VO2peak compared to controls.179 The authors postulated that the reduction in diffusive O2 transport was due to lower Hb concentration, and\/or an increase in type-II fibers and decreased muscle-capillary interface. Taken together, it appears that systemic oxygen extraction and diffusive O2 delivery are impaired (particularly in more severe COPD), and that this may be due to reduced capillarization and surface area, reduced perfusion of the skeletal muscle and\/or intrinsic oxidative myopathies.  25 1.4.3 Cardiovascular Contributions to Exercise Limitation Relatively less attention has traditionally been given to understanding the cardiovascular contributions to exercise limitation in COPD despite Q and O2 delivery being a critical element governing maximal oxygen consumption. Although the linear relationship between Q and VO2 appears preserved in COPD, peak Q is generally reduced compared to healthy controls.128,178,180\u2013183 In patients without overt cardiovascular disease, the reduction in peak Q is traditionally considered to be due to early exercise cessation before \u2018true\u2019 maximum Q is reached due to a ventilatory limitation and\/or peripheral muscle fatigue. Therefore, HRpeak is usually well-below HRmax and a significant cardiac reserve exists (i.e., there is a significant ability to continue increasing HR).62 However, some patients have been documented to reach close to HRmax during incremental CPET.58,184 This observation indicates that a true maximum Q is attained in some patients, and as such convective O2 delivery is a limiting factor to exercise. The HR response to incremental exercise is typically linear in patients with COPD, however the slope of the HR-VO2 relationship may be increased likely due to an attenuated submaximal SV response.180,185 The exercise SV response is more varied in patients with COPD compared to healthy individuals such that SV may rise but the response may be attenuated.128,178,180\u2013183,186,187 In some individuals (i.e., those that reach HRmax), SV may be attenuated due to systemic deconditioning (e.g., smaller cardiac volumes, decreased blood volume, reduced myocardial compliance and increased vascular afterload). In others with a more classic ventilatory limitation to exercise, the reduced SV during exercise may also be due to adverse heart-lung interactions as a result of greater levels of static and\/or dynamic hyperinflation and large intrathoracic pressure swings.  The large negative swings in intrathoracic pressure associated with dynamic hyperinflation have been proposed to alter the exercising SV response in COPD. Several studies have made an inference to this relationship through a correlation between the reduction in IC\/TLC ratio (indicating more severe dynamic hyperinflation and thus larger 26 swings in intrathoracic pressure) and a low peak O2pulse (a surrogate for SV, calculated as VO2\/HR) during incremental exercise in COPD.188\u2013191 However, dynamic hyperinflation itself has also been demonstrated to affect hemodynamics (i.e., reduced left ventricular filling and right and left ventricular SV) by increasing pulmonary vascular resistance (PVR) and thus right ventricular afterload through the direct effect of elevated alveolar pressure which acts to compress alveolar capillaries at high lung volumes.149,192 Additionally, dynamic hyperinflation increases cardiac transmural pressure thereby acting as a mediastinal constraint to the heart which may impede ventricular compliance and further reduce left ventricular filling.193 The influence of negative intrathoracic pressure swings and dynamic hyperinflation on the exercise SV response is further complicated by the phasic effects of the respiratory and cardiac cycles, differences in breathing pattern, rate of dynamic hyperinflation between patients, and the exercise mediate increases in Q and venous return.  As such, even in patients who have a ventilatory limitation to exercise due to the well-documented adverse lung mechanics and significant dynamic hyperinflation, a reduction in SV at submaximal intensities likely attenuates convective O2 delivery due to adverse heart-lung interactions, thus contributing to exercise limitation.  The pathophysiological changes in the pulmonary vasculature that are associated with the development of COPD also contribute to the attenuated SV response. The emphysema-related destruction of the cross-sectional area of the pulmonary vascular bed and reduced alveolar-capillary interface limits the normal distension and recruitment of pulmonary capillaries during exercise in COPD.194 This reduces pulmonary vascular compliance and increases PVR.194 Additionally, reduced alveolar PO2 due to increased physiological dead space and gas trapping can result in hypoxic pulmonary vasoconstriction to improve VA\/Q matching.195 Collectively, these factors can contribute to a disproportionate rise in PVR and\/or pulmonary artery pressure that significantly increases the afterload of the pressure-sensitive right ventricle thus impairing right ventricular SV.181,187,196\u2013199 Under conditions of altered lung 27 mechanics due to increased dynamic hyperinflation, a reduction in right ventricular SV can act to amplify heart-lung interactions via series and\/or direct ventricular interaction that can lead to reduced left ventricular filling and Q.200,201 Another means by which convective and diffusive O2 delivery can be reduced in non-hypoxemic patients is anemia and reduced Hb concentration.  As the majority of O2 is transported bound to Hb (~20ml\/dL) compared to dissolve in plasma (~0.3ml\/dL), anemia will have a significant impact on O2 delivery and tissue availability even when O2 content measured by PaO2 or SpO2 is adequate.  Anemia is estimated to occur in ~20-30% of the COPD population,202 and although the pathogenesis is not fully understood it is thought to be related to an increase in systemic inflammatory mediators that affect iron homeostasis, Hb, and the bone marrow response to erythropoietin.203 In studies comparing anemic and non-anemic patients with moderate to severe COPD, anemic patients achieved a lower VO2peak and Wmax, a lower workload at the anaerobic threshold, and higher dyspnea scores compared to non-anemic patients.204,205 Authors have interpreted these findings to suggest that an earlier reliance on anaerobic glycolysis due to an overall reduction in gross O2 delivery, resulted in increased H+ and CO2 production at a lower workload, thus increasing ventilatory demand leading to a ventilatory limitation and exercise cessation at a lower Wmax.   Considerable evidence also supports a cardiovascular contribution to exercise limitation in COPD through a re-distribution of blood flow away from the exercising peripheral muscles and towards the respiratory muscles. In the presence of a ventilatory limitation due to adverse change in lung mechanics and increased work of breathing, the oxygen cost of the respiratory muscles dramatically increases as dynamic hyperinflation worsens.139,141,146\u2013148 As Q is finite, this creates a competition for blood flow between the respiratory muscles and exercising peripheral muscles. In certain patients, respiratory muscle \u201csteal\u201d may occur resulting in a re-distribution of blood flow away from the peripheral muscles predisposing the peripheral muscles to fatigue at lower exercise intensities, and thus contributing to exercise 28 cessation. This phenomenon was proposed following the observation that in certain patients with severe COPD undergoing incremental CPET, leg VO2, Q and arterial-venous O2 difference plateaued while limb vascular conductance increased despite a continued increased in whole-body VO2 and Q.129 Subsequent studies that used heliox to reduce the work of breathing during high intensity constant load exercise supported this hypothesis by demonstrating an increase in exercise time and quadricep blood flow during the heliox trials.206,207 Further work performed by Amann et al., more directly demonstrated the link between increased work of breathing and peripheral muscle fatigue during constant load exercise in COPD.208 It was reported that non-voluntary quadriceps muscle fatigue was attenuated by ~1\/3 during high-intensity constant load exercise following various intervention that reduced the work of breathing and\/or increased SpO2 (i.e., proportional assist ventilation plus heliox, heliox, hyperoxia, or hyperoxic-heliox).208 Taken together, these studies demonstrate that the altered ventilatory mechanics that are classically associated with exercise limitation in COPD can contribute to a cardiovascular contribution to exercise limitation via a re-distribution of Q and O2 delivery away from the exercising peripheral muscles.    1.4.4 Integrated Exercise Response in COPD It is acknowledged that the underlying mechanisms that result in an altered ventilatory, metabolic, and\/or cardiovascular response to exercise in COPD are complex and multifactorial (as discussed above). Therefore, the degree to which these factors contribute to exercise limitation is unique to each individual patient. However, despite this acknowledgement, it is traditionally accepted that VO2peak is predominantly limited by the pulmonary system in COPD as indicated by the abnormal changes in lung volumes and\/or gas exchange abnormalities. Yet, when utilizing incremental CPET, a wide range in ventilatory, metabolic, and cardiac responses at submaximal and peak exercise have been 29 reported between patients. As such, the assumption that all patients are primarily limited by the pulmonary system appears to be an oversimplification. For example, although rarely acknowledged within the literature, it has been documented that certain patients with COPD are able to reach HRmax with or without the presence of a ventilatory reserve at VO2peak.58,184 This finding suggests that in a subset of patients, Q and convective O2 delivery are maximized (once HRmax is reached), indicating a predominant cardiovascular limitation to exercise similar to that observed in healthy individuals. Therefore, it appears that within the COPD population, certain patients may present with a primary ventilatory limitation, a primary cardiovascular limitation or a combined limitation (consisting of both a ventilatory and cardiovascular limitation).  Incremental CPET is a clinical important tool in the risk stratification of patients due to the integrative assessment of physiological responses. A lower VO2peak has been associated with increased mortality in COPD209,210 and correlated with other disease characteristics such as hypoxemia, IC\/TLC ratio and VE\/VCO2 nadir.211,212 However, incremental CPET may also be used to distinguish subgroups of patients with unique disease characteristics (i.e., phenotypes). Investigating different phenotypes of exercise limitation based on traditionally accepted classification criteria may offer more insight into the integrative physiological factors that ultimately govern VO2peak in patients with COPD. To date, delineating different exercise limitation phenotypes in COPD and their associated physiological exercise responses has not been explored. However, identifying these phenotypes may provide novel prognostic utility to distinguish clinically important disease characteristics (i.e, a significantly reduced VO2peak). As such, Study #1 of this dissertation aimed to determine the distinct ventilatory, cardiovascular, and metabolic responses to incremental CPET in patients with COPD who present with either a ventilatory, cardiovascular or combined exercise limitation phenotype.  30 1.5 Exercise Prescription in COPD The current ATS\/ European Respiratory Society (ERS) aerobic exercise prescription recommendations (i.e., >60%Wmax for >20 minutes of continuous aerobic exercise) are based on evidence of a beneficial effect on important outcomes such as exercise tolerance, exertional dyspnea and health related quality in the majority of patients with COPD.44 However, closer inspection of studies assessing outcomes following pulmonary rehabilitation demonstrate that there is considerable heterogeneity in the magnitude of responses between patients following exercise training.50\u201354 One reason for this may be an inability to achieve the required duration and\/or intensity of aerobic exercise even after completing 6-12 weeks of pulmonary rehabilitation. As the ability for patients with COPD to achieve a higher training volume (i.e., higher intensity and\/or a longer duration) may to be related to the primary exercise limitation,57\u201360 classifying patients according to exercise limitation phenotype may help explain the high variability in the relative intensity that patients can sustain continuously for 20-30 minutes. For example, it may be that patients with a cardiovascular or combined limitation are able to attain a higher Wmax and\/or VO2peak during an incremental CPET compared to their ventilatory limited counterparts in whom exercise is likely to be curtailed earlier due to static and\/or dynamic hyperinflation resulting in VT constraint and ensuing dyspnea. Additionally, other physiological factors that govern O2 delivery and utilization such as the oxidative capacity of skeletal muscle, blood flow re-distribution and the ability to augment Q may also differ depending on the primary exercise limitation, further contributing to the individual variance in exercise tolerance.  In patients with COPD, it has been demonstrated that when VT constraint occurs (i.e., VT\/IC ~70%, EILV\/TLC \u226590%, and IRV is reduced to ~500ml) exercise duration becomes finite due to an inflection in the IRV-dyspnea relationship.110,117,138,139,213 Therefore, it would seem logical that the highest exercise intensity at which patients would be able to sustain continuous exercise (and thus accumulate the largest training volume) would be slightly below this point. 31 While this may be true for patients with a ventilatory limitation, the use of this inflection point to determine the highest sustainable exercise intensity may be problematic in those with a cardiovascular or combined limitation, as changes in lung volumes are unlikely to be the primary mechanism to exercise limitation. In these patients, exertional dyspnea may not be the primary limiting symptom and\/or may be associated with different causes other than dynamic hyperinflation and VT constraint. Thus, exercise intolerance may be due to limitations in convective O2 delivery and the associated metabolic perturbations that occur at higher exercise intensities. Determining this critical intensity is necessary for optimizing the volume of training achieved during an exercise program and could be used as an important benchmark to help guide exercise prescription in pulmonary rehabilitation programs.  1.6 The Power-Duration Relationship One method commonly used to differentiate sustainable from non-sustainable exercise intensities is the hyperbolic power-duration relationship. This relationship describes the ability to sustain a particular power output as a function of time and is governed by 2 parameters: critical power (CP) and W\u2019. Critical power represents the maximal workload (i.e., watts) or metabolic rate (i.e., %VO2peak) that separates sustainable intensities form those that are predictably limited.214\u2013219 In healthy individuals, CP is largely determined by changes in aerobic versus anaerobic metabolic contributions and the accumulation of fatigue-related metabolites such that during exercise above CP, steady-state is not achieved and VO2 and [Bla] continue to rise to maximal values.216,220\u2013222 Therefore, CP is considered to demarcate the upper boundary of the heavy exercise intensity domain and the lower boundary of the severe intensity exercise domain.216,221,223 W\u2019 represents a finite amount of work that can be performed above CP that is independent of rate. W\u2019 is regarded as a composite of several factors including anaerobic energy contributions (i.e., intramuscular phosphocreatine (PCr) and glycogen stores), the accumulation of fatiguing metabolites (e.g., Pi, H+, adenosine 32 diphosphate (ADP), and potassium (K)), and the development of the VO2 slow component.221,222,224\u2013228 Numerous methodological approaches have been used to determine CP and W\u2019, each producing slightly different estimates of both parameters.229\u2013235 However, traditionally, CP and W\u2019 are determined by performing 3-5 high-intensity constant load exercise trials (CLTs) ranging between ~80-120%Wmax that result in a time at the limit of tolerance (Tlim) between ~2-15 minutes. From the original hyperbolic relationship, CP is determined as the asymptote of the hyperbola and W\u2019 is the curvature constant.215 However, this hyperbolic relationship is commonly transformed into a linear relationship by plotting the corresponding workloads and Tlims, expressed as 1\/time.214 In this linear model, CP is calculated as the y-intercept and W\u2019 is represented by slope of the line (Figure 1.1).  Mathematically, the power-duration relationship can be used to predict the highest workload that can be sustained for a specific duration or vice versa. As such, the CP concept has been implemented across multiple sporting disciplines (e.g., cycling, running, swimming, rowing etc.) to predict time-trial and time-to-exhaustion performance, monitor training progress and differentiate exercise intensity domains.236\u2013240 Therefore, by using the individually calculated CP and W\u2019 it is possible to differentiate workloads that can be sustained for a longer duration (i.e., below CP) versus those that are finite (i.e., above CP) and to mathematically predict a workload that will elicit a specific Tlim and vice versa.  33  Figure 1.1. An illustrative example depicting the power-duration relationship as determined from three high-intensity constant load trials performed at varying workloads. Using the (A) hyperbolic power-duration relationship, critical power (CP) is calculated as the asymptote and W\u2019 is the curvature constant. In the (B) linear transformation of the power-duration relationship by using 1\/time, critical power is calculated as the y-intercept and W\u2019 is the slope of the line. When using the workloads and corresponding durations to calculate critical power and W\u2019 utilizing the linear transformation, critical power and W\u2019 were estimated as 161watts and 12.9KJ, respectively.  1.6.1 Determining the Power-Duration Relationship in Health  Critical power is considered to represent the highest workload or metabolic rate at which steady-state can be achieved largely due to contributions from oxidative metabolism such that substrate phosphorylation, [Bla] and VO2 reach steady-state despite being significantly elevated above resting values. In a seminal study performed by Poole et al., the physiological exercise responses at and above CP were investigated in healthy, physically active, younger males.216 Critical power occurred at ~68%Wmax (corresponding to ~80%VO2max) and was situated approximately mid-way between the anaerobic threshold and VO2max.  Following several minutes of exercise at CP (~18minutes), VO2, pH, and [Bla] (5.6\u00b10.9mmol\/L) eventually reached steady-state and exercise was maintained for 24-minutes (the investigators predetermined end-point). However, when workload was increased 5% above CP, VO2 progressed to VO2max, pH continued to decline, [Bla] increased inexorably (11.3\u00b11.4mmol\/L), and exercise was terminated by the participants at ~18-minutes.216 Due to 34 the different physiological responses observed above and below CP, it was concluded that CP demarcates the heavy exercise intensity domain (whereby VO2 and [Bla] do not continue to rise) from the severe exercise intensity domain (where a continued increase to maximal values is observed).216 Although similar findings in the physiological responses to exercise below, at or above CP have been reported in other studies and are generally accepted, a large variance exists within the literature likely due to differences in methodology and the population studied.241\u2013245  The responses of various measures of intramuscular metabolism have also been shown to differ above and below CP. Using 31P magnetic resonance spectroscopy during knee-extensor exercise, Jones et al., reported that a steady-state was achieved in PCr, Pi and pH following 1-3 minutes during exercise at 10% below CP.221 As such, exercise was maintained for 20 minutes. However, when exercise was performed at 10% above CP (corresponding to a relatively small mean difference of 2watts), PCr and pH continued to decrease while Pi increased significantly beyond resting values until exercise was terminated by participants at ~15-minutes. Using a similar methodology, Vanhatalo et al., studied the intramuscular responses in normoxia versus hyperoxia.246 During exercise above CP, the end-values of PCr, Pi and pH reached at exercise cessation were the same under both conditions, however a longer Tlim was achieved during the hyperoxic trials.246 This finding suggests that exercise tolerance above CP in healthy individuals is related to the attainment of a specific metabolic milieu that directly affects muscle contractile function and\/or limits muscle activation. Taken together, these investigations have established CP as a boundary below which, a physiological steady-state is attainable but above, intramuscular metabolism cannot be stabilized.   W\u2019 is a finite amount of work that can be performed above CP that is independent of the rate at which it is utilized. Classically, W\u2019 was considered to represent the sum of anaerobic energy stores, with a small aerobic contribution from myoglobin and venous Hb 35 bound O2 stores.215,219 However, it has been demonstrated that W\u2019 is not only affected by intramuscular PCr and glycogen stores (as demonstrated by creatine supplementation and glycogen depletion, respectively), but also the accumulation of fatigue related metabolites (e.g., Pi, H+, extracellular K and ADP).221,224,226\u2013228 Additionally, W\u2019 has been correlated with the VO2 slow component that occurs during exercise in the severe intensity domain (i.e., above CP where steady state is not achieved).225,247 The VO2 slow component is considered an additional, slowly developing increase in VO2 thought to reflect a progressive loss of skeletal muscle contractile efficiency due to the negative effects of fatigue-inducing metabolites on excitation-contraction coupling and\/or the recruitment of additional type II muscle fibers to maintain power output.248\u2013250 Therefore, the rate at which W\u2019 is utilized (and thus the Tlim at a given workload above CP) is considered to be dependent upon the rate of depletion of a finite amount of anaerobic energy stores, the accumulation of fatigue-related metabolites and a reduction in skeletal muscle efficiency.   1.6.2 Methodological Considerations for the Calculation of CP and W\u2019 Although previous studies have provided a physiological basis for both CP and W\u2019, it is imperative to acknowledge that differences in the methodology used to construct the power-duration relationship (i.e., the number and intensity\/duration of CLTs, and rest between trials) as well as the mathematical model employed (i.e., work vs. time, linear vs. nonlinear, two vs. three parameters, CLTs vs. the 3-minute all out test) will produce different estimates of CP and W\u2019.229\u2013232,234,235,251 Of particular note, prior studies have demonstrated that when higher intensities resulting in shorter Tlims are used to construct the power-duration relationship, the calculated CP is higher while W\u2019 is reduced. This has been demonstrated by various investigators utilizing different mathematical models, varying ranges of intensities and across sporting disciplines (i.e., running versus cycling).252\u2013257  For example, Bishop et al., compared parameter estimates during cycling exercise derived from trials lasting between 1-10 minutes, 36 ~1-3 minutes (the three shortest trials) and ~3-10 minutes (the three longest trials), and reported that CP was highest when the shortest predictive trials were used.252 In a subsequent study by Jenkins et al., a similar finding was observed even when predictive trials, on average, were longer and ranged from ~4-10 minutes (CP short), ~4-26 minutes (CP medium), and ~10-26 minutes (CP long).253 All CPs significantly differed from each other, with CP short being the highest and CP long being the lowest. When a verification trial was performed at the corresponding CPs, CP long could be sustained for ~45 minutes compared to 40 minutes and 34 minutes in the CP medium and short trials, respectively.  One proposed reason for why CP is higher when higher intensities resulting in shorter Tlims are used, is due to a greater \u2018metabolic inertia\u2019 that must be overcome at intensities above the anaerobic threshold. Therefore, when maximal or supramaximal intensities are used it is likely that aerobic metabolism is still rising when exercise tolerance is reached, thus biasing W\u2019 and CP in the negative and positive directions, respectively. In a review by Vandewalle and colleagues,258 the authors support this contention through mathematical simulations and propose that CP should be determined from heavy submaximal exercise trials as opposed to maximal and supramaximal trials to account for both the aerobic and anaerobic metabolic contributions to sustained exercise. This seems logical, since if the purpose or goal of utilizing the power-duration relations is to predict a sustainable workload that can be maintained for ~25-30 minutes, the physiological responses governed by both aerobic and anaerobic metabolism should be accounted for in the trials and thus mathematical calculation used to determine CP and W\u2019.  These differences in methodology are important as they will have a large effect on where CP lies relative to other physiological boundaries (i.e., AT, RCP and maximal lactate steady state (MLSS)), and in part, likely explains why some studies have reported a CP equal to AT while others have positioned it closer to RCP and Wmax.242,244,259\u2013262 The relative position of CP will also affect the sustainability of exercise at CP, and also likely explains the 37 wide range in durations (~15-60 minutes) reported between studies when a verification trial was performed at the calculated CP.216,238,242,254,263 In addition, CP and W\u2019 have also been demonstrated to vary according to age and training status.220,243,264 Therefore, when interpreting and comparing CP and W\u2019 reported within the literature, it is important to consider differences in the context of the methodology used and population studied.   1.6.3 Determining the Power-Duration Relationship in COPD The power-duration relationship has been previously examined in non-hypoxemic patients with COPD, and similar to healthy individuals is hyperbolic in nature.265\u2013269 Two previous studies using high-intensity CLTs ranging between ~70-120%Wmax and Tlims between ~2-10 minutes have compared CP and W\u2019 in patients with moderate to severe COPD and healthy controls.268,269 In both studies, absolute CP and W\u2019 were significantly reduced in patients; however, when expressed relative to Wmax, CP was similar (70\u00b114%Wmax COPD vs. 75\u00b14%Wmax controls268) or higher (82\u00b13%Wmax COPD vs. 68\u00b14%Wmax controls269) compared to controls. A comparably low CP and W\u2019 were also reported in three additional studies without a control group.265\u2013267 Although, on average, CP occurred at a relatively lower workload in these studies (50-65%Wmax), the authors noted a wide range in CP between individual patients (i.e., 50-90%Wmax).266,267  The significantly reduced absolute CP and W\u2019 reported in COPD has been attributed to differences in the physiological factors that govern Tlim in patients versus healthy controls. In non-hypoxemic moderate to severe COPD, ventilatory response dynamics appear to constrain tolerance to high-intensity exercise.269 This was determined by the consistent attainment of a ventilatory limitation (i.e., VEpeak reaching MVC), maximal dyspnea values and VO2peak at Tlim while HR remained below HRmax.269 With higher CLT workloads, a faster VO2 and VE kinetic response occurred due to a greater reliance on anaerobic glycolysis resulting in intolerable dyspnea and exercise cessation at progressively shorter Tlims.269 As such, W\u2019 38 was severely reduced in patients compared to controls and was interpreted to represent (at least partly) a consistent maximum limit of exertional dyspnea that a patient was willing to endure above CP.269 This was in contrast to the healthy controls, in whom a large and variable ventilatory reserve was reported at Tlim while the attainment of VO2peak and HRmax occurred during all CLTs.269 These finding demonstrate that the physiological underpinnings of CP and W\u2019 differ in patients with COPD versus healthy controls. Interestingly, no significant relationships were found between CP and the first lactate threshold or time constant for VO2 in patients, however these relationships were significant in controls.269 The authors interpreted this finding to suggest that cardiovascular or peripheral factors do not play a predominant role in limiting exercise endurance at high intensities in patient with moderate to severe COPD.    In three of the aforementioned studies, patients performed an exercise trial at CP. Despite the wide range in CP calculated between studies (53-82%Wmax), it was documented that all patient cohorts could sustain CP for ~15-20 minutes.266,267,269 In the study performed by Neder et al., VO2 and VE remained stable albeit at relatively high values (92\u00b17%VO2peak and 80.1\u00b16.5%VE\/MVC) and a large cardiac reserve was maintained during exercise at CP.269 The near-maximal VE occurred at the expense of an increased VT\/FVC ratio and lower Ti\/Ttot (measured at end-exercise) indicating dynamic hyperinflation likely occurred, however IC was not measured in this study. Interestingly, despite these changes in breathing patterns, dyspnea did not progress to intolerable levels, thus enabling patients to sustain exercise for ~20 minutes at 82\u00b13%Wmax. In two other studies that also performed an exercise trial at CP, similar end-exercise values for VO2, VE and HR were reported supporting that exercise at CP results in a relatively high but still submaximal VO2 and VE response associated with lower exertional symptoms.266,267 Therefore, it appears that in moderate to severe COPD CP demarcates workloads at which a ventilatory limitation is inevitable and exercise duration is finite, from workloads that can be sustained for a longer duration due to maintaining a ventilatory reserve and tolerable symptoms. This finding differs in comparison to healthy 39 individuals, in whom CP is governed by changes in aerobic-anaerobic energy contributions and the accumulation\/clearance of fatigue-related metabolites rather than a mechanical constraint to ventilation.    1.6.4 Using the Power-Duration Relationship to Accurately Predict Exercise Tolerance in COPD   To date, no study has utilized the power-duration relationship to accurately predict exercise tolerance in COPD for exercise prescription purposes. However, using the individual values obtained for CP and W\u2019, it is possible to calculate a predicted workload that can be maintained for a specific duration. As such, the power-duration relationship may prove to be a powerful tool to better individualize training intensities for patients. For example, by using the individual power-duration relationship it may be possible to predict the highest intensity that can be maintained for 30-minutes (i.e., CP30). If accurately predicted, CP30 may serve as an ideal initial intensity when starting an aerobic training program.   As previously stated, CP occurs across a wide range of intensities in patients with COPD.265\u2013269 As such, predicting an intensity that can be maintained for a specific period of time from the individually calculated CP and W\u2019 will also likely vary between patients. From previous studies, it appears that the development of a mechanical constraint to ventilation and ensuing dyspnea at a given high-intensity constant workload determines CP in certain patients with COPD.269 While this is most likely to apply to patients who have a classic ventilatory limitation to exercise, it is interesting to consider how the determinants of the power-duration relationship might differ in patients with a cardiovascular exercise limitation in whom VO2peak is primarily limited by Q and convective O2 delivery and a ventilatory reserve exists at peak exercise.  In patients who have a greater cardiovascular contribution to exercise limitation, CP may be more governed by changes in aerobic-anaerobic metabolic 40 contributions and\/or the accumulation of fatigue-related metabolites rather than the development of a constraint to ventilation. Additionally, if patients who are primarily cardiovascular limited are able to reach a higher Wmax and VO2peak during incremental CPET, CP and CP30 may also occur at a higher absolute workload compared to patients with a classic ventilatory limitation.  Therefore, the aim of Study #2 of this dissertation was to utilize the power-duration relationship to compare differences in CP and W\u2019 between different phenotypes of exercise limitation in COPD, and to assess the efficacy of this concept to accurately predict exercise tolerance as determined by CP30, in each of the phenotypes.   1.7 Additional Indices of Sustainability  In addition to CP, other indices have been used to differentiate sustainable from non-sustainable exercise intensities. An index of sustainability commonly used in sport performance is the maximal lactate steady-state (MLSS).270 The MLSS is conventionally derived from a series (typically 4-5) of 30-minute continuous exercise bouts completed on separate days at varying workloads.270,271 During each trial [Bla] is measured every 5 minutes, and MLSS is defined as the highest workload that does not result in a rise in [Bla] >1mmol\/L between 10 to 30 minutes.271 As such, MLSS is considered to represent the highest workload and thus metabolic load that can be sustained for ~30-minutes. Therefore, MLSS can be a useful intensity for effective exercise prescription and to assess improvements in performance. Alternative indices of sustainability have also been investigated using a variety of physiological variables such as VO2, HR, electromyography and rating of perceived exertion (RPE).272\u2013276 Similar to the methodology used to determine CP or MLSS, these indices are mathematically derived from the exercise response relationship (i.e., RPE versus time) over a range of submaximal workloads. However, in order to determine the workload associated with the attainment of a steady-state, the exercise response relationships are analyzed with 41 linear regression to determine the slope of each response. The slope coefficient from each submaximal trial (on the y-axis) and the corresponding workloads (on the x-axis) are then plotted. The x-intercept of this relationship (when y (slope) = 0) is thought to represent the workload at which steady-state can be maintained (i.e., the workload at which RPE does not continue to rise over time). An illustrative example of this analysis is presented in Figure 1.2.     Figure 1.2. An illustrative example of the slope coefficient analysis to determine the workload at which RPE will remain at a steady-state during constant load exercise. The RPE responses to exercise performed at three different high-intensity workloads are shown in (A) and the slope of the RPE-time response corresponding to each workload is depicted. When each slope is then plotted relative to the corresponding workload (B), linear regression can be used to determine the steady-state workload (on the x-axis) when y=0, indicating no change in the slope of the RPE response. In this example, the workload resulting in a steady-state in RPE was predicted to be 166watts.       42 1.7.1 Maximal Dyspnea Steady-State  As previously discussed, exercise training programs that achieve a greater training volume frequently report superior physiological training effects.47,55\u201357,59,60 However, certain patients with COPD may not be able to achieve the necessary intensity and\/or duration required to induce physiological adaptations as exercise may be stopped prematurely due to a rise in dyspnea to intolerable levels. Previous investigations in patients with COPD have demonstrated that tolerance to constant load exercise performed at moderate to high intensities (~60-85%Wmax) is reduced due to dynamic hyperinflation and VT constraint.110,138,139 When IRV is reduced to ~500ml, VT plateaus and an inflection in the IRV-dyspnea relation occurs due to a widening disparity between the increased neural drive to breathe and the mechanical response of the respiratory system.138,139,213 As such, dyspnea rises steeply and exercise is curtailed, on average, often before 10-minutes.111,277,278 In the exercise training setting, this would result in a reduction in total training volume and physiological training stimulus achieved. Alternatively, prescribing exercise at a relatively low intensity may result in tolerable symptoms; however, the physiological training stimulus may be too low to induce physiological adaptations. Therefore, the ability for patients with COPD to maintain exercise at as high an intensity as possible while maintaining tolerable symptoms may be key to improving the likelihood of physiological adaptations occurring.    Previous studies have investigated the use of exertional dyspnea ratings obtained from an incremental CPET to prescribe exercise intensity during constant load exercise. In a study by Horowitz et al., it was found that constant load exercise performed at the same Borg intensity that was reported at 80%VO2peak during an incremental CPET resulted in a similar VO2 during constant load and incremental exercise.279 As such, the authors concluded that individual Borg dyspnea ratings obtained from an incremental CPET could be used to reliably gauge constant load exercise intensity. In a subsequent study using a similar methodology, VO2 measured during treadmill exercise performed at the Borg intensity corresponding to 43 80%VO2peak during a cycling incremental CPET was similar to during treadmill exercise and incremental CPET.280 This finding suggested that exercise intensity can be accurately prescribed by using individual Borg dyspnea ratings even in a cross-modal prescription.280 However, in both of these studies, dyspnea and VO2 were only assessed over a 10-minute period, and as such, it is possible that dyspnea may have continued to rise overtime had exercise continued for a longer duration. Additionally, the ability to sustain exercise for a given duration is largely dependent upon the intensity prescribed. While it may be easier to predict a lower versus higher intensity that is associated with tolerable symptoms from an incremental CPET, the lower intensity may not represent an optimal training stimulus for physiological adaptations.  Considering that dyspnea is the most common symptom that curtails constant load exercise in COPD, identifying the highest workload at which dyspnea remains at a steady-state (MDSS, maximal dyspnea steady-state) may be highly beneficial, particularly if the goal of an exercise prescription is to optimize exercise duration at the highest intensity possible. Theoretically, MDSS will differentiate workloads at which dyspnea will remain stable allowing exercise to continue for a longer duration from those at which dyspnea will continue to rise leading to earlier exercise cessation. By utilizing a similar concept to MLSS and methodology to the slope coefficient analysis it may be possible to calculate MDSS in patients with COPD. However, since the physiological factors governing exercise tolerance may differ depending on the presence of a ventilatory and\/or cardiovascular contribution to exercise limitation, it may also be important to consider how MDSS differs in patients with different phenotypes of exercise limitation. As such, Study #3 of this dissertation sought to accurately predict the novel concept of MDSS, which represents the maximal workload at which the intensity of dyspnea would remain stable during constant load exercise in patients with COPD who have different exercise limitation phenotypes.   44 1.8 Purpose, Aims and Hypothesis Despite the wide-range in ventilatory, metabolic, and cardiovascular responses to incremental CPET observed in COPD, it is accepted that VO2peak is determined by abnormal ventilatory mechanics and\/or skeletal function and that exercise is terminated well-before a significant reduction in cardiac reserve occurs. Therefore, cardiovascular contributions to exercise limitation are generally considered minimally important. However, certain patients have been documented to attain HRmax at peak exercise indicating that Q and convective O2 delivery are maximized, with or without the presence of a ventilatory reserve.58,184 As such, in a cohort of patients, cardiovascular limitations to exercise may play a larger role than previously considered. Additionally, previous evidence has demonstrated that tolerance to high-intensity submaximal constant load exercise is determined by the development of a mechanical constraint to ventilation and ensuing dyspnea.110,117,138,139,269  While this may be the case in patients who are predominately ventilatory limited, exercise tolerance may be more governed by changes in aerobic-anaerobic metabolic contributions and\/or the accumulation of fatigue-related metabolites rather than the development of a constraint to ventilation in patients with a cardiovascular contribution to exercise limitation. These underlying physiological differences likely have important implications when determining a submaximal exercise intensity that can be sustained for a given duration.  Therefore, the purpose of this dissertation was to identify distinct phenotypes of exercise limitation in COPD based on the integrative submaximal and maximal physiological responses to incremental exercise. Additionally, we sought to investigate the different physiology factors governing submaximal constant load exercise in patients with different exercise limitation phenotypes, in an effort to accurately predict exercise tolerance at two novel indices of sustainable exercise intensity (CP30 and MDSS) that may be implemented to optimize exercise prescription for rehabilitative purposes.   45 Specific Aims  Study 1: To determine the distinct ventilatory, cardiovascular, and metabolic responses to incremental exercise in individuals with COPD who present with either a ventilatory, cardiovascular or combined exercise limitation.   Study 2: To utilize the power-duration relationship to compare differences in CP and W\u2019 between different phenotypes of exercise limitation in COPD, and to assess the efficacy of this concept to accurately predict the maximum workload that can be sustained for 30-minutes of continuous exercise (CP30) in each phenotype.   Study 3: To predict the novel concept of MDSS in patients with COPD who have different exercise limitation phenotypes.   Specific Hypothesis  Study 1: The cardiovascular limited phenotype would have the least amount of static hyperinflation at rest and dynamic hyperinflation during exercise, and thus the greatest VT expansion. Consequently, VO2peak and Wmax would be higher in the cardiovascular phenotype compared to the ventilatory and combined phenotypes.   Study 2: The workload corresponding to CP would be higher in the cardiovascular phenotype compared to both ventilatory and combined phenotypes. It was also hypothesized that patients would be able to sustain 30-minutes of continuous exercise at CP30, independent of the phenotype of exercise limitation.  Study 3: By using the slope of the dyspnea-time relationship, it would be possible to predict MDSS, defined as the maximum workload at which dyspnea would not increase >1 Borg unit 46 following 10 minutes of constant load exercise up to a total of 30 minutes, in all patients regardless of the phenotype of exercise limitation. 47 Chapter 2: Phenotyping Cardiopulmonary Exercise Limitations in Patients with Chronic Obstructive Pulmonary Disease  2.1 Background  Chronic obstructive pulmonary disease is a complex heterogenous condition with diverse clinical presentations and prognoses that cannot be entirely explained by differences in airflow limitation, dyspnea, and exacerbation history.281,282 As such, delineating the clinical phenotypes of COPD is important to facilitate the prescription of targeted therapies to optimize clinical outcomes. Incremental CPET is a clinically important tool in the risk stratification of patients due to the integrative assessment of physiological responses that can help distinguish subgroups of patients with unique disease characteristics.209,211,212 To date, identification of distinct phenotypes of exercise limitation in COPD based on the integrative physiological responses to maximal exercise have not been explored but may provide novel prognostic and therapeutic utility independent of disease severity.   In COPD, exercise limitation has classically been attributed to EFL causing an abnormal rise in lung volumes leading to a mechanical constraint to ventilation. In patients with greater static and\/or dynamic hyperinflation, EILV rises close to TLC during exercise.110,138 Tidal volume expansion becomes constrained and a further increase in VE can only be achieved by increasing breathing frequency.110,138 The greater mechanical work associated with breathing at higher lung volumes with a greater frequency increases inspiratory neural drive while the ability to increase VE is reduced.110,120,138,139 The resulting imbalance ultimately leads to the sensation of dyspnea and exercise termination. It is intuitive that patients with COPD would be primarily limited by the pulmonary system; however, considerable evidence supports that all systems in the O2 cascade integratively contribute to the body\u2019s inability to meet metabolic demand even in patients with advanced lung disease.114,178,179 Furthermore, during incremental exercise, a number of individuals with COPD reach HRmax with or without the presence of a ventilatory reserve.58,184 48 In health, SV plateaus at ~50% of VO2peak, and as such Q becomes limited once HRmax is reached.63\u201367 The observation that a subgroup of patients with COPD can reach HRmax suggests that the attainment of peak Q and thus convective O2 delivery also contribute to exercise limitation (as CaO2 does not significantly increase283) in certain patients. However, whether the physiological responses to exercise differ in patients with different cardiopulmonary exercise limitations has not been studied but may identify important phenotypes of patients, independent of FEV1.   The aim of this study was to determine the distinct pulmonary, cardiovascular, and metabolic responses to incremental exercise in individuals with COPD who present with either a ventilatory, cardiovascular or combined exercise limitation. We hypothesized that the cardiovascular limited phenotype would have the least amount of static hyperinflation at rest and dynamic hyperinflation during exercise, and thus the greatest VT expansion. Consequently, cardiopulmonary fitness (VO2peak) and Wmax would be higher in the cardiovascular phenotype compared to the ventilatory and combined phenotypes.  2.2 Methodology  2.2.1 Participants and Study Design  Stable individuals with physician confirmed COPD (post bronchodilator FEV1\/FVC<0.7 and <LLN24) free from concomitant conditions that could impair exercise capacity were included. Full details of recruitment and exclusion criteria are detailed in Figure 1. Testing was performed at the Universities of British Columbia (n=55) and Calgary (n=6). All participants signed an informed consent that had received local institutional research ethics board approval. Additionally, 54 CPETs previously conducted to screen for exercise contraindications in prior studies were retrospectively analyzed. Thirty-four CPETs met inclusion criteria and were included in the analysis. While the submaximal exercise responses 49 have never been published, some of the peak exercise responses (n=22) have been published elsewhere.284,285 Pulmonary function (6200 Autobox; SensorMedics, California, USA) was assessed according to the ATS\/ERS guidelines.24,286\u2013288 An incremental CPET was performed to symptom limitation on an electrically braked cycle ergometer (Ergoselect 200, SensorMedics GmbH, Bitz, Germany) with expired breath-by-breath gas analysis (Vmax 29C, SensorMedics, CA, USA [n=89] or QuarkCPET, COSMED, Italy [n=6]) according to ATS\/ACCP guidelines.62 Oxyhemoglobin saturation (Radial 7, Maximo, CA, USA) and HR (12-lead ECG; CardioSoftTM, GE Healthcare, WI, USA) were monitored continuously. Exertional dyspnea and leg fatigue (modified Borg Scale289) and IC (to assess changes in lung volumes137) were measured every 2-minutes. VO2peak and VEpeak were selected as the highest 30-second average, while HRpeak was the highest HR recorded. The VE-VCO2 slope and intercept were determined by plotting 30-second averages of VE versus VCO2 following minute-1 of exercise up to the RCP, which was considered the lowest VE\/VCO2 (VE\/VCO2 nadir) before a consistent rise and confirmed by the modified Beaver plot.290,291 If RCP could not be identified, all data up to symptom limitation were included and the lowest VE\/VCO2 was considered the nadir. Exercise limitation was determined according to current ATS\/ACCP recommendations, whereby MVC was estimated as FEV1* 35 and HRmax was calculated as 220-age.62 Phenotypes were classified as ventilatory (VEpeak\/MVC \u226585% or MVC-VEpeak \u226411L\/min, and HRpeak <90%HRmax), cardiovascular (VEpeak\/MVC <85% or MVC-VEpeak >11L\/min, and HRpeak \u226590%HRmax) or combined (met both ventilatory and cardiovascular criteria).   2.2.2 Statistical Analysis  Normality was assessed with the Shapiro-Wilk test. Parametric data was analyzed with a one-way ANOVA and Tukey HSD post-hoc, while Kruskal Wallis H-test and Dunn\u2019s test were used for non-parametric data at rest, 40watts (isoload-1) and peak exercise. Differences between 50 the cardiovascular and combined phenotypes at 60watts (isoload-2) were assessed with an independent t-test or Mann-Whitney rank sum. Isoloads represented the highest workloads achieved by >90% of participants. All data is presented as mean\u00b1SD. Utilizing data from our laboratory, it was anticipated that 55, 35 and 10% of individuals with COPD would be ventilatory, combined or cardiovascular limited, respectively. Assuming similar proportions, a minimum difference in the change in IC between groups of 200ml, a SD of 300ml, a \u03b2=0.8, and a two tailed \u03b1=0.017 (to correct for multiple comparisons), a minimum of 69 participants were required.   2.3 Results  2.3.1 Participant Characteristics and Pulmonary Function Ninety-five individuals were included in the study (Figure 2.1). Phenotypic characteristics are presented in Table 2.1. More patients were classified with a ventilatory phenotype (48%) compared to the combined (35%) or cardiovascular (17%) phenotypes.  Age, body mass index, and smoking history were similar between phenotypes, however the ventilatory phenotype included more males (59%) and reported a higher mMRC dyspnea compared to the other phenotypes. FEV1 and FEV1\/FVC were significantly different between all three phenotypes, with a large range of airflow obstruction within each: ventilatory (FEV1:23-75%pred), combined (28-90%pred), and cardiovascular (68-113%pred) (Figure 2.2). The cardiovascular phenotype had a lower RV\/TLC and larger VC compared to the other phenotypes, while RV and FRC were lower, and IC\/TLC and DLCO were higher compared to the ventilatory phenotype.   51   Figure 2.1. Study participant flow         52  Figure 2.2. The distribution of airflow limitation severity within each exercise limitation phenotype.         53 Table 2.0.1. Phenotypic Characteristics and Pulmonary Function  Variable Ventilatory (n=46) Combined (n=33) Cardiovascular (n=16) ANOVA P value Males:Females 27:19 17:16 8:8 - Age (years) 68 \u00b1 7 71 \u00b1 7 68 \u00b1 8 0.24 Height (m)  1.69 \u00b1 0.09  1.68 \u00b1 0.11  1.70 \u00b1 0.09 0.77 Body mass index (kg\/m2) 27.1 \u00b1 5.9 27.5 \u00b1 3.6 26.3 \u00b1 3.4 0.61 Smoking history (pk yr) 38 \u00b1 23 33 \u00b1 19 26 \u00b1 18 0.11 MRC dyspnea score 3 \u00b1 1*\u2020 2 \u00b1 1 2 \u00b1 1 <0.01 FEV1 (L) 1.34 \u00b1 0.45*\u2020 1.66 \u00b1 0.45\u2021 2.43 \u00b1 0.59 <0.01 FEV1 (% pred) 49 \u00b1 13*\u2020 64 \u00b1 15\u2021 88 \u00b1 14 <0.01 GOLD stage (%) (I\/II\/III-IV) 0\/50\/50 18\/61\/21 62\/38\/0 - FVC (L) 3.42 \u00b1 0.97 3.53 \u00b1 0.94 4.02 \u00b1 0.78 0.09 FVC (% pred) 93 \u00b1 16* 100 \u00b1 14\u2021 113 \u00b1 13 <0.01 FEV1\/FVC (%) 40 \u00b1 11*\u2020 49 \u00b1 11\u2021 60 \u00b1 7 <0.01 VC (L) 3.33 \u00b1 0.91* 3.33 \u00b1 0.78\u2021 4.05 \u00b1 0.81 0.01 VC (% pred) 90 \u00b1 17* 97 \u00b1 14\u2021 113 \u00b1 13 <0.01 TLC (L) 6.85 \u00b1 1.71 6.32 \u00b1 1.39 6.54 \u00b1 1.00 0.35 TLC (% pred) 108 \u00b1 17 105 \u00b1 13 107 \u00b1 14 0.64 IC\/TLC (%) 36 \u00b1 10* 40 \u00b1 11 45 \u00b1 9 <0.01 RV (L) 3.52 \u00b1 1.11* 2.98 \u00b1 0.92 2.48 \u00b1 0.50 <0.01 RV (% pred) 154 \u00b1 41* 133 \u00b1 32 112 \u00b1 26 <0.01 RV\/TLC (%) 51 \u00b1 8* 47 \u00b1 7\u2021 38 \u00b1 7 <0.01 FRC (L) 4.65 \u00b1 1.33* 4.03 \u00b1 1.21 3.66 \u00b1 0.75 0.01 FRC (% pred) 147 \u00b1 32* 132 \u00b1 29 117 \u00b1 24 <0.01 DLCO (ml\/mmHg\/min) 14.5 \u00b1 4.9 16.1 \u00b1 5.2 18.2 \u00b1 6.4 0.11 DLCO (% pred) 63 \u00b1 18* 72 \u00b1 19 79 \u00b1 21 <0.01 DLCO\/VA (ml\/mmHg\/min) 3.25 \u00b1 0.91 3.60 \u00b1 0.84 3.43 \u00b1 0.81 0.23 DLCO\/VA (% pred) 79 \u00b1 22 86 \u00b1 19 82 \u00b1 19 0.27 Medications (n [%]) SABA 28 (61) 18 (55) 7 (44)  Anticholinergic 28 (61) 15 (45) 4 (25)  LABA\/LAMA 12 (26) 1 (3) 0 (0)  ICS\/LABA 21 (46) 16 (48) 4 (25)  Inhaled corticosteroid 6 (13) 3 (9) 1 (6)  Statin 10 (22) 3 (9) 6 (4)  ARBs 7 (15) 4 (12) 4 (25)  ACE inhibitor 5 (11) 6 (18) 0 (0)  Diuretic 6 (13) 6 (18) 0 (0)  Abbreviations: MRC dyspnea score, measured with the medical research council breathlessness scale; FEV1, forced expiratory volume in 1 second; FVC, forced vital capacity; VC, vital capacity; TLC, total lung capacity; RV, residual volume; FRC, functional residual capacity; DLCO, diffusion capacity for carbon monoxide; DLCO\/VA, diffusion capacity for carbon monoxide corrected for alveolar ventilation. SABA, short-acting \u03b22-adrenergic receptor agonist; LABA\/LAMA, long-acting \u03b22-adrenergic receptor agonist and long-acting muscarinic antagonist; ICS\/LABA, inhaled corticosteroid and long-acting \u03b22-adrenergic receptor agonist; ARBs, angiotensin II receptor blocker; ACE inhibitor, angiotensin-converting enzyme inhibitor. Between phenotype comparisons: *p=0.05, ventilatory vs. cardiovascular. \u2020p=0.05, ventilatory vs. combined. \u2021p=0.05, combined vs. cardiovascular.   54 2.3.2 Peak Cardiopulmonary Exercise Responses  The cardiovascular phenotype reached a higher HRpeak and VEpeak compared to the ventilatory phenotype, however both HRpeak and VEpeak were similar to the combined phenotype (Table 2.2). Patients with a cardiovascular phenotype had a larger VT (Figure 2.3), and lower EELV and EILV compared to both other phenotypes (Figures 2.4). End-expiratory lung volume was lower at peak exercise in the combined versus ventilatory phenotype (Figure 2.4). Peak IC was larger in the cardiovascular phenotype (Figure 2.5), however there were no differences in the change in IC (\u0394IC) from rest-to-peak between phenotypes (-0.33\u00b10.43L, -0.51\u00b10.26L and -0.54\u00b10.33L in the cardiovascular, combined and ventilatory phenotypes, respectively, p=0.09).  At peak exercise, IRV (Figure 2.4 & 2.7), VT\/IC (Figure 2.5 & 2.7), and O2pulse (Figure 2.6) were similar between phenotypes.  The cardiovascular and combined phenotypes achieved a similar VO2peak, Wmax, VCO2, RER, SpO2 and VD\/VT at peak exercise (Table 2.2).  Compared to both other phenotypes, the ventilatory phenotype reached a lower VO2peak, Wmax, VCO2, RER, and SpO2, and higher VD\/VT at peak exercise.  Peak PETO2 was higher in the cardiovascular versus ventilatory phenotype, however PETCO2, dyspnea, and leg fatigue were not different between any phenotypes (Table 2.2).             55 Table 0.2.2. Incremental CPET Responses Between Phenotypes   Variable Ventilatory (n=46) Combined (n=33) Cardiovascular (n=16) ANOVA  P value VEpeak (L\/min) 45.3 \u00b1 15.5*\u2020 54.0 \u00b1 15.4 60.8 \u00b1 11.5 <0.01 VEpeak (% pred MVC) 101 \u00b1 15* 98 \u00b1 12 \u2021 73 \u00b1 10 <0.01 Ventilatory Reserve (L\/min) -0.2 \u00b1 5.8* +1.3 \u00b1 5.9 \u2021 +24.4 \u00b1 15.8 <0.01 HRpeak (beats\/min) 120 \u00b1 12*\u2020 146 \u00b1 11 147 \u00b1 10 <0.01 HRpeak (% pred) 79 \u00b1 7*\u2020 98 \u00b1 7 97 \u00b1 6 <0.01 Cardiac reserve (beats\/min) 32 \u00b1 11*\u2020  3 \u00b1 10 5 \u00b1 9  <0.01 Maximum workload (watts) 72 \u00b1 27*\u2020 91 \u00b1 30 103 \u00b1 34 <0.01 VO2peak (ml\/kg\/min) 15.2 \u00b1 3.3*\u2020 18.3 \u00b1 4.3 20.6 \u00b1 4.0 <0.01 VO2peak (% pred)a 63 \u00b1 19*\u2020 86 \u00b1 26 87 \u00b1 17 <0.01 VO2peak (L\/min) 1.19 \u00b1 0.40*\u2020 1.43 \u00b1 0.40 1.57 \u00b1 0.40 <0.01 VCO2 (L\/min) 1.22 \u00b1 0.46*\u2020 1.53 \u00b1 0.48 1.77 \u00b1 0.45 <0.01 RER 1.02 \u00b1 0.10*\u2020 1.07 \u00b1 0.09 1.13 \u00b1 0.10 <0.01 VE\/VCO2 nadir 38 \u00b1 7*\u2020 34 \u00b1 5 33 \u00b1 5 <0.01 VE-VCO2 slope 30 \u00b1 6 27 \u00b1 5 28 \u00b1 4 0.13 VE-VCO2 intercept  8 \u00b1 4 9 \u00b1 3 7 \u00b1 2 0.17 PETO2 (mmHg) 102.9 \u00b1 6.8* 106.2 \u00b1 6.5 109.9 \u00b1 6.2 <0.01 PETCO2 (mmHg) 35.3 \u00b1 4.3 35.3 \u00b1 4.4 34.4 \u00b1 5.2 0.79 VD\/VT  0.28 \u00b1 0.08*\u2020 0.22 \u00b1 0.06 0.18 \u00b1 0.04 <0.01 SpO2 (%) 92 \u00b1 4*\u2020 95 \u00b1 4 96 \u00b1 2 <0.01 \u0394 SpO2 (%) -3 \u00b1 3* -2 \u00b1 3 -1 \u00b1 2 0.01 Dyspnea (Borg units) 5.3 \u00b1 2.2 5.5 \u00b1 1.9 5.1 \u00b1 2.8 0.76 Leg Fatigue (Borg units) 5.4 \u00b1 2.5 5.9 \u00b1 2.5 5.9 \u00b1 2.7 0.69 Dyspnea\/LF\/Both (%) 39\/46\/15 42\/42\/15 31\/44\/25 0.89b       Abbreviations: VEpeak, peak minute ventilation; MVC, estimated maximum ventilatory capacity; HRpeak, peak heart rate; VO2peak, peak oxygen consumption; VCO2, volume of exhaled carbon dioxide; RER, respiratory exchange ratio; VE\/VCO2, ratio of minute ventilation to volume of exhaled carbon dioxide; PETO2,  pressure of end-tidal oxygen;  PETCO2, pressure of end-tidal carbon dioxide; VD\/VT, ratio of dead space ventilation to tidal volume obtained in ventilatory n=33, combined n=26, cardiovascular n=12; SpO2, oxyhemoglobin saturation; \u0394 SpO2, change in oxyhemoglobin saturation from rest to peak exercise; LF, leg fatigue; Both, both dyspnea and leg fatigue. a= calculated using the FRIEND database312.  b= p-value determined from Chi-Square test. Between phenotype comparisons: *p=0.05, ventilatory vs. cardiovascular. \u2020p=0.05, ventilatory vs. combined. \u2021p=0.05, combined vs. cardiovascular. 56  Figure 2.3. Phenotypic responses in (A) absolute ventilation, (B) relative ventilation (expressed as percentage of estimated maximal ventilatory capacity [MVC]), (C) tidal volume, and (D) breathing frequency during an incremental CPET. Between phenotype comparisons: *p=0.05, ventilatory vs. cardiovascular. \u2020p=0.05, ventilatory vs. combined. \u2021p=0.05, combined vs. cardiovascular.   57  Figure 2.4. Changes in global lung volumes in the (A) ventilatory, (B) combined and (C) cardiovascular phenotypes, and (D) end-expiratory lung volume, (E) end-inspiratory lung volume, and (F) inspiratory reserve volume during an incremental CPET. Between phenotype comparisons: *p=0.05, ventilatory vs. cardiovascular. \u2020p=0.05, ventilatory vs. combined. \u2021p=0.05, combined vs. cardiovascular 58  Figure 2.5. Phenotypic responses in (A) inspiratory capacity and (B) tidal volume to inspiratory capacity ratio during an incremental CPET.  Between phenotype comparisons: *p=0.05, ventilatory vs. cardiovascular. \u2020p=0.05, ventilatory vs. combined. \u2021p=0.05, combined vs. cardiovascular. 59  Figure 2.6. Changes in (A) absolute heart rate, (B) relative heart rate (expressed as a percentage of age-predicted heart rate maximum), and (C) O2pulse during an incremental CPET between phenotypes of exercise limitation in COPD. Between phenotype comparisons: *p=0.05, ventilatory versus cardiovascular. \u2020p=0.05, ventilatory versus combined. 60   Figure 2.7. The change in (A) exertional dyspnea, and the interrelationships between exertional dyspnea and (B) inspiratory reserve volume and (C) tidal volume to inspiratory capacity ratio during an incremental CPET between phenotypes of exercise limitation in COPD.  The dashed lines represent IRV (\u2264500ml) and VT\/IC % \u226570.61 2.3.3 Submaximal Cardiopulmonary Exercise Responses Absolute VE, VT and breathing frequency were similar between phenotypes at rest and both isoloads (Figure 2.3). The cardiovascular phenotype had the greatest ventilatory reserve (Figure 2.3B).  At rest and isoload-1, EELV and EILV were reduced and IRV was larger in the cardiovascular versus ventilatory phenotype (Figure 2.4). Compared to the combined phenotype, EELV was lower and IRV was greater at both isoloads, however EILV was lower only at isoload-2 in the cardiovascular phenotype (Figure 2.4). In the ventilatory phenotype, EILV was elevated at rest and EELV was higher at isoload-1 versus the combined phenotype (Figure 2.4). The cardiovascular phenotype had a larger inspiratory capacity at both isoloads (Figure 2.5) and a smaller \u0394IC from rest at isoload-1 (p=0.017) and isoload-2 (p=0.056) than both other phenotypes. VT\/IC was lower in the cardiovascular phenotype at isoload-1 and isoload-2 compared to the ventilatory and combined phenotypes, respectively (Figure 2.5). Heart rate was lower at isoload-1 in the ventilatory compared to combined phenotype, and O2pulse was similar at both isoloads between all phenotypes (Figure 2.6). VE\/VCO2 nadir was higher in the ventilatory phenotype than both other phenotypes, while VE-VCO2 slope and intercept were not different between phenotypes (Table 2.2). Absolute and relative VO2 and exertional symptoms did not differ between phenotypes at any isoload.   2.4 Discussion  This study is the first to provide empirical evidence that distinct phenotypes of exercise limitation can be identified in COPD by utilizing traditional exercise limitation criteria. In partial support of our hypothesis, patients with a cardiovascular phenotype had the least amount of static hyperinflation at rest and IC was larger throughout exercise compared to the other two phenotypes. While the extent of dynamic hyperinflation (\uf044IC) was not statistically reduced in the cardiovascular phenotype, the consistently lower EELV throughout exercise resulted in a 62 larger peak VT compared to the other two phenotypes, and a greater VEpeak compared to the ventilatory phenotype. Patients with a cardiovascular phenotype had a significantly higher VO2peak and Wmax compared to the ventilatory but were similar to the combined phenotype.  2.4.1 Lung Volume Responses to Incremental Exercise Compared to the ventilatory phenotype, patients with a cardiovascular phenotype had less static hyperinflation which allowed a greater reserve for VT expansion and thus a greater VT and VE were achieved at peak exercise. Additionally, EELV remained lower in the cardiovascular phenotype due to a slower rise in dynamic hyperinflation as IC was only reduced by ~330ml over ~100W compared to ~540ml over ~70W in the ventilatory phenotype. Despite 100% of the cardiovascular phenotype reaching an EILV \u226586%TLC, peak IRV (0.65\u00b10.40L) and  VT (1.96\u00b10.39L) were highly comparable to values previously reported at maximal exercise in healthy older adults of the same sex distribution and similar average age.97,292,293 Furthermore, when utilizing recently published CPET normative reference equations, peak VT was >LLN in 100% of patients with a cardiovascular phenotype,293 demonstrating normal VT expansion in these individuals. The cardiovascular phenotype also exhibited a normal hyperventilatory response to incremental exercise as evidenced by reaching an PETCO2 within a normal range after the RCP.294,295 This was accompanied by an RER >1.10, indicating that ventilation was adequate to remove additional CO2 produced due to a greater reliance on glycolytic metabolism. These observations demonstrate that not all patients with COPD have a clear ventilatory constraint to exercise and that the underlying physiological responses to maximal exercise in patients with a cardiovascular phenotype differ to what has classically been reported in COPD.110,139 As such, while mild alterations in pulmonary mechanics may partially contribute to exercise limitation in patients with a cardiovascular phenotype, they do not appear to be the primary limitation to exercise in these individuals.    63 In patients with a ventilatory phenotype, greater static and dynamic hyperinflation resulted in a higher EELV and EILV throughout exercise constraining VT expansion. Peak VT and VEpeak were reduced as EILV rose to ~94% of TLC and IRV reached ~420ml at a lower Wmax than the other two phenotypes. The greater elastic work of breathing  associated with breathing at higher lung volumes146 ultimately creates an imbalance between the increased inspiratory neural drive to breathe and inability to increase ventilatory output,139 leading to intolerable dyspnea and early exercise termination.110,139 At isoload-1, EELV, EILV and VT\/IC were lower and IRV was ~175% larger in the cardiovascular versus ventilatory phenotype. The ability to maintain a more efficient breathing pattern due to breathing at lower lung volumes in the cardiovascular phenotype may have contributed to the lower dyspnea at isoload-1 (p=0.03, ANOVA main effect), enabling those with a cardiovascular phenotype to reach a significantly higher Wmax despite eventually reaching a similar peak dyspnea to the ventilatory phenotype. Patients with a combined phenotype exhibited a similar lung volume response to the ventilatory phenotype; however, EELV remained lower at isoload and peak exercise compared to the ventilatory phenotypes. Therefore, the combined phenotype reached both a ventilatory and cardiovascular limitation relatively simultaneously but at a higher VO2peak, Wmax, and VEpeak compared to the ventilatory phenotype.      2.4.2 Cardiovascular Responses to Exercise  Age-predicted HRmax was obtained in 52% of individuals with COPD demonstrating a cardiovascular contribution to exercise limitation. To date, only one previous study has sought to categorized exercise limitations in COPD to better understand the variable adaptations gained following pulmonary rehabilitation.58 Utilizing slightly different criteria (attaining \u226580%pred HRmax), a similar percentage of patients (56%) were reported to achieve a cardiovascular limitation with or without a ventilatory limitation.58 It may appear that the different phenotypes of exercise limitation simply reflect differences in airflow limitation 64 severity. However, the exercise limitation phenotypes are not solely dependent upon airflow limitation severity due to the varied distribution of FEV1 observed across phenotypes (Figure 2.2). For example, patients with moderate airflow obstruction were included in all three phenotypes of exercise limitation. Additionally, ~40% of patient with a cardiovascular phenotype had moderate obstruction indicating that a subset of patients had a relatively normal ventilatory response at maximal exercise despite having moderate airflow limitation. In the combined phenotype, ~20% of patients presented with an FEV1<50%pred demonstrating that despite having severe airflow limitation and a significant reduction in ventilatory reserve at VO2peak, certain patients could still reach HRmax. As such, considering that these individuals were limited by both the pulmonary and cardiovascular systems at maximal exercise, it cannot be assumed that exercise limitation is primarily attributed to abnormal ventilatory mechanics before a significant reduction in cardiac reserve occurs in all patients with severe airflow limitation.  Acknowledging the limitations of O2pulse as a surrogate of SV296 and assuming a similar arterial-venous O2 response across phenotypes, the comparable O2pulse response suggests that SV is not significantly different between phenotypes. However, peak Q would be expected to be significantly greater in the cardiovascular and combined phenotypes due to reaching a higher HRpeak, which may partly explain the higher VO2peak achieved compared to the ventilatory phenotype. The ability to reach HRmax supports that the attainment of peak Q and limitation in convective O2 delivery to the skeletal muscle contributes to exercise limitation in many individuals with COPD, as previously proposed by others.177,178  2.4.3 Metabolic Responses to Exercise  VO2peak was higher in the cardiovascular and combined phenotypes compared to the ventilatory phenotype. Fifty-six percent of patients with a cardiovascular and 48% of those 65 with a combined phenotype had a normal cardiopulmonary fitness indicated by a VO2peak >84% of age-and sex-predicted.62 In each of these phenotypes, ~25% of patients reached \u2265100%pred VO2peak, demonstrating that cardiopulmonary fitness is well-preserved in certain individuals. In contrast, only 13% and 2% of patients with a ventilatory phenotype reached >84% or \u2265100% of predicted VO2peak, respectively.  VE\/VCO2 nadir was highest in the ventilatory phenotype due to exercise cessation at a lower workload (often before the RCP) as a result of VT constraint and likely a higher VD\/VT ratio. However, the VE-VCO2 slope and intercept were not different between phenotypes. Therefore, despite differences in disease severity, dynamic hyperinflation at isoloads, and mechanical constraint to ventilation at peak exercise, the ventilatory response to VCO2 and the CO2 set-point were similar and independent of the phenotype of exercise limitation. In COPD, it has been recently suggested that an EILV \u226590%TLC and a VE\/VCO2 nadir >34 more strongly predicts a reduction in VO2peak compared to ventilatory reserve.121 In the current cohort, 94% of all patients reached an EILV \u226588%TLC and the presence of a VE\/VCO2 nadir >34 was highly variable between phenotypes (67% in ventilatory, 39% in combined, and 50% in cardiovascular). Regardless, VO2peak was significantly higher in the cardiovascular and combined phenotypes compared to the ventilatory phenotype suggesting that the classification of patients according to phenotype of exercise limitation may be a more robust method to identify reductions in an integrative measure like VO2peak. Additionally, given that 40\/46 patients with a ventilatory phenotype reached a VO2peak <84%pred, the identification of a ventilatory phenotype may have important prognostic implications.297  2.4.4 Skeletal Muscle Contributions to Exercise Limitation  It must be acknowledged that all systems in the O2 cascade integratively contribute to the body\u2019s inability to meet metabolic demand even in patients with advanced lung disease.114,178,179 In COPD, alterations in skeletal muscle structure and function contribute to 66 exercise limitation.114,126 In many patients, skeletal muscle deconditioning and\/or dysfunction leads to a greater reliance on anaerobic glycolysis resulting in increased H+ and CO2 production above the anaerobic threshold.127,130 Increased CO2 production and an increased drive to breathe from type III & IV afferents131,298 could accelerate VT constraint leading to a ventilatory limitation at a lower workload, independent of airflow limitation severity. However, with maintained or improved skeletal muscle quality, ventilatory drive is likely reduced47,114 and HR can continue to rise closer to maximal values. Therefore, the ability for some patients with a cardiovascular or combined phenotype to achieve a VO2peak close to age-predicted normative values may be attributed to lower H+ and CO2 production, and afferent activity from metabo- and mechano-receptors due to preserved or enhanced skeletal muscle quality.   2.4.5 Clinical Relevance The physiological responses to incremental CPET differ between phenotypes. As such, the physiological stimulus associated with a generic exercise prescription (i.e., 60%Wmax) will result in different ventilatory, cardiovascular, and metabolic stressors depending on the phenotype of exercise limitation. This is likely to result in different physiological adaptations gained and outcomes achieved between patients following exercise training, and may explain previous findings in which COPD patients with a cardiovascular limitation achieved the greatest improvement in VO2peak following a pulmonary rehabilitation program compared to their ventilatory limited counterparts.58 Although we have identified three exercise limitation phenotypes, it is important to acknowledge that the range of exercise responses varied between individuals within the same phenotype. Therefore, exercise limitation in COPD is likely a continuum with the phenotypes representing specific physiological transitions. An initial regression from a cardiovascular to combined phenotype, followed by a slow eventual transition to a ventilatory phenotype may occur due to an increase in static and\/or dynamic hyperinflation, greater drive to breathe due to deconditioning and\/or changes in skeletal 67 muscle function. However, this transition may be offset or even somewhat reversed through optimal exercise conditioning, physical activity and medical therapy.   2.4.6 Study Considerations  Using estimates of MVC and HRmax have a number of limitations that have been previously documented.299,300 While other techniques (i.e., the maximal voluntary ventilation maneuver or VECAP method) may provide a more comprehensive evaluation of ventilatory capacity,299 they are not simple to perform or interpret clinically. The current estimates are routinely used in clinical practice to objectively identify adequate or abnormal cardiovascular and breathing reserves as per ATS\/ACCP recommendations62 as they provide an estimate of exercise limitation without additional complex physiological testing or analysis. Therefore, the classification of these phenotypes may be more likely to be implemented in clinical practice.   2.4.7 Conclusion  The categorization of patients with COPD according to traditional exercise limitation criteria resulted in the identification of different phenotypes of exercise limitations associated with distinct ventilatory, cardiovascular and metabolic responses to incremental CPET, that are not solely dependent upon FEV1. Maximum age-predicted HR was attained in approximately half of the study cohort indicating that maximum Q and convective O2 delivery contributed to exercise limitation in a significant number of patients. The relative contribution of a ventilatory and\/or cardiovascular limitation and thus phenotype classification is likely to be further mediated by skeletal muscle function.  Classifying patients phenotypically may have important prognostic implications and aid in the better individualization of exercise prescriptions in COPD.   68 Chapter 3: Exercise Limitation Phenotype Alters the Power-Duration Relationship in COPD 3.1 Background  The hyperbolic power-duration relationship can be used to predict a power output that can be maintained for a given duration and vice versa. Examining the physiological responses associated with this relationship can provide insight into the principal factor(s) that determine exercise tolerance at a given intensity. The power-duration relationship is governed by two parameters: 1) CP, the work rate (i.e. watts) or metabolic rate (i.e., % VO2peak) that separates sustainable intensities from those that are predictably limited, and 2) W\u2019, a finite amount of work that can be performed above CP.215\u2013218 In healthy individuals, CP is largely determined by changes in aerobic versus anaerobic metabolic contributions and the accumulation of fatigue-related metabolites. As such, during exercise above CP, steady state is not achieved indicated by a continued rise in [BLa] and VO2 while HR reaches maximal values and a large ventilatory reserve is maintained.216,220\u2013222,269  In contrast, patients with COPD commonly curtail high-intensity constant load exercise (CLT) due to a ventilatory limitation to exercise.112,269,301 In these patients, it has been postulated that CP represents the highest workload at which mechanical ventilatory constraint and dyspnea intensity remain below a limiting threshold.269 However, we have recently demonstrated that at least three distinct phenotypes of exercise limitation exist in COPD (Study #1, Chapter 2). In ~50% of patients, HRmax is reached during incremental exercise (with or without a ventilatory limitation) and thus an inability to further increase Q and convective O2 delivery contribute to exercise limitation.  As such, it is likely that CP is not solely governed by ventilatory constraint and ensuing dyspnea in all patients with COPD, and therefore CP and W\u2019 may differ depending upon the phenotype of exercise limitation. Traditionally, CP and W\u2019 are determined by performing 3-5 high-intensity CLTs (i.e., 80-120% Wmax) that elicit a time at the limit of tolerance (Tlim) between ~2-10 69 minutes216,234,302,303 and a CP that can be maintained for 15-20 minutes in COPD.266,267,269  However, both CP and W\u2019 are sensitive to the intensity (and thus duration) of CLTs, such that shorter trials produce a higher CP and lower W\u2019 and vice versa.252\u2013255 This methodological consideration has important implications as it will affect where CP is positioned relative to other physiological thresholds, and the duration for which CP can be sustained.  As the volume of exercise is important for maximizing the physiological benefits of exercise training in COPD,57,59 identifying the optimal intensity that can be sustained for longer than 15-20 minutes may be particularly beneficial. Using CLT intensities that result in Tlims between ~3-15 minutes, it is theoretically possible to predict the highest workload that each patient can sustain for 30-minutes (CP30).   The aim of this study was to utilize the power-duration relationship to compare differences in CP and W\u2019 between different phenotypes of exercise limitation in COPD, and to assess the efficacy of this concept to accurately predict exercise tolerance in each phenotype. It was hypothesized that the workload corresponding to CP would be higher in the cardiovascular phenotype compared to both ventilatory and combined phenotypes. It was also hypothesized that patients would be able to sustain 30-minutes of continuous exercise at CP30, independent of the phenotype of exercise limitation.   3.2 Methods 3.2.1 Participants and Study Design   This study included 30 stable individuals (exacerbation free >3months) with physician confirmed COPD (post bronchodilator FEV1\/FVC<0.7 and <LLN24) that simultaneously participated in Study #1 (Chapter 2). Participants were recruited from the pulmonary function laboratory at the Kelowna General Hospital and community pulmonary rehabilitation programs. Exclusion criteria included: taking a \u03b2-adrenoreceptor antagonist, known 70 cardiovascular disease, concomitant respiratory disease(s), other conditions known to limit exercise (i.e., musculoskeletal conditions, hypoxemia, diabetes, neurological disease), cardiovascular contraindications to exercise, or did not achieve the predetermined exercise limitation criteria. Participants signed an informed consent that had received institutional research ethics board approval. Five visits were performed >48-hours apart, and all testing was performed at the Kelowna General Hospital. During visit 1, a pulmonary function test and incremental CPET on a cycle ergometer was performed. On visits 2-4, CLTs were performed at varying workloads to determine the power-duration relationship and calculate CP and W\u2019. Visit 5 consisted of an exercise session at the predicted workload corresponding to CP30.   3.2.2 Pulmonary Function Testing and Incremental CPET A pulmonary function test consisting of routine spirometry and lung volumes determined by constant body-plethysmography (6200 Autobox; SensorMedics, Yorba Linda CA) was performed according to ATS\/ERS guidelines.286,288 An incremental CPET to symptom limitation was performed on an electrically braked cycle ergometer (Ergoselect 200, SensorMedics GmbH, Bitz, Germany) with expired breath-by-breath gas analysis (Vmax 29C, SensorMedics, California, USA) according to ATS\/ ACCP guidelines.62 Following 5-minutes of stable resting ventilatory values, participants cycled unloaded for 1-minute followed by an increase in 10 watts per minute until symptom limitation determined by the participant as reaching a dyspnea and\/or leg fatigue intensity at which they had to stop. Oxyhemoglobin saturation (pulse oximetry; Radical 7, Maximo, Irvine, CA) and HR (12-lead ECG; CardioSoftTM, GE Healthcare, Waukesha, WI) were monitored continuously. Blood pressure (measured manually with a sphygmomanometer and stethoscope), exertional dyspnea and leg fatigue (modified Borg Scale289) and IC (to assess changes in lung volumes137) were measured at rest, every 2-minutes, and at peak exercise. Finger capillary blood lactate samples (5\u00b5l) were collected at rest and peak exercise and analyzed using a Lactate Pro 71 analyzer (Arkray KDK Corporation, Kyoto, Japan). VO2peak and VEpeak were selected as the highest 30-second average, while HRpeak was the highest HR recorded. Anaerobic threshold was identified as the lowest VE\/VO2 before a consistent rise with a stable VE\/CO2.291 The RCP was considered the lowest VE\/VCO2 before a consistent rise and confirmed by the modified Beaver plot.290 Exercise limitations were determined according to ATS\/ACCP criteria whereby MVC was estimated as FEV1*35 and HRmax was calculated as 220-age.62 Phenotypes were classified as ventilatory (VEpeak\/MVC \u226585% or MVC-VEpeak \u226411L\/min, and HRpeak <90%pred), cardiovascular (VEpeak\/MVC <85% or MVC-VEpeak >11L\/min, and HRpeak \u226590%pred) or combined (met both ventilatory and cardiovascular criteria).  3.2.3. Determination of Critical Power and W\u2019 Constant load tests were performed on the same cycle ergometer as the incremental CPET and all physiological parameters were measured identically to the incremental CPET. Exercise consisted of 2-minutes of unloaded cycling, followed by cycling at a sustained workload to Tlim as determined by symptom limitation or the inability to maintain >50rpm despite verbal encouragement. As estimated CP and W\u2019 are known to be influenced by the duration of time trials, workloads were chosen to elicit an approximate Tlim between 14-18 minutes (CLT-1), 8-13 minutes (CLT-2), and 3-7 minutes (CLT-3) to extrapolate the workload corresponding to 30-minutes for CP30. Previous research has demonstrated that exercise tolerance is greatly reduced in individuals with COPD when rising lung volumes constrain VT expansion and IRV reaches ~500ml.110,138,139,304 Reaching this critical inspiratory constraint to ventilation has been shown to result in a dramatic rise in dyspnea depicted by an inflection in the IRV-dyspnea relationship that rapidly leads to exercise cessation.110,138,139,304 As such, CLT intensities were initially individualized based on the identification of an obvious inflection in the IRV-dyspnea relationship and workloads were selected as the power outputs corresponding to 10% below (CLT-1), at (CLT-2), and 10% above (CLT-3) the inflection point. 72 However, 43% of participants demonstrated no inflection point and workloads corresponding to 10% below, at, and 10% above the RCP were selected instead. If no inflection point or RCP could be identified (4\/30 participants), 60, 70 and 80%Wmax were used. These ranges of exercise intensities were selected to account for the slowed VE, VO2 and HR kinetics that are characteristics of COPD.305 Additionally, it was predicted that these intensities would result in a Tlim within the targeted duration to account for changes in aerobic-anaerobic metabolic contributions that occur with longer duration exercise, in an effort to more accurately predict CP30. Constant load trials were performed in a randomized order and participants remained blinded to the workload and duration of each trial. Verbal encouragement was standardized and administered by an assessment team member blinded to study outcomes. CP and W\u2019 were calculated by linearizing the hyperbolic power-time relationship by using the reciprocal of time (1\/time) to solve for power output using equation A: (A) Power Output = CP + (W\u2019 * [1\/time]) Where power output (watts) is a linear function of 1\/time (sec-1), and CP (watts; intercept) and W\u2019 (KJ; slope) were extracted by least-square linear regression.214   3.2.4 CP30 Exercise Session  The workload corresponding to CP30 was calculated using equation A with time set to 1800 seconds and using the individually determined CP and W\u2019. The exercise protocol was identical to the CLTs, however ventilatory parameters, [BLa] and exertional symptoms were measured at 10-minute intervals. To assess the accuracy of CP30, participants were not told that 30-minutes was the test\u2019s maximum duration and were instructed to cease exercise if their dyspnea or leg fatigue reached an intensity at which they would normally stop. 73 3.2.5 Statistical Analysis  Normality was assessed with the Shapiro-Wilk test. Parametric data was analyzed with a one-way ANOVA and Tukey HSD post-hoc, while Kruskal Wallis H-test and Dunn\u2019s test were used for non-parametric data. Exercise responses were analyzed at Tlim for the CLTs and at rest, 10-minutes, 20-minutes, and end-exercise for CP30. A z-proportions test was used to assess differences in the proportion of patients who successfully completed 30-minutes at CP30 between phenotypes. All data is presented as mean\u00b1SD. Assuming a minimally clinically important difference in workload between phenotypes at CP of 6.8 watts,306 a SD of 15 watts, a \u03b2=0.8, and a two-tailed \u03b1=0.05, 9 participants per phenotype were required.   3.3 Results  3.3.1 Phenotype Characteristics and Incremental CPET Responses The participant flow diagram is presented in Figure 3.1. Consecutive patients were recruited until ten patients with each phenotype were included, after which all other patients with that specific phenotype were excluded (n=6). Phenotype characteristics and peak incremental CPET responses are presented in Table 3.1. There were no differences in sex, age, height, or body mass index between phenotypes. FEV1 and FEV1\/FVC were significantly different between all phenotypes with a wide range of airflow obstruction within each. Residual volume was highest in the ventilatory phenotype. The cardiovascular phenotype had a great change in tidal volume (\u0394VT) and higher VO2peak compared to the ventilatory phenotype. Maximum workload and the \u0394IC were not different between any phenotypes.  74  Figure 3.1. Study participant flow        75 Table 3.1. Phenotype Characteristics and Incremental CPET Responses  Variable Ventilatory (n=10) Combined (n=11) Cardiovascular (n=9) P Male: Female 6:4 6:5 5:4 - Age (years) 64 \u00b1 8 72 \u00b1 5 69 \u00b1 9 0.09 Height (m) 1.70 \u00b1 0.10 1.69 \u00b1 0.11 1.68 \u00b1 0.10 0.93 BMI (kg\/m2) 27.2 \u00b1 6.2 26.7 \u00b1 2.7 26.0 \u00b1 3.8 0.83 FEV1 (L) 1.49 \u00b1 0.48* 1.78 \u00b1 0.41\u2021 2.38 \u00b1 0.62 0.02 FEV1 (% pred) 51 \u00b1 15*\u2020 69 \u00b1 13\u2021 89 \u00b1 15 <0.01 Range in FEV1 (% pred) 31-75 49-90 68-111 - FVC (L) 3.73 \u00b1 0.86 3.81 \u00b1 1.11 3.92 \u00b1 3.81 0.90 FVC (% pred) 98 \u00b1 13 101 \u00b1 13 113 \u00b1 12 0.05 FEV1\/FVC 41 \u00b1 12*\u2020 52 \u00b1 9\u2021 60 \u00b1 7 <0.01 RV (L) 3.54 \u00b1 0.96*\u2020 2.59 \u00b1 0.55 2.67 \u00b1 0.39 <0.01 TLC ( L) 7.31 \u00b1 1.60 6.08 \u00b1 1.18 6.61 \u00b1 0.97 0.11 RV\/TLC (%) 48 \u00b1 6 43 \u00b1 7 41 \u00b1 6 0.05 VEpeak (L\/min) 47.5 \u00b1 18.0 58.0 \u00b1 16.6 57.6 \u00b1 10.8 0.25 VEpeak (% pred MVC) 92 \u00b1 8* 95 \u00b1 10\u2021 70 \u00b1 8 <0.01 VE reserve (L\/min) 3.1 \u00b1 3.5* 2.7 \u00b1 5.4\u2021 26.1 \u00b1 13.4 <0.01 HRpeak (beats\/min) 125 \u00b1 11*\u2020 140 \u00b1 8 147 \u00b1 13 <0.01 HRpeak (% pred) 81 \u00b1 8*\u2020 95 \u00b1 6 97 \u00b1 7 <0.01 Cardiac reserve (beats\/min) 30 \u00b1 13*\u2020 8 \u00b1 9 4 \u00b1 11 <0.01 Maximum Workload (watts) 83 \u00b1 29 103 \u00b1 36 111 \u00b1 38 0.20 VO2peak (ml\/kg\/min) 16.1 \u00b1 4.5* 19.4 \u00b1 4.7 21.6 \u00b1 4.5 0.04 VO2peak (L\/min) 1.26 \u00b1 0.46 1.51 \u00b1 0.50 1.60 \u00b1 0.50 0.31 VO2peak (% pred)a 59 \u00b1 18*\u2020 86 \u00b1 18 88 \u00b1 19 <0.01 IRV (L) 0.54 \u00b1 0.28 0.50 \u00b1 0.29 0.59 \u00b1 0.25 0.76 Tidal volume (L) 1.51 \u00b1 0.46 1.69 \u00b1 0.54 2.00 \u00b1 0.74 0.15 \u0394 Tidal Volume (L) 0.59 \u00b1 0.45* 0.82 \u00b1 0.44 1.19 \u00b1 0.63 0.02 \u0394 IC (L) 0.63 \u00b1 0.29 0.44 \u00b1 0.30 0.43 \u00b1 0.41 0.33 RER 1.04 \u00b1 0.13 1.08 \u00b1 0.08 1.08 \u00b1 0.06 0.51 SpO2 (%) 92 \u00b1 4 95 \u00b1 3 96 \u00b1 3 0.12 Blood lactate (mmol\/L) 3.6 \u00b1 1.8 4.6 \u00b1 2.0 5.1 \u00b1 1.1 0.06 Dyspnea (Borg units) 5.3 \u00b1 1.8 5.4 \u00b1 2.0 4.4 \u00b1 2.1 0.54 Leg fatigue (Borg units) 5.2 \u00b1 2.6 5.6 \u00b1 2.9 5.6 \u00b1 1.9 0.92 Respiratory Medications (n [%]) SABA 7 (70) 7 (64) 5 (56)  Anticholinergic 8 (80) 5 (45) 1 (11)  LABA\/LAMA 4 (40) 1 (9) 0 (0)  ICS\/LABA 5 (50) 6 (55) 3 (33)  Inhaled corticosteroid 3 (30) 0 (0) 1 (11)  Abbreviations: BMI, body mass index; FEV1, forced expiratory volume in 1 second; FEV1\/FVC, ratio of forced expiratory volume in 1 second to forced vital capacity; RV, residual volume; TLC, total lung capacity; VEpeak, peak minute ventilation; MVC, estimated maximum ventilatory capacity determined as 35*FEV1; VE reserve, calculated as MVC-VEpeak; \u0394 IC, change in inspiratory capacity from rest to peak; HRpeak, peak heart rate; Cardiac reserve, calculated as (220-age)-HRpeak; VO2peak, peak oxygen consumption; RER, respiratory exchange ratio; \u0394 SpO2, change in oxyhemoglobin saturation from rest to peak; SABA, short-acting \u03b22-adrenergic receptor agonist; LABA\/LAMA, long-acting \u03b22-adrenergic receptor agonist and long-acting muscarinic antagonist; ICS\/LABA, inhaled corticosteroid and long-acting \u03b22-adrenergic receptor agonist. a= calculated using the FRIEND database [38]. Between phenotype comparisons: *p=0.05, ventilatory vs. cardiovascular. \u2020p=0.05, ventilatory vs. combined. \u2021p=0.05, combined vs. cardiovascular.   76 3.3.2 Physiological Responses during the CLTs The physiological responses to the CLTs are presented in Table 3.2 and Figure 3.2.  Absolute and relative workloads and VO2 at Tlim were higher in the cardiovascular versus ventilatory phenotype during all CLTs. In the combined phenotype, relative workload was higher during CLT-2 & 3, and lower during CLT-1 compared to the ventilatory and cardiovascular phenotypes, respectively. There were no differences in Tlim during any CLTs between phenotypes. Ventilatory reserve was larger in the cardiovascular phenotype versus both other phenotypes during all CLTs. The cardiovascular and combined phenotypes reached a similar HR and [BLa], and both variables were higher compared to the ventilatory phenotype during all CLTs. Patients with a cardiovascular phenotype reached a higher VO2, maintained a lower EELV, and had a greater \u0394VT compared to the ventilatory phenotype during all CLTs. In the combined phenotype, EELV was similar to the cardiovascular phenotype, but only lower during CLT-3 compared to the ventilatory phenotype. There were no significant differences in EILV, dyspnea or leg fatigue between phenotypes during any CLTs.                      77 Table 3.2. Constant Load Exercise Responses Between Phenotypes Variable  Ventilatory Combined Cardiovascular P Constant load trial #1 Workload (watts) 44 \u00b1 19* 64 \u00b1 25 77 \u00b1 28 0.02 Workload (%Wmax) 51 \u00b1 6* 61 \u00b1 6\u2021 69 \u00b1 3 <0.01 Tlim (sec) 1070 \u00b1 401 1196 \u00b1 525 825 \u00b1 635 0.30 VO2 (ml\/kg\/min) 13.1 \u00b1 3.2* 16.8 \u00b1 4.6 19.2 \u00b1 3.8 <0.01 Heart rate (beats\/min) 113 \u00b1 11*\u2020 130 \u00b1 14 140 \u00b1 14 <0.01 VE (L\/min) 37.7 \u00b1 11.8 47.7 \u00b1 14.3 49.3 \u00b1 6.7 0.08 VE reserve (L\/min) 12.8 \u00b1 6.1* 13.0 \u00b1 6.2\u2021 34.4 \u00b1 16.0 <0.01 IRV (L) 0.73 \u00b1 0.34 0.66 \u00b1 0.31 0.97 \u00b1 0.47 0.19 Tidal volume (L) 1.42 \u00b1 0.33 1.54 \u00b1 0.51 1.82 \u00b1 0.57 0.19 \u0394 Tidal volume (L) 0.45 \u00b1 0.27* 0.63 \u00b1 0.50 1.02 \u00b1 0.54 0.03 \u0394 IC (L) -0.52 \u00b1 0.30 -0.29 \u00b1 0.32 -0.16 \u00b1 0.45 0.11 SpO2 (%) 93 \u00b1 3* 94 \u00b1 2 96 \u00b1 3 0.04 Blood lactate (mmol\/L) 2.2 \u00b1 0.8*\u2020 3.9 \u00b1 2.0 4.5 \u00b1 0.8 <0.01 Dyspnea (Borg unit)  5.4 \u00b1 2.1 5.5 \u00b1 2.1 5.2 \u00b1 2.6 0.95 Leg fatigue (Borg unit) 5.3 \u00b1 2.0 5.7 \u00b1 3.1 6.0 \u00b1 2.7 0.87 Constant load trial #2 Workload (watts) 49 \u00b1 20* 72 \u00b1 28 84 \u00b1 30 0.02 Workload (%Wmax) 58 \u00b1 4*\u2020 69 \u00b1 6 76 \u00b1 3 <0.01 Tlim (sec) 747 \u00b1 454 848 \u00b1 526 557 \u00b1 380 0.33 VO2 (ml\/kg\/min) 14.1 \u00b1 3.1* 17.8 \u00b1 3.7 20.3 \u00b1 3.8 <0.01 Heart rate (beats\/min) 112 \u00b1 12*\u2020 132 \u00b1 13 144 \u00b1 11 <0.01 VE (L\/min) 39.0 \u00b1 12.3* 48.6 \u00b1 13.2 53.7 \u00b1 7.6 0.03 VE Reserve (L\/min) 11.6 \u00b1 4.6* 12.1 \u00b1 6.9\u2021 30.0 \u00b1 16.8 <0.01 IRV (L) 0.62 \u00b1 0.35 0.57 \u00b1 0.41 0.73 \u00b1 0.38 0.47 Tidal volume (L) 1.40 \u00b1 0.36 1.60 \u00b1 0.52 1.96 \u00b1 0.68 0.12 \u0394 Tidal volume (L) 0.44 \u00b1 0.36* 0.65 \u00b1 0.45 1.18 \u00b1 0.62 <0.01 \u0394 IC (L) -0.59 \u00b1 0.33* -0.4 \u00b1 0.25 -0.23 \u00b1 0.31 0.04 SpO2 (%) 94 \u00b1 2 93 \u00b1 2\u2021 96 \u00b1 3 0.02 Blood lactate (mmol\/L) 2.9 \u00b1 1.3*\u2020 4.8 \u00b1 1.9 4.8 \u00b1 1.2 0.02 Dyspnea (Borg unit) 5.1 \u00b1 1.7 5.2 \u00b1 2.2 4.8 \u00b1 2.3 0.68 Leg fatigue (Borg unit) 5.5 \u00b1 2.5 5.5 \u00b1 2.5 5.7 \u00b1 2.9 0.91        78 Table 3.2 Constant Load Exercise Responses Between Phenotypes (Cont\u2019d) Constant load trial #3 Workload (watts) 57 \u00b1 21* 80 \u00b1 31 92 \u00b1 33 0.03 Workload (%Wmax) 68 \u00b1 3*\u2020 77 \u00b1 7 82 \u00b1 3 <0.01 Tlim (sec) 576 \u00b1 461 386 \u00b1 165 332 \u00b1 131 0.76 VO2 (ml\/kg\/min) 14.7 \u00b1 3.7*   19.2 \u00b1 4.6 20.7 \u00b1 4.8 0.02 Heart rate (beats\/min) 118 \u00b1 5*\u2020 137 \u00b1 13 145 \u00b1 14 <0.01 VE (L\/min) 40.2 \u00b1 12.8* 54.9 \u00b1 16.0 53.3 \u00b1 10.0 0.04 VE reserve (L\/min) 10.3 \u00b1 4.3* 5.8 \u00b1 8.3\u2021 30.1 \u00b1 14.5 <0.01 IRV (L)  0.62 \u00b1 0.30 0.57 \u00b1 0.29 0.76 \u00b1 0.33 0.36 Tidal volume (L) 1.38 \u00b1 0.40 1.66 \u00b1 0.53 1.96 \u00b1 0.67 0.07 \u0394 Tidal volume (L) 0.49 \u00b1 0.35* 0.68 \u00b1 0.53 1.25 \u00b1 0.53 <0.01 \u0394 IC (L) -0.64 \u00b1 0.24 -0.29 \u00b1 0.31 -0.29 \u00b1 0.43 0.05 SpO2 (%) 91 \u00b1 4* 94 \u00b1 2 96 \u00b1 1 <0.01 Blood lactate (mmol\/L) 3.2 \u00b1 1.5*\u2020 5.2 \u00b1 1.6 5.6 \u00b1 1.3 <0.01 Dyspnea (Borg unit) 5.3 \u00b1 1.8 5.4 \u00b1 2.2 4.1 \u00b1 1.9 0.28 Leg fatigue (Borg unit) 5.5 \u00b1 2.4 5.5 \u00b1 2.5 5.8 \u00b1 2.6 0.99    Abbreviations: Tlim, time at the limit of tolerance; VO2, volume of oxygen consumed; VE, minute ventilation; \u0394 Tidal volume, the change in tidal volume from rest to Tlim; \u0394 IC, the change in inspiratory capacity from rest to Tlim; SpO2, oxyhemoglobin saturation; LF, leg fatigue; Both, participants stated both dyspnea and leg fatigue as the limiting symptom at end-exercise. Blood lactate concentration obtained in ventilatory n=10, combined n=9, cardiovascular n=7. Between phenotype comparisons: *p=0.05, ventilatory vs. cardiovascular. \u2020p=0.05, ventilatory vs. combined. \u2021p=0.05, combined vs. cardiovascular.  79  Figure 3.2. Phenotypic responses in (A, B & C) relative minute ventilation (expresses as a percentage of estimated maximal ventilatory capacity [MVC], the dashed line represents 85% of MVC), (D, E & F) relative heart rate (expressed as a percentage of age-predicted maximum heart rate, the dashed line represents 90% predicted), (G, H & I) end-inspiratory lung volume (EILV) and end-expiratory lung volume (EELV), and (J, K & L) VO2. Between phenotype comparisons: *p=0.05, ventilatory vs. cardiovascular. \u2020p=0.05, ventilatory vs. combined. \u2021p=0.05, combined vs. cardiovascular. 80 3.3.3 CP and W\u2019  The power-duration relationships between different phenotypes are presented in Figure 3.3. Absolute and relative CP were lower in the ventilatory phenotype compared to both other phenotypes (Table 3). No differences in W\u2019 were found between phenotypes (15.8\u00b120.6KJ ventilatory vs. 10.4\u00b18.1KJ combined vs. 9.8\u00b19.0KJ cardiovascular, p=0.98).  The exercise intensity domains for each phenotype are presented in Table 3.3 and Figure 3.4. In the ventilatory phenotype, both CP and CP30 were lower than AT (ANOVA p<0.01). In the combined phenotype, CP was lower however CP30 was comparable to AT (ANOVA p=0.02). There were no differences between CP, CP30 and AT in the cardiovascular phenotype (ANOVA p=0.41).        81  Figure 3.3. The individual power-duration relationships depicted as the (A) hyperbolic relationship, (B) linear relationship (absolute CP) and (C) linear relationship relative to percentage maximum workload (relative CP) between different phenotypes of exercise limitation in COPD.    82   Table 3.3. Exercise Intensity Domains in Different Phenotypes of COPD Training Zone Ventilatory Combined Cardiovascular P Critical power (watts) 33 \u00b1 13*\u2020 56 \u00b1 22 64 \u00b1 22 <0.01 Critical power (%Wmax) 40 \u00b1 9*\u2020 54 \u00b1 6 58 \u00b1 8 <0.01 CP30 (watts) 39 \u00b1 18* 60 \u00b1 25 68 \u00b1 24 0.02 CP30 (%Wmax) 44 \u00b1 8*\u2020 57 \u00b1 6 61 \u00b1 7 <0.01 AT (watts) 56 \u00b1 15 64 \u00b1 22 65 \u00b1 33 0.80 AT (%Wmax) 72 \u00b1 19 63 \u00b1 9 56 \u00b1 12 0.08 RCP (watts) - 99 \u00b1 27 93 \u00b1 36 0.75 RCP (%Wmax) - 85 \u00b1 4 84 \u00b1 10 0.64 Maximum workload (watts) 83 \u00b1 29 103 \u00b1 36 111 \u00b1 38 0.20   Abbreviations: CP30, the estimated highest power output that could be sustained continuously for 30-minutes; AT, anaerobic threshold determined from the CPET; RCP, respiratory compensation point determined from the CPET and clearly identified according to the specified methodology in ventilatory n=0, combined n=7, cardiovascular n=9. Wmax, maximum workload achieved during the CPET. Between phenotype comparisons: *p=0.05, ventilatory vs. cardiovascular. \u2020p=0.05, ventilatory vs. combined.  83   Figure 3.4. Different exercise intensity domains in the (A) ventilatory phenotype, (B) combined phenotype, and (C) cardiovascular phenotype. The first dashed line represents the average power output corresponding to the anaerobic threshold (AT). The second dashed line represents the average power output corresponding to the respiratory compensation point (RCP), which was clearly identified according to the specified methodology in ventilatory n=0, combined n=7, and cardiovascular n=9.  84 3.3.4 Responses to Exercise at CP30 Exercise responses to CP30 are presented in Tables 3.4 and Figure 3.5. The relative workload corresponding to CP30 was lower in the ventilatory phenotype versus both other phenotypes. There were no statistical differences in the proportion of patients who achieved 30-minutes at CP30 between the ventilatory and cardiovascular (p=0.07) or combined (p=0.51) phenotypes. In the cardiovascular phenotype at all timepoints, relative VE was lower compared to both other phenotypes, and EILV and EELV were lower than the ventilatory phenotype. At end-exercise only, EELV was lower in the combined versus ventilatory phenotype. Throughout exercise and at end-exercise, HR, VO2, and [BLa] were similar in the cardiovascular and combined phenotypes and higher in both the cardiovascular and combined phenotypes versus the ventilatory phenotype. At end-exercise, the cardiovascular phenotype had a larger IRV compared to both other phenotypes, and a larger \u0394VT compared to the ventilatory phenotype. There were no differences in dyspnea or leg fatigue between phenotypes.             85 Table 3.4 End Exercise Responses at CP30 Between Phenotypes   Variable Ventilatory Combined Cardiovascular P Achieved 30-minutes (n) 10\/10 9\/11 5\/9 - Average time (sec) 1800 \u00b1 0 1706 \u00b1 248 1597 \u00b1 248 0.06 Workload (watts) 39 \u00b1 18* 60 \u00b1 25 68 \u00b1 24 0.02 Workload (%Wmax) 44 \u00b1 8*\u2020 57 \u00b1 6 61 \u00b1 7 <0.01 VO2 (% peak) 76 \u00b1 7 82 \u00b1 8 80 \u00b1 8 0.21 Heart rate (beats\/min) 104 \u00b1 13* 118 \u00b1 12 131 \u00b1 18 <0.01 Cardiac reserve (beats\/min) 52 \u00b1 17*\u2020 30 \u00b1 15 20 \u00b1 16 <0.01 VE (L\/min) 33.4 \u00b1 10.3 43.5 \u00b1 11.3 42.7 \u00b1 6.6 0.05 Ventilatory reserve (L\/min) 17.1 \u00b1 6.6*  17.2 \u00b1 7.3\u2021 41.0 \u00b1 18.5 <0.01 EILV (L) 6.61 \u00b1 1.58 5.36 \u00b1 1.01 5.43 \u00b1 0.98 0.05 EELV (L) 5.26 \u00b1 1.43*\u2020 3.76 \u00b1 0.68 3.81 \u00b1 0.65 <0.01 IRV (L) 0.70 \u00b1 0.27* 0.71 \u00b1 0.35\u2021 1.14 \u00b1 0.46 0.03 Tidal volume (L) 1.35 \u00b1 0.22 1.60 \u00b1 0.50 1.66 \u00b1 0.45 0.28 \u0394 Tidal volume (L) 0.44 \u00b1 0.20* 0.73 \u00b1 0.40 0.90 \u00b1 0.34 0.02 IC (L) 2.05 \u00b1 0.34 2.32 \u00b1 0.70 2.80 \u00b1 0.69 0.06 \u0394 IC (L) -0.41 \u00b1 0.26 -0.22 \u00b1 0.18 -0.21 \u00b1 0.28 0.14 SpO2 (%) 93 \u00b1 2 94 \u00b1 1 95 \u00b1 2 0.17 Blood lactate (mmol\/L) 1.8 \u00b1 0.7*\u2020 3.5 \u00b1 1.2 3.3 \u00b1 1.2 <0.01 Dyspnea (Borg unit) 3.0 \u00b1 1.6 3.4 \u00b1 1.4 3.3 \u00b1 2.0 0.84 Leg fatigue (Borg unit) 3.9 \u00b1 1.2 4.2 \u00b1 2.1 4.8 \u00b1 2.4 0.80 Abbreviations: VO2, volume of oxygen consumed; VCO2, volume of carbon dioxide produced; Cardiac reserve, calculated as (220-age) \u2013 end-exercise heart rate; VE, minute ventilation; VE reserve, calculated as VEpeak from CPET \u2013 end-exercise VE;  EILV, end-inspiratory lung volume; EELV, end-expiratory lung volume; IRV, inspiratory reserve volume; \u0394 Tidal volume, change in tidal volume from rest to end-exercise. IC, inspiratory capacity; \u0394IC, change in inspiratory capacity from rest to end-exercise. Between phenotype comparisons: *p=0.05, ventilatory vs. cardiovascular. \u2020p=0.05, ventilatory vs. combined. \u2021p=0.05, combined vs. cardiovascular.   86  Figure 3.5. The change in (A) relative minute ventilation (expresses as a percentage of estimated maximal ventilatory capacity [MVC]), (B) relative heart rate (expressed as a percentage of age-predicted maximum heart rate), (C) VO2, (D) capillary blood lactate concentration, (E) end-inspiratory lung volume (EILV) and end-expiratory lung volume (EELV), and (F) dyspnea during the CP30 exercise session.*p=0.05, ventilatory vs. cardiovascular. \u2020p=0.05, ventilatory vs. combined. \u2021p=0.05, combined vs. cardiovascular. 87 3.3.5 Secondary Analysis: Lung Volume Responses Within Phenotype   In order to assess the within phenotype lung volumes responses to exercise during the CLTs and CP30, a secondary post-hoc analysis was performed using a 1-way repeated-measures ANOVA. The within phenotype changes in lung volumes are presented in Figure 3.6. In the ventilatory phenotype, the change in EELV from rest to end-exercise was greater during CLT-3 versus CP30. In the cardiovascular phenotype, relative EILV was higher at end-exercise during CLT-2 and CLT-3 and the change in EILV was larger during CLT-3 compared to CP30. The \u0394VT was greater during all CTLs versus CP30 in the cardiovascular phenotype.     i    Figure 3.6. Within phenotype responses during CLT-1, CLT-2, CLT-3 and CP30 in (A,B,C) relative end-inspiratory lung volume (EILV) and end-expiratory lung volume (EELV), (D,E,F) the absolute change in tidal volume from rest to end-exercise, (G,H,I) the absolute change in end-inspiratory lung volume from rest to end-exercise, and (J,K,L) the absolute change in end-expiratory lung volume from rest to end-exercise. Dashed lines represent the mean change in tidal volume from rest to peak exercise achieved during the incremental cardiopulmonary exercise test. *p=0.05, significantly different from CP30. \u2020p=0.05, significantly different from CLT-3.   89 3.4 Discussion  In partial support of our primary hypothesis, CP occurred at a higher absolute and relative workload in patients with a cardiovascular phenotype compared to ventilatory but was not different from the combined phenotype. CP30 also occurred at a higher relative workload in both the cardiovascular and combined phenotypes versus the ventilatory phenotype. Compared to the ventilatory phenotype, there was a greater variance in the duration achieved at CP30 in both other phenotypes.   3.4.1 Differences in CP and W\u2019 between Phenotypes  The CLT exercise responses observed in the cardiovascular phenotype were similar to those documented in healthy individuals in whom CP is considered to represent the maximum workload and metabolic rate associated with sustainable oxidative steady-state.216,220\u2013222 All patients with a cardiovascular phenotype maintained a ventilatory reserve at Tlim (i.e. VEpeak\/MVC <85% and MVC-VEpeak >11L\/min), and HRpeak was \u226590%HRmax in 90% of the CLTs. Additionally, in the cardiovascular phenotype EELV remained lower and \u0394VT was greater during all CLTs despite exercising at a higher workload, VO2, HR, and [BLa] compared to the ventilatory phenotype. As such, in the cardiovascular phenotype, CP appear to be determined by changes in aerobic-anaerobic energy contributions and the production\/clearance of fatigue related metabolites rather than a mechanical constraint to ventilation. In contrast, no patient with a ventilatory phenotype reached \u226590%HRmax and HR was consistently lower compared to the cardiovascular phenotype during all CLTs; however, a ventilatory limitation was reached in 80% of CLTs. The smaller \u0394VT in the ventilatory versus cardiovascular phenotype was due to a higher EELV, as EILV was equivalent during all CLTs. Therefore, CP in the ventilatory phenotype was likely governed by a mechanical constraint to VT expansion due to greater static and dynamic hyperinflation. In the combined phenotype, CP appeared determined by both the limits of aerobic metabolism and VT constraint as end-exercise HR and [BLa] were 90 similar to the cardiovascular phenotype and higher than the ventilatory phenotype; however, \u0394VT expansion was more comparable to the ventilatory phenotype. Interestingly, dyspnea and leg fatigue reached similar peak values and were not different between phenotypes during any CLTs suggesting that the mechanisms of exertional symptoms vary between phenotypes as the physiological responses at Tlim clearly differ. Critical power occurred at a higher absolute and relative workload in both the cardiovascular and combined phenotypes compared to the ventilatory phenotype. This demonstrates that the upper limit which separates sustainable intensities from those that are predictably limited is higher in patients who can reach \u226590%HRmax during an incremental CPET, independent of the presence of a ventilatory limitation. This novel finding emphasizes the importance of identifying phenotypes of exercise limitation when prescribing sustainable workloads for patients. In previous studies of heterogeneous COPD cohorts which used high-intensity CLTs ranging between ~70-120%Wmax and Tlims of ~2.5-12 minutes, CP occurred on average between ~53-82%Wmax.266\u2013269 Despite the range in CP between studies, it was reported that patients were able to complete ~15-20 minutes at CP.266,267,269 In general, our CP predictions were somewhat lower ranging from 40%Wmax in the ventilatory phenotype to 58%Wmax in the cardiovascular phenotype. These discrepancies may be due to methodological differences such as performing multiple CLTs on the same day.267,268 Additionally, CP has been shown to be estimated at a higher workload when higher intensity and shorter trials are used.252\u2013255 In the current study, lower intensities (~50-80%Wmax vs. 70-120%Wmax) and longer Tlims were used to predict the highest workload that could be sustained for 30-minutes, thus resulting in a lower CP.   W\u2019 did not differ between phenotypes. In healthy adults, CP is situated between AT and Wmax (its relative position dependent on age and fitness level220,243,264), and therefore, W\u2019 represents a finite amount of work that can be performed above CP. Conventionally, W\u2019 is considered a composite measure of factors that govern glycolytic metabolism including the 91 depletion of intramuscular substrates (i.e., PCr and glycogen), the accumulation of fatigue-related metabolites (i.e., H+, Pi, ADP, and K), and the VO2 slow component. 221,224\u2013228 It is likely that these variables still contributed to the sum of W\u2019 in all phenotypes of exercise limitation in COPD. However, considering that CP was positioned close to (i.e., cardiovascular and combined phenotypes) or below AT (i.e., ventilatory phenotype) other factors not specifically governed by glycolytic metabolism (e.g., work of breathing, blood flow redistribution, and dyspnea) likely also play a role in determining the amount of work that patients are able to endure above CP.   3.4.2 Exercise Responses to CP30  CP30 occurred at a similar relative workload in the cardiovascular and combined phenotypes, and both workloads were higher compared to the ventilatory phenotype. Throughout exercise, VO2, HR, and [BLa] were comparable between the cardiovascular and combined phenotypes, and consistently higher versus ventilatory phenotype indicating that patients with a cardiovascular or combined phenotype could sustain 30-minutes of continuous exercise at a greater metabolic load compared to the ventilatory phenotype. In the cardiovascular phenotype \u0394VT was greater than the ventilatory phenotype due to a lower EELV. As such, EILV also remained lower throughout exercise in the cardiovascular versus ventilatory phenotype. Additionally, in the cardiovascular phenotypes, \u0394VT during CP30 reached 0.90\u00b10.34L versus 1.19\u00b10.52L during the incremental CPET (p=0.01) demonstrating a small reserve to further expand VT (Figure 3.6). However, in both the combined and ventilatory phenotypes, \u0394VT during CP30 was comparable to that observed in the incremental CPET (both p>0.10). Thus, in the cardiovascular phenotype, CP30 represented the highest metabolic load that could be tolerated without excessive fatigue or dyspnea. In the combined phenotype, CP30 occurred at a comparable metabolic load to the cardiovascular phenotype but also at the limit of VT expansion. However, this occurred without resulting in intolerable 92 exertional symptoms in the majority of these patients. In the ventilatory phenotype, CP30 occurred at a much lower workload and metabolic load than the other two phenotypes, which resulted in a delayed rise in EELV. This preserved VT at or below a level where VT constraint would occur (Figure 3.6) enabling all patients to complete 30-minutes of continuous exercise. The sensitivity of the power-duration relationship and the importance of accurately prescribing exercise intensity to optimize exercise tolerance is clearly demonstrated by our data, as a mere 5-watt reduction between CLT-1 and CP30 in the ventilatory phenotype resulted in a difference in \u0394EELV of 110ml but an increase in exercise duration of ~70% (Figure 3.6).  Although no differences were found between phenotypes in the proportion of patients who successfully achieved 30-minutes at CP30, 4\/9 individuals with a cardiovascular phenotype did not reach 30-minutes as opposed to 2\/11 and 0\/10 in the combined and ventilatory phenotypes, respectively. Patients who did not reach 30-minutes in the cardiovascular phenotype had a higher end-exercise [BLa] (4.2\u00b11.0mmol\/L vs. 2.6\u00b10.9mmol\/L, p=0.04) and completed ~22.5\u00b11.5minutes at CP30. While underpowered to detect subgroup differences, individuals who did not reach 30-minutes achieved shorter CLT durations despite exercising at lower workloads on average, compared to the rest of the cardiovascular cohort which likely influenced the accuracy of the CP30 prediction.      3.4.3 Practical Application and Clinical Relevance The current study demonstrates that patients who reach \u226590%HRmax during an incremental CPET can exercise continuously at a higher intensity for the same duration compared to those with a ventilatory phenotype. This speaks to the importance of tailoring exercise prescriptions to each phenotype as the use of generic exercise prescriptions, even when individualized (i.e., 60%Wmax) will result in the over or undertraining of certain patients. On average, patients with a ventilatory phenotype can maintain continuous cycling for 30-minutes at an intensity ~20-30% below AT, where exercise is predominantly metabolically aerobic and the 93 accumulation of fatigue-related metabolites does not occur. However, individuals with a combined or cardiovascular phenotype can sustain 30-minutes of cycling at an intensity ~10% below or ~10% above AT, respectively (Figure 3.4).   3.4.4 Study Considerations  As no CP verification trial was performed we cannot say with certainty which physiological variables govern CP in each phenotype. Also, the use of different incremental CPET protocols (i.e., 10W\/min incremental versus ramp protocol) and power outputs\/durations to construct the power-duration relationship makes the comparison of results to previous studies challenging. It is possible that AT and\/or RCP may have been slightly over predicted by using a 10W\/min incremental CPET protocol, thus positioning CP30 at a somewhat higher intensity relative to AT compared to if a longer protocol (i.e., 2\u20133minute increments) was used.   3.4.5 Conclusion Critical power and CP30 occurred at a higher workload in the cardiovascular and combined phenotypes compared to the ventilatory phenotype. This demonstrates that patients with COPD who reach \u226590%HRmax during incremental exercise testing with or without a ventilatory limitation, can exercise continuously for 30-minutes at a higher intensity and metabolic load compared to those with a ventilatory phenotype. This novel finding demonstrates the importance of identifying phenotypes of exercise limitation to accurately prescribe the optimal intensity for sustainable exercise, to increase exercise volume and the physiological training stimulus achieved.  94 Chapter 4: Predicting Maximal Dyspnea Steady-State in Different Phenotypes of Exercise Limitation in COPD  4.1 Background  Previous investigations in patients with COPD have demonstrated that exercise tolerance to constant load exercise performed at moderate-high intensities (~60-85%Wmax) is reduced due to dynamic hyperinflation resulting in VT constraint and intolerable dyspnea.110,138,139,148 When IRV is reduced to ~500-600ml, VT plateaus and an inflection in the IRV-dyspnea relationship occurs.138,139,213 At this inflection point, there is a widening disparity between the increased neural drive to breathe and the mechanical response of the respiratory system that results in a steep rise in dyspnea to intolerable levels followed by exercise cessation.120,138,139 As such, moderate-high intensity constant load exercise is curtailed, on average,  at less than 10-minutes.111,148,277,278 In the exercise training setting, this would result in a reduction in total training volume, and thus physiological training stimulus achieved. Alternatively, prescribing exercise at a relatively low intensity may result in tolerable exertional symptoms; however, the physiological training stimulus may be too low to induce physiological adaptations. Therefore, the ability for patients with COPD to maintain exercise at as high an intensity as possible while maintaining tolerable symptoms may be key to improving the likelihood of physiological adaptations occurring.  The identification of a maximum steady-state intensity has been investigated in healthy individuals and athletic populations using several different physiological variables to determine the highest workload that demarcates sustainable versus non-sustainable intensities. The most common indices used to define maximum steady-state intensity include CP and MLSS.223,270 However, this concept has also been applied to other variables such as VO2, HR, electromyography, and RPE.272\u2013276 Similar to the methods used to determine CP or MLSS, these alternative indices are derived mathematically from the exercise response relationship (i.e., RPE vs. time) over a range of submaximal workloads. Considering that 95 dyspnea is the most common symptom that curtails constant load exercise in COPD, identifying the highest workload at which dyspnea remains at a steady-state during continuous exercise (maximal dyspnea steady-state (MDSS)) may be highly advantageous.  Theoretically, MDSS will differentiate the maximum workload at which dyspnea will remain stable allowing exercise to continue for a longer duration from workloads at which dyspnea will continue to rise leading to earlier exercise cessation.  It has recently been demonstrated by our group that different phenotypes of exercise limitation exist in COPD that are not solely dependent on the severity of airflow limitation, that influence the maximal sustainable intensity determined from the power-duration relationship (Study #2, Chapter 3). It was found that patients with a cardiovascular contribution to exercise limitation (i.e., attained \u226590% HRmax during an incremental CPET) could sustain constant load exercise at a much higher intensity for the same duration (30-minutes) compared to patients with a classic ventilatory phenotype. However, exertional dyspnea responses during and at end-exercise, were similar across phenotypes. As such, it is important to consider these phenotypes when assessing whether MDSS can be accurately predicted in patients with COPD. Therefore, the primary aim of this study was to predict the maximum workload at which the intensity of dyspnea will remain stable during constant load exercise in patients with COPD who have different exercise limitation phenotypes. It was hypothesized that by using the slope of the dyspnea-time relationship, it would be possible to predict MDSS, defined as the highest workload at which dyspnea would not increase >1 Borg unit following 10-minutes of constant load exercise up to a total of 30-minutes, in all patients regardless of the phenotype of exercise limitation.   96 4.2 Methodology  4.2.1 Participants and Study Design This study included 23 patients with physician confirmed COPD (post-bronchodilator FEV1\/FVC<0.7 and <LLN24) who simultaneously participated in study #2 (chapter 3, n=30). Seven patients did not participate in Study #3 as ethical approval for this study was obtained after these individuals had already completed Study #2. Study inclusion and exclusion criteria were the same as detailed in Study #2. In brief, patients were excluded if they had recently experienced an exacerbation (<3 months), were taking a \u03b2-adrenoreceptor antagonist, had a concomitant condition that could influence exercise limitation (i.e., other respiratory condition, neuromuscular disease, diabetes, or hypoxemia), presented with a cardiovascular contraindication to exercise or did not achieve the predetermined exercise limitation criteria. All participants signed an informed consent form that had received institutional research ethics board approval. The incremental CPET and CLTs performed as part of Study #2 were used in Study #3 for the analysis of MDSS. The MDSS verification trial was performed >48-hours following the CP30 exercise trial performed as part of Study #2.   4.2.2 Pulmonary Function, Incremental CPET and CLTs Detailed methods for pulmonary function testing, incremental CPET, and CLTs are described in Study #2 (chapter 3). Briefly, visit 1 involved pulmonary function testing comprising of routine spirometry and lung volumes determined by constant body-plethysmography (6200 Autobox; SensorMedics, Yorba Linda CA) performed according to ATS\/ERS guidelines.286,288  An incremental CPET was also performed on an electrically braked cycle ergometer (Ergoselect 200, SensorMedics GmbH, Bitz, Germany) with expired breath-by-breath gas analysis (Vmax 29C, SensorMedics, California, USA) according to ATS\/ACCP guidelines.62 Exercise limitations were determined according to ATS\/ACCP criteria whereby MVC was 97 estimated as FEV1*35  and HRmax was calculated as 220-age.62 Phenotypes were classified as ventilatory (VEpeak\/MVC \u226585% or MVC-VEpeak \u226411L\/min, and HRpeak <90%pred), cardiovascular (VEpeak\/MVC <85% or MVC-VEpeak >11L\/min, and HRpeak \u226590%pred) or combined (met both ventilatory and cardiovascular criteria). Visits 2-4 consisted of a CLT performed on the same cycle ergometer as the incremental CPET. Constant load trial intensities were initially individualized based on the identification of an obvious inflection in the IRV-dyspnea relationship and workloads were selected as the power outputs corresponding to 10% below (CLT-1), at (CLT-2), and 10% above (CLT-3) the inflection point. However, 48% of participants demonstrated no inflection point and workloads corresponding to 10% below, at, and 10% above the RCP were selected instead. If no inflection point or RCP could be identified (4\/23 participants), 60, 70 and 80% of maximum workload were used. All physiological parameters were measured identically to the incremental CPET with the exception of dyspnea and leg fatigue, which were measured every minute. Constant load trials were performed in a randomized order and participants remained blinded to the workload and duration of each trial. Verbal encouragement was standardized and administered by an assessment team member blinded to study outcomes.  4.2.3 Determination of MDSS For each participant, the individual dyspnea-time relationship for each CLT was plotted starting at 1-minute of exercise up to end-exercise. In order to linearize the dyspnea response, dyspnea ratings were plotted only when a change in dyspnea from the previous minute was reported. Least-square linear regression was used to calculate the slope of each dyspnea-time relationship for each CLT. The corresponding slopes (y-axis) and workloads (x-axis) were then plotted, and linear regression was used to calculate the workload corresponding to the x-intercept when y=0. This workload was predicted as the intensity corresponding to MDSS as theoretically it represented the workload at which the dyspnea slope would remain 98 at zero, indicating that dyspnea had reached a steady-state. An example of this analysis is illustrated in Figure 4.1.    Figure 4.1. An illustrative example of the dyspnea-time slope analysis used to calculate MDSS depicted using the average dyspnea-time ratings during each constant load trial in the (A) ventilatory phenotype, (B) combined phenotype and (C) cardiovascular phenotype. Panel (D) illustrates the slope-workload relationships generated from each dyspnea-time slopes calculated during the CLTs in each phenotype. The dyspnea-time slope analysis was performed for each individual patient and the workload and R2 values reported in this figure represent the mean values of all individuals for each phenotype.    99 4.2.4 MDSS Verification Trial  Participants performed a verification trial at the workload corresponding to the predicted MDSS. The exercise protocol was identical to the CLTs, however ventilatory parameters, HR, [BLa] and exertional symptoms were measured at rest, and 10, 20, and 30-minutes during exercise. Participants were not told that 30-minutes was the test\u2019s maximum duration and were instructed that they may cease exercise if their dyspnea or leg fatigue reached an intensity at which they would normally stop.    4.2.5 Statistical Analysis  Normality was assessed with the Shapiro-Wilk test. Participant characteristics, incremental CPET and CLT exercise responses were analyzed with a one-way ANOVA and Holm-Sidak post-hoc for parametric data, and Kruskal Wallis H-test and Dunn\u2019s test for non-parametric data. Exercise responses during the MDSS verification trial were analyzed with a 2-way repeated measures ANOVA for parametric data and Friedman\u2019s ANOVA on ranks for non-parametric data with Holm-Sidak post-hoc analysis to assess phenotype x time interactions. All data is presented as mean\u00b1SD. The successful prediction of MDSS was determined as a change in dyspnea of \u22641 Borg unit from 10 to 30-minutes during exercise. This time period was chosen as it was presumed that, if possible, VO2 and VE kinetics would have reached steady-state by 10-minutes into exercise in patients with COPD.305   4.3 Results  4.3.1 Characteristics and Incremental CPET Responses  The study participant flow is presented in Figure 4.2. Patient characteristics and incremental CPET responses are presented in Table 4.1. The proportion of females was greater in the combined phenotype (63%) compared to the ventilatory (29%) and cardiovascular (50%) 100 phenotypes. There were no differences in age, height or BMI between any phenotypes. The cardiovascular phenotype had a higher FEV1 and FEV1\/FVC compared to both other phenotypes. The ventilatory phenotype had a higher RV and TLC versus the combined phenotype only. There were no differences in VO2peak, Wmax, [BLa] or exertional dyspnea or leg fatigue at peak exercise between phenotypes.     Figure 4.2. Study participant flow  101 Table 4.1. Patient Characteristics and Incremental CPET Responses Variable Ventilatory (n=7) Combined (n=8) Cardiovascular (n=8)  P Male:Female 5:2 3:5 4:4 - Age (years) 65 \u00b1 9 71 \u00b1 6 68 \u00b1 8 0.30 Height (m) 1.71 \u00b1 0.10 1.66 \u00b1 0.11 1.66 \u00b1 0.09 0.50 BMI (kg\/m2) 25.3 \u00b1 3.7 26.7 \u00b1 3.1 26.2 \u00b1 4.0 0.75 FEV1 (L) 1.53 \u00b1 0.57* 1.65 \u00b1 0.41\u2021 2.42 \u00b1 0.65 0.01 FEV1 (% pred) 50 \u00b1 14*\u2020 68 \u00b1 15\u2021 91 \u00b1 14 <0.01 Range in FEV1 (% pred) 31-72 49-90 78-111 - FEV1\/FVC 37 \u00b1 8*\u2020 51 \u00b1 10\u2021 62 \u00b1 7 <0.01 RV (L) 3.62 \u00b1 0.97\u2020 2.57 \u00b1 0.61 2.72 \u00b1 0.39 0.05 TLC (L) 7.74 \u00b1 1.40\u2020 5.86 \u00b1 1.33 6.65 \u00b1 1.03 0.03 RV\/TLC (%) 46 \u00b1 6 44 \u00b1 8 41 \u00b1 7 0.38 DLCO\/VA (%) 71 \u00b1 17 76 \u00b1 16 85 \u00b1 22 0.37 VEpeak (L\/min) 50.2 \u00b1 21.0 52.2 \u00b1 15.9 58.6 \u00b1 11.0 0.58 VEpeak (% pred MVC) 93 \u00b1 8* 94 \u00b1 12\u2021 70 \u00b1 9 <0.01 VE reserve (L\/min) 2.6 \u00b1 3.4* 3.2 \u00b1 6.0\u2021 26.2 \u00b1 14.3 <0.01 HRpeak (beats\/min) 125 \u00b1 13* 139 \u00b1 9 147 \u00b1 14 <0.01 HRpeak (% pred) 80 \u00b1 9*\u2020 93 \u00b1 6 96 \u00b1 6 <0.01 Cardiac reserve (beats\/min) 31 \u00b1 16*\u2020 10 \u00b1 9 6 \u00b1 10 <0.01 Maximum workload (watts) 84 \u00b1 34 91 \u00b1 36 111 \u00b1 40 0.39 VO2peak (ml\/kg\/min) 17.0 \u00b1 5.0 17.9 \u00b1 4.6 22.0 \u00b1 4.7 0.13 VO2peak (L\/min) 1.30 \u00b1 0.54 1.34 \u00b1 0.49 1.61 \u00b1 0.53 0.45 VCO2 (L\/min) 1.34 \u00b1 0.66 1.43 \u00b1 0.51 1.71 \u00b1 0.53 0.44 RER 1.02 \u00b1 0.15 1.07 \u00b1 0.10 1.08 \u00b1 0.06 0.76 VE\/VCO2 nadir 38 \u00b1 6 35 \u00b1 4 34 \u00b1 7 0.42 PETO2 (mmHg) 106.2 \u00b1 7.3 107.3 \u00b1 4.0 109.5 \u00b1 7.5 0.29 PETCO2 (mmHg) 34.5 \u00b1 3.3 35.4 \u00b1 3.0 34.2 \u00b1 0.72 0.39 VD\/VT  0.26 \u00b1 0.09 0.25 \u00b1 0.03 0.18 \u00b1 0.06 0.04 Tidal volume (L) 1.62 \u00b1 0.52 1.56 \u00b1 0.57 2.02 \u00b1 0.67 0.23 \u0394 Tidal Volume (L) 0.63 \u00b1 0.52 0.74 \u00b1 0.50 1.21 \u00b1 0.55 0.07 \u0394 IC (L) -0.71 \u00b1 0.27 -0.44 \u00b1 0.18 -0.46 \u00b1 0.43 0.21 SpO2 (%) 93 \u00b1 2 95 \u00b1 6 96 \u00b1 3 0.11 Blood Lactate (mmol\/L) 3.7 \u00b1 2.3 4.5 \u00b1 2.1 5.3 \u00b1 1.0 0.13 Dyspnea (Borg Units) 5.6 \u00b1 1.4 4.9 \u00b1 2.1 4.5 \u00b1 2.3 0.58 Leg fatigue (Borg Units) 4.7 \u00b1 2.8 5.1 \u00b1 2.9 5.8 \u00b1 2.0 0.74 Dyspnea\/LF\/both (%) 57\/29\/14 38\/50\/12 12\/75\/12 -  Abbreviations: BMI, body mass index; FEV1, forced expiratory volume in 1 second; FEV1\/FVC, ratio of forced expiratory volume in 1 second to forced vital capacity; RV, residual volume; TLC, total lung capacity; DLCO\/VA, the ratio of diffusion of carbon monoxide to alveolar ventilation; VEpeak, peak minute ventilation; MVC, estimated maximum ventilatory capacity; HRpeak, peak heart rate; VO2peak, peak oxygen consumption; VCO2, volume of carbon dioxide produced; RER, respiratory exchange ratio; PETO2, partial pressure of end-tidal oxygen; PETCO2, partial pressure of end-tidal carbon dioxide; VD\/VT, dead space to tidal volume ratio; \u0394 Tidal volume, change in tidal volume from rest to peak exercise; \u0394 IC, change in inspiratory capacity from rest to peak exercise; SpO2, oxyhemoglobin saturation; Both, patients stated both dyspnea and leg fatigue as the limiting symptom at peak exercise. Comparisons: *p=0.05, ventilatory vs. cardiovascular. \u2020p=0.05, ventilatory vs. combined. \u2021p=0.05, combined vs. cardiovascular.   102 4.3.2 Physiological Responses during the CLTs End-exercise responses during the CLTs are presented in Table 4.2. The cardiovascular phenotype exercised at a higher relative workload compared to both other phenotypes during all CLTs. Relative workload was also higher in the combined phenotype versus the ventilatory phenotype during CLT-2 & CLT-3. Cardiac reserve was higher in both the cardiovascular and combined phenotypes versus the ventilatory phenotype during all CLTs. Additionally, the cardiovascular phenotype had a larger ventilatory reserve compared to both other phenotypes during all CLTs. Patients with a cardiovascular phenotype had a larger \u0394VT during CLT-2 & CLT3 and reached a higher [BLa] during CLT-1 compared to the ventilatory phenotype. No significant differences in dyspnea or leg fatigue were found between phenotypes during any CLTs.                             103 Table 4.2. Constant Load Exercise Responses Between Phenotypes Variable  Ventilatory (n=7) Combined (n=8) Cardiovascular (n=8)  P Constant Load Trial #1     Workload (watts) 44 \u00b1 22 56 \u00b1 24 78 \u00b1 30 0.06 Workload (%Wmax) 51 \u00b1 7* 60 \u00b1 6\u2021 69 \u00b1 3 <0.01 Duration (sec) 959 \u00b1 437 1012 \u00b1 441 628 \u00b1 249 0.12 Heart rate (beats\/min) 111 \u00b1 11*\u2020 129 \u00b1 15 138 \u00b1 14 <0.01 Cardiac reserve (beats\/min)  45 \u00b1 18*\u2020 20 \u00b1 18 14 \u00b1 10 <0.01 VE (L\/min) 39.1 \u00b1 14.1 44.5 \u00b1 15.4 49.7 \u00b1 7.0 0.29 VE reserve (L\/min) 13.8 \u00b1 6.9* 11.0 \u00b1 6.8\u2021 35.1 \u00b1 16.9 <0.01 IRV (L) 0.78 \u00b1 0.35 0.54 \u00b1 0.25 0.96 \u00b1 0.50 0.19 \u0394 Tidal volume (L) 0.47 \u00b1 0.32 0.51 \u00b1 0.55 1.05 \u00b1 0.57 0.04 \u0394 IC (L) -0.58 \u00b1 0.32 -0.44 \u00b1 0.19 -0.22 \u00b1 0.45 0.16 RER  0.92 \u00b1 0.09 0.97 \u00b1 0.06 1.00 \u00b1 0.02 0.10 Blood Lactate (mmol\/L) 2.6 \u00b1 1.6* 3.6 \u00b1 2.0 4.5 \u00b1 0.85 0.04 SpO2 (%) 93 \u00b1 2 93 \u00b1 2 96 \u00b1 3 0.14 Dyspnea (Borg unit) 5.4 \u00b1 1.5 5.1 \u00b1 1.8 5.4 \u00b1 2.8 0.96 Leg Fatigue (Borg unit) 5.0 \u00b1 2.2 5.0 \u00b1 2.9 6.4 \u00b1 2.7 0.50 Dyspnea\/LF\/both (%) 43\/28\/28 38\/38\/25 50\/38\/12 - Constant Load Trial #2     Workload (watts) 50 \u00b1 23 62 \u00b1 27 84 \u00b1 32 0.07 Workload (%Wmax) 58 \u00b1 5*\u2020 68 \u00b1 7\u2021 75 \u00b1 3 <0.01 Duration (sec) 732 \u00b1 508 890 \u00b1 600 443 \u00b1 176 0.18 Heart rate (beats\/min) 108 \u00b1 11*\u2020 132 \u00b1 15 143 \u00b1 11 <0.01 Cardiac reserve (beats\/min) 48 \u00b1 19*\u2020 17 \u00b1 16 10 \u00b1 6 <0.01 VE (L\/min) 40.7 \u00b1 14.4 45.8 \u00b1 14.4 53.8 \u00b1 8.1 0.15 VE reserve (L\/min) 12.1 \u00b1 5.6* 9.7 \u00b1 6.2\u2021 31.0 \u00b1 17.6 <0.01 IRV (L) 0.72 \u00b1 0.36 0.48 \u00b1 0.33 0.74 \u00b1 0.40 0.22 \u0394 Tidal volume (L) 0.50 \u00b1 0.41* 0.62 \u00b1 0.49 1.18 \u00b1 0.66 0.03 \u0394 IC (L) -0.57 \u00b1 0.39 -0.49 \u00b1 0.22 -0.29 \u00b1 0.29 0.18 RER  0.93 \u00b1 0.11 1.00 \u00b1 0.07 1.03 \u00b1 0.05 0.05 Blood Lactate (mmol\/L) 2.9 \u00b1 1.5 4.1 \u00b1 2.0 4.9 \u00b1 1.3 0.08 SpO2 (%) 93 \u00b1 2 93 \u00b1 2\u2021 96 \u00b1 3 0.03 Dyspnea (Borg unit) 4.9 \u00b1 1.1 4.9 \u00b1 2.1 4.9 \u00b1 2.5 1.00 Leg Fatigue (Borg unit) 4.6 \u00b1 1.6 5.1 \u00b1 2.6 5.9 \u00b1 3.0 0.61 Dyspnea\/LF\/both 43\/43\/14 25\/38\/38 50\/50\/0 -        104 Table 4.2. Constant Load Exercise Responses Between Phenotypes (Cont\u2019d) Constant Load Trial #3     Workload (watts) 58 \u00b1 25 69 \u00b1 29 93 \u00b136 0.10 Workload (%Wmax) 68 \u00b1 3*\u2020 75 \u00b1 7\u2021 83 \u00b1 3 <0.01 Duration (sec) 490 \u00b1 411 365 \u00b1 179 302 \u00b1 104 0.81 Heart rate (beats\/min) 116 \u00b1 5*\u2020 136 \u00b1 13 144 \u00b1 14 <0.01 Cardiac reserve (beats\/min) 39 \u00b1 12*\u2020 13 \u00b1 15 9 \u00b1 9 <0.01 VE (L\/min) 41.8 \u00b1 15.0 50.4 \u00b1 15.5 54.0 \u00b1 10.6 0.25 VE reserve (L\/min) 11.0 \u00b1 4.9* 5.1 \u00b1 5.8\u2021 30.8 \u00b1 15.4 <0.01 IRV (L) 0.71 \u00b1 0.27 0.49 \u00b1 0.26 0.74 \u00b1 0.35 0.23 \u0394 Tidal volume (L) 0.55 \u00b1 0.40* 0.64 \u00b1 0.61 1.28 \u00b1 0.56 0.02 \u0394 IC (L) -0.68 \u00b1 0.22 -0.36 \u00b1 0.21 -0.32 \u00b1 0.45 0.05 RER 0.96 \u00b1 0.12 1.04 \u00b1 0.10 1.07 \u00b1 0.07 0.12 Blood Lactate (mmol\/L) 3.0 \u00b1 1.8 4.2 \u00b1 2.3 5.0 \u00b1 1.8 0.19 SpO2 (%) 92 \u00b1 2* 94 \u00b1 3 96 \u00b1 2 0.01 Dyspnea (Borg unit) 5.3 \u00b1 1.3 5.1 \u00b1 2.2 4.1 \u00b1 2.0 0.45 Leg Fatigue (Borg unit) 4.6 \u00b1 1.3 5.3 \u00b1 2.5 5.9 \u00b1 2.7 0.57 Dyspnea\/LF\/both 86\/0\/14 25\/25\/50 12\/38\/50 -             Abbreviations: VO2, volume of oxygen consumed; VE, minute ventilation; IRV, inspiratory reserve volume; \u0394 Tidal volume, the change in tidal volume from rest to end-exercise; \u0394 IC, the change in inspiratory capacity from rest to end-exercise; SpO2, oxyhemoglobin saturation; LF, leg fatigue; Both, participants stated both dyspnea and leg fatigue as the limiting symptom at end-exercise. Comparisons: *p=0.05, ventilatory vs. cardiovascular. \u2020p=0.05, ventilatory vs. combined. \u2021p=0.05, combined vs. cardiovascular.  105 4.3.3 Exercise Responses during the MDSS Verification Trial Exercise responses during the MDSS verification trial are presented in Table 4.3 and Figures 4.3 & 4.4. MDSS could not be calculated in a subgroup of patients (n=4) (Figure 2). Two patients did not reach 30-minutes and stopped exercise at ~25-minutes due to general fatigue (n=1; cardiovascular phenotype), and ~16-minutes due to dyspnea and leg fatigue accompanied by a relatively sudden drop in diastolic blood pressure of 20mmHg (n=1; combined phenotype). The patient with a cardiovascular phenotype reported a dyspnea rating of 2 Borg units at 10-minutes which increased to 3 Borg units at exercise cessation. The patient with a combined phenotype reported a dyspnea rating of 4 Borg units at 10-minutes and a final dyspnea rating of 5 Borg units at exercise cessation. As such, it is possible that MDSS was accurately predicted had these individuals managed to complete 30-minutes.  However, as MDSS was defined as the change in dyspnea of \u22641 Borg unit from 10 to 30-minutes, these patients were removed from the verification trial analysis.  The relative workload corresponding to MDSS was lower in the ventilatory phenotype compared to both the cardiovascular and combined phenotypes (ANOVA p=0.05). MDSS was accurately predicted in 100% of patients with a ventilatory phenotype as all patients completed 30-minutes and no patient reported an increase in dyspnea >1 Borg unit from 10 to 30-minutes. In the ventilatory phenotype, dyspnea at 10-minutes was 1.5\u00b11.3 Borg units and rose to 2.3\u00b11.5 Borg units at 30-minutes, representing an average change in dyspnea (\u0394dyspnea) of 0.8\u00b10.3 Borg units. In the cardiovascular phenotype, MDSS was accurately predicted in only 2\/5 patients and \u0394dyspnea was 1.7\u00b11.0 Borg units (2.1\u00b11.0 Borg units at 10-minutes versus 3.8\u00b11.6 Borg units at 30-minutes).  In the combined phenotype, 3\/6 patients successfully achieved MDSS and \u0394dyspnea was 1.3\u00b10.8 Borg units (1.3\u00b11.4 Borg units at 10-minutes versus 2.7\u00b11.8 Borg units at 30-minutes). There was a wide range in individual dyspnea ratings at end-exercise across all phenotypes, however no differences in dyspnea or \u0394dyspnea were found between phenotypes. Absolute EILV and EELV were higher at end-106 exercise in the ventilatory phenotype compared to both other phenotypes, however there were no differences between phenotypes when lung volumes were expressed relative to TLC. There were no significant differences in IRV, \u0394IC or \u0394VT between phenotypes at end-exercise. Throughout exercise and at end-exercise, ventilatory reserve was greater in the cardiovascular phenotype compared to both other phenotypes. At end-exercise, the combined phenotype reached a higher percentage of HRmax compared to the ventilatory phenotype, however HR at end exercise was not different compared to the cardiovascular phenotype. Throughout exercise, [BLa] was higher in the combined versus ventilatory phenotype, and the combined phenotype reached a higher end-tidal CO2 compared to the cardiovascular phenotype. Leg fatigue was higher in the cardiovascular phenotype versus both other phenotypes at end-exercise.                 107 Table 4.3. End Exercise Responses in Patients who Achieved MDSS Variable Ventilatory (n=6) Combined  (n=6) Cardiovascular (n=5) P Male:Female 4:2 2:4 1:4 - Achieved MDSS (n) 6\/6 3\/6 2\/5 - Dyspnea range (Borg units) 1-4 0.5-5 1-5 - Workload (watts) 29 \u00b1 14 50 \u00b1 26 46 \u00b1 27 0.22 Workload(%Wmax) 38 \u00b1 9 52 \u00b1 7 47 \u00b1 12 0.05 Heart rate (beats\/min) 97 \u00b1 13 116 \u00b1 13 112 \u00b1 22 0.12 Cardiac reserve (beats\/min) 61 \u00b1 20\u2020 33 \u00b1 16 36 \u00b1 16 0.03 VE (L\/min) 30.3 \u00b1 9.5 36.1 \u00b1 10.0 38.2 \u00b1 10.4 0.52 Ventilatory reserve (L\/min) 18.1 \u00b1 9.5* 21.1 \u00b1 9.3\u2021 36.3 \u00b1 10.4 <0.01 VO2 (ml\/kg\/min) 11.4 \u00b1 2.7 13.5 \u00b1 3.3 16.0 \u00b1 4.9 0.15 VO2 (% peak) 72 \u00b1 11 79 \u00b1 14 75 \u00b1 8 0.53 VO2 (L\/min) 0.81 \u00b1 0.24 1.03 \u00b1 0.32 0.99 \u00b1 0.33 0.41 VCO2 (L\/min) 0.72 \u00b1 0.24 0.95 \u00b1 0.30 0.90 \u00b1 0.32 0.38 RER  0.89 \u00b1 0.08 0.93 \u00b1 0.06 0.91 \u00b1 0.02 0.68 PETO2 (mmHg) 104.5 \u00b1 6.5 104.6 \u00b1 1.4 108.3 \u00b1 2.3 0.08 PETCO2 (mmHg) 33.2 \u00b1 3.6 35.2 \u00b1 1.4\u2021 30.9 \u00b1 1.7 0.04 VD\/VT  0.31 \u00b1 0.07 0.28 \u00b1 0.03 0.25 \u00b1 0.05 0.13 EILV (L) 6.68 \u00b1 1.25*\u2020 4.95 \u00b1 0.87 5.03 \u00b1 0.69 0.02 EELV (L) 5.37 \u00b1 1.14*\u2020 3.51 \u00b1 0.67 3.74 \u00b1 0.70 <0.01 IRV (L) 0.78 \u00b1 0.28 0.83 \u00b1 0.51 1.02 \u00b1 0.48 0.57 Tidal volume (L) 1.31 \u00b1 0.25 1.40 \u00b1 0.45 1.29 \u00b1 0.18 0.84 \u0394 Tidal volume (L) 0.43 \u00b1 0.18 0.60 \u00b1 0.38 0.58 \u00b1 0.10 0.50 \u0394 IC (L) -0.40 \u00b1 0.28 -0.15 \u00b1 0.16 -0.26 \u00b1 0.21 0.20 SpO2 (%) 94 \u00b1 2 94 \u00b1 2 95 \u00b1 1 0.28        Abbreviations: MDSS, maximum dyspnea steady-state; VE, minute ventilation; VO2, volume of oxygen consumed; VCO2, volume of carbon dioxide produced; RER, respiratory exchange ratio; VE\/VCO2, ratio of minute ventilation to carbon dioxide produced; PETO2, partial pressure of end-tidal oxygen; PETCO2, partial pressure of end-tidal carbon dioxide; VD\/VT, deadspace to tidal volume ratio; EILV, end-inspiratory lung volume; EELV, end-expiratory lung volume; IRV, inspiratory reserve volume; ; \u0394Tidal volume, the change in tidal volume from rest to 30-minutes; \u0394IC, change in inspiratory capacity from rest to 30-minutes; SpO2, oxyhemoglobin saturation;  ; Both, participants stated both dyspnea and leg fatigue as the limiting symptom at 30-minutes. Comparisons: *p=0.05, ventilatory vs. cardiovascular. \u2020p=0.05, ventilatory vs. combined. \u2021p=0.05, combined vs. cardiovascular. \u00a7p=0.05, subgroup of patients in whom MDSS could not be calculate vs. phenotype.   108   Figure 4.3. Phenotypic responses in (A) dyspnea, (B) the change in dyspnea from 10 to 30-minutes, (C) end-inspiratory lung volume (EILV) and end-expiratory lung volume (EELV), and (D) the change in tidal volume from rest to 30-minutes. The dashed lines in panel D represent the mean change in tidal volume from rest to peak exercise achieved during the incremental CPET. The vertical distance from the last data point at 1800 seconds to the dashed line represents the reserve in tidal volume expansion. The ventilatory phenotype (VL) reached the same amount of tidal volume expansion during the MDSS verification trial as during the incremental CPET, and therefore had a minimal reserve to further expand tidal volume. However, both the combined phenotype (COM) and cardiovascular phenotypes (CVL) had a larger reserve for tidal volume expansion depicted by the brackets.   109  Figure 4.4. The change in (A) relative minute ventilation (expressed as a percentage of estimate maximal ventilatory capacity [MVC]), (B) relative heart rate (expressed as a percentage of age-predicted maximum heart rate), (C) capillary blood lactate concentration, and (D) leg fatigue during the MDSS verification trial. *p=0.05, ventilatory vs. cardiovascular. \u2020p=0.05, ventilatory vs. combined. \u2021p=0.05, combined vs. cardiovascular.         110 4.3.4 Secondary Analysis: Patients in Whom MDSS Could not be Calculated  As MDSS could not be calculate in a subset of patients (n=1 ventilatory, n=1 combined, n=2 cardiovascular), a secondary post-hoc analysis was performed to assess differences in characteristics and incremental CPET responses between the subgroup and phenotypes (Table A.1 (Appendix)). In the subgroup, all patients were male and had a higher FEV1 compared to the ventilatory phenotype and a larger TLC compared to the combined phenotype.  At peak exercise, patients in the subgroup reached a higher HR (%pred) and had a smaller cardiac reserve compared to the ventilatory phenotype. Additionally, VT and \u0394VT were larger and [BLa] was higher versus the ventilatory phenotype. VD\/VT was lower in the subgroup compared to both the ventilatory and combined phenotypes. The subgroup analysis of peak CLT responses are presented in Table A.2 (Appendix). Relative workload was higher during CLT-1, whereas HR and \u0394VT were greater during CLT-2 and CLT-3 in the subgroup compared to the ventilatory phenotype. Patients in the subgroup also reached a higher [BLa] during CLT-2 compared to the ventilatory phenotype, and IRV was greater during CLT-3 versus the combined phenotype.            111 4.5 Discussion  Across all phenotypes, the novel concept of MDSS could be calculated in 83% of patients with COPD. However, our hypothesis was only partially supported as MDSS, defined as an increase of \u22641 Borg unit between 10 to 30-minutes of continuous exercise was achieved in 58% of patients. During the MDSS verification trial, 100% of patients with a ventilatory phenotype successfully achieved MDSS; however, dyspnea remained stable in only 50% and 40% of patients with a combined or cardiovascular phenotype, respectively.  4.5.1 Physiological Responses at the Predicted MDSS  MDSS was accurately predicted in 100% of patients with a ventilatory phenotype as dyspnea did not rise >1 Borg unit following 10 to 30-minutes of exercise during the verification trial. In the ventilatory phenotype, dyspnea remained constant likely due to optimal coupling between inspiratory neural drive and the efferent output of the respiratory system. During exercise, inspiratory neural drive increases due to increased afferent input from different stimuli including increased VCO2 and H+, changes in PaCO2 and PaO2, increased type III & IV afferent activity and stretch receptors within the lungs and chest wall.307 In COPD, static and dynamic hyperinflation increases the work of breathing and can act to mechanically constrain the ability of the respiratory system to efficiently increase VE (and thus VA) to match inspiratory neural drive.110,138,139,148,213 When a mismatch occurs between the inspiratory neural drive to breathe and the mechanical response of the respiratory system (i.e., neuromechanical uncoupling), dyspnea rises to intolerable levels and exercise is ceased.138,139,151 In the ventilatory phenotype, exercise at MDSS was associate with a moderate amount of dynamic hyperinflation (~400ml), a reduction in IRV to ~780ml, and a minimal reserve for VT expansion (Figure 3D).  However, [BLa] at MDSS was relatively low in the ventilatory phenotype such that at it did not differ from resting values at 30-minutes (p=0.15). This indicates that ATP production was primarily met by aerobic metabolism and that additional contributions to the 112 drive to breathe from increased VCO2, PaCO2, and H+ were likely minimal. Therefore, as VA was sufficient to meet metabolic demand without a continued rise in VE or lung volumes, MDSS was likely positioned just below the intensity where neuromechanical uncoupling would have occurred. As such, inspiratory neural drive reached a steady-state enabling all patients with a ventilatory phenotype to sustain continuous exercise for 30-minutes without a rise in dyspnea. While the predicted MDSS represented a sustainable intensity in almost all patients independent of phenotype (i.e., 17\/19 patients completed 30-minutes of continuous exercise), MDSS was successfully achieved in only ~50% of patients with a combined or cardiovascular phenotype. This demonstrates that MDSS, as per our definition, was less applicable to patients in whom Q and convective O2 delivery contribute to exercise limitation. At end-exercise, IRV reached ~800ml and ~1.00L in the combined and cardiovascular phenotypes, respectively, and both phenotypes had a large VT reserve (Figure 3D). Blood lactate concentration was higher in both the combined and cardiovascular phenotypes compared to the ventilatory phenotype (2-way RM ANOVA p=0.01) as MDSS was predicted at a higher relative workload. Therefore, in certain patients with a combined or cardiovascular phenotype, inspiratory neural drive likely continued to rise due to chemoreceptor stimulation associated with increased CO2 and H+ production from anaerobic glycolysis, as well as potentially other peripheral sensors (i.e. type III & IV afferents in skeletal muscles). However, all patients with a cardiovascular or combined phenotype still successfully completed 30-minutes at MDSS as [BLa] reached a steady state in all patients (as demonstrated by a change in [BLa] of <1.0mmol\/L between 10 to 30-minutes of exercise271) indicating that lactate production was appropriately matched by metabolism\/clearance. Considering that there was no evidence that a mechanical constraint to ventilation had been reached, any further increase in neural drive to breathe could have been matched by a mechanical increase in VE. As such, the rise in 113 dyspnea reported in these individuals appeared to be due to an increase in VE rather than to neuromechanical uncoupling, per se.  In the six patients with a combined or cardiovascular phenotype who did not achieve MDSS, leg fatigue alone (n=5) or in combination with dyspnea (n=1) was reported as the limiting symptom at peak exercise during the incremental CPET. This was in contrast to the ventilatory phenotype, in which MDSS was successfully achieved regardless of if patients reported dyspnea or leg fatigue as the main limiting symptom. It has been previously demonstrated that exercise tolerance in patients with a ventilatory phenotype is governed by changes in lung volumes and VT constraint (Study #2, Chapter 3). However, in patients with a cardiovascular contribution to exercise limitation, convective O2 delivery and\/or the accumulation of fatigue related metabolites play a larger role in governing exercise tolerance, and therefore the concept of MDSS may be less relevant in patients with a cardiovascular or combined phenotype. At end exercise during the MDSS verification trial, leg fatigue was significantly higher in the cardiovascular phenotype versus both other phenotypes. On average, MDSS was positioned ~5% below AT in the cardiovascular phenotype, compared to ~45% and ~20% below AT in the ventilatory and combined phenotypes, respectively.  Therefore, patients with a cardiovascular phenotype were exercising at a higher intensity relative to AT than both other phenotypes, which may explain the higher leg fatigue intensity reported in these patients.   4.5.2 Subgroup Analysis  Using the dyspnea-time slope methodology, MDSS could not be mathematically calculated in 4 patients. In two patients (n=1 cardiovascular and n=1 combined), the calculated workload corresponding to MDSS was negative (-110watts and -96watts, respectively). However, in the two other patients (n=1 cardiovascular and n=1 ventilatory), the MDSS workload was extremely high (+1092watts and +469watts, respectively) due to a negative slope-workload 114 relationship (i.e., as workload increased slope decreased). One reason for this may be the similarity of the dyspnea-time slopes calculate between each CLTs. In all subgroup patients, the dyspnea-time slopes and duration of CLTs were less variable compared to other patients within the same phenotype. This indicates that these patients experienced a similar dyspnea response across all three CLTs regardless of the difference in workload and suggests that dyspnea may not be a key limiting factor to exercise in these individuals. By using three data points that occurred at similar slopes but different workloads, the slope coefficient of the slope-workload relationship (Figure 2D) was likely to be closer to zero indicating a weak relationship, and thus reducing the accuracy of predicting MDSS within a physiologically possible range.  This finding suggests that MDSS may not be able to be calculated in patients in whom the dyspnea-time response is less sensitive to different workloads, and\/or that the workloads selected for the CLTs were too similar in these individuals. It is likely that an accurate MDSS workload could have been predicted had the difference between CLT workloads been greater.    4.5.3 Practical Application of MDSS A wide-range of individual dyspnea ratings (between 0.5-5 Borg units) were reported across all phenotypes at end-exercise during the MDSS verification trial. As such, using a standardized dyspnea rating of 4-6 Borg units (somewhat severe-very severe) to prescribe a moderate-high intensity exercise session will: 1) result in a considerable range of exercise durations, and 2) elicit highly different metabolic responses between patients. Therefore, the volume of training achieved as well as the internal metabolic load elicited will vary considerably between patients resulting in a wide range of physiological adaptations. Although MDSS was successfully achieved in 100% of patients with a ventilatory phenotype, the absolute and relative workloads associated with MDSS were only ~30watts and ~40%Wmax, respectively. This intensity is well below the recommended intensity for aerobic exercise training in COPD of >60%Wmax for >20-minutes.44 However, previous investigations 115 in moderate-severe COPD have reported that physiological training effects can still be attained when cycling at lower intensities (ranging between 25-60%Wmax and at a similar average absolute workload of 30-35watts) for a similar duration of ~30-minutes over the course of a 12-week training program.308 Therefore, exercising at MDSS should induce important physiological adaptations for patients with a ventilatory phenotype, and as such, may be the ideal initial intensity to initiate important cardiometabolic adaptations when beginning an aerobic training program. Considering that the ventilatory phenotype represents the largest percentage of patients compared to the cardiovascular or combined phenotypes (as reported in Study #1, Chapter 2), MDSS be a useful intensity to prescribe a longer duration (i.e., 30-minute) continuous aerobic exercise session in a rehabilitation setting where patients may be more likely to present with a classic ventilatory limitation to exercise.   4.5.4 Study Considerations  It should be acknowledged that the small sample size within each phenotype could have resulted in a type II error.  Also, sex-related differences in dyspnea have been documented in COPD, with females reporting greater dyspnea than males for a given absolute workload or VE.309  As a greater portion of patients with a combined or cardiovascular phenotype were female compared to the ventilatory phenotype, this may have resulted in a MDSS being predicted at a lower absolute workload. However, as MDSS represents a workload based on the individual dyspnea responses to exercise, the influence of sex-differences in dyspnea was minimized.  The linearization of the dyspnea responses during the CLTs could have reduced the accuracy of the predicted MDSS as the dyspnea response may be more curvilinear or exponential in certain patients. Finally, as we did not perform a verification trial slightly above MDSS we cannot say with absolute certainty that MDSS was accurately predicted. However, relatively small changes in workload (i.e.,~5watts) have been previously demonstrate to have a significant effect on exercise tolerance, especially in patients with a ventilatory phenotype 116 (Study #2, Chapter 3). As such, the finding that all patients could complete 30-minutes of continuous exercise at MDSS suggest that MDSS is a feasible intensity to initially prescribe a longer duration aerobic exercise session in patients with COPD.  4.5.6 Conclusion  The novel concept of MDSS could be calculated in a large portion of patients with COPD regardless of exercise limitation phenotype. MDSS was successfully achieved in 100% of patients with a ventilatory phenotype and occurred at a relatively low workload that was primarily associated with aerobic metabolism. MDSS occurred at a higher workload and metabolic load in the cardiovascular and combined phenotypes compared to the ventilatory phenotype. Although MDSS was achieved in only ~50% of patients with a cardiovascular contribution to exercise limitation, almost all patients still completed 30-minutes of exercise despite dyspnea continuing to rise. While MDSS was less applicable to patients with a combined or cardiovascular phenotype who reported leg fatigue as a contributing factor to exercise limitation, it may be an ideal intensity to prescribe a longer duration aerobic exercise session in patients with a ventilatory phenotype.    117 Chapter 5: General Discussion and Conclusion  5.1 Overall Summary  Exercise limitation in COPD is multifactorial and complex due to the pathophysiological heterogeneity of the disease. However, due to the considerable volume of literature that has identified that abnormal changes in lung volumes lead to intolerable dyspnea that impairs exercise tolerance, other potential limitations (especially in sub-groups of patients) have often been overlooked. Traditionally, cardiovascular contributions to exercise limitation have been considered minimally important as it is commonly assumed that incremental exercise is terminated due to pulmonary causes well-before a significant reduction in cardiac reserve occurs. However, this dissertation provides novel empirical evidence that by using well-accepted criteria for meeting a ventilatory or cardiovascular limitation to exercise, distinct phenotypes of exercise limitation can be identified in COPD that are not solely dependent upon FEV1. Collectively, these findings indicate that a cardiovascular contribution to exercise limitation occurs in a substantial portion of patients with COPD (52%). Additionally, this work supports that the three phenotypes of exercise limitation are clinically relevant as they are associated with recognized prognostic indicators (i.e., markedly different VO2peak, Wmax and ventilatory responses), and that classifying patients phenotypically may aid in the optimization of exercise prescription for rehabilitative purposes.  In Study #1, it was observed that approximately half of the study cohort reached HRpeak \u226590%HRmax independent of the presence of a ventilation limitation indicating that Q and convective O2 delivery contribute to exercise limitation in a significant number of patients. Furthermore, the integrated exercise responses observed in patients with a cardiovascular phenotype (~20% of the study cohort) were similar to healthy aging. This novel finding demonstrated that while mild alterations in pulmonary mechanics may partially contribute to exercise limitation, they are not the primary limitation to VO2peak in a subgroup of patients with COPD. Additionally, mean VO2peak was significantly higher in the cardiovascular and 118 combined phenotypes versus the ventilatory phenotype with 56% and 48% of the patients with a cardiovascular or combined phenotype, respectively, attaining >84% of age-and sex-predicted VO2peak compared to only 13% of the ventilatory phenotype. These data suggest that the exercise limitation phenotypes may be useful to identify important physiological differences between patients that could have prognostic implications.  The findings of Studies #2 & #3 provide further evidence to support that the three phenotypes represent distinct subgroups of patients as the physiological responses to constant load exercise performed at 2 novel indices of sustainability (i.e., CP30 and MDSS) differed between phenotypes. Both CP30 and MDSS occurred at a higher workload and greater metabolic load in patients with a cardiovascular contribution to exercise limitation compared to those with a classic ventilatory limitation phenotype. These data also demonstrate the pitfalls of using a generic exercise prescription and offer a new pragmatic framework to prescribe intensity and duration for aerobic exercise training in COPD. Additionally, while the highly novel concept of MDSS appears to be an appropriate alternative index to determine a sustainable exercise intensity in patients with a ventilatory phenotype, it is less applicable in patients with a cardiovascular or combined phenotype who are less limited by the pulmonary system and likely have different etiologies of exertional dyspnea.  This finding challenges the dogma that exertional dyspnea is a key cause of exercise intolerance in all individuals with COPD110,113,138,148 and supports that other factors (e.g., O2 delivery and peripheral muscle fatigue) may contribute to exercise limitation to a greater extent in certain individuals.   5.2 Clinical Implications of the Exercise Limitation Phenotypes  Three phenotypes of exercise limitation were identified in this dissertation that were associated with distinct submaximal and maximal physiological exercise responses observed during an incremental CPET. In patients with a cardiovascular phenotype, VO2peak was limited 119 by Q and convective O2 delivery indicated by the obtainment of HRmax. This was possible due to reduced static and dynamic lung volumes, normal VT expansion, and reduced isoload dyspnea in the cardiovascular versus ventilatory phenotype. Additionally, the cardiovascular phenotype had a relatively normal hyperventilatory response beyond the RCP, and achieved a higher Wmax, VO2peak, and VEpeak compared to the ventilatory phenotype. In contrast, patients with a ventilatory phenotype had greater static and dynamic hyperinflation that resulted in VT constraint and an increase in dyspnea leading to the curtailment of exercise at a lower Wmax, VO2peak and HRpeak compared to both other phenotypes. Although the combined phenotype exhibited a similar degree of dynamic hyperinflation (i.e., \u0394IC) to the ventilatory phenotype, relative EELV was lower at isoload and peak exercise due to less static hyperinflation (FRC was lower on average, by ~620ml) compared to the ventilatory phenotype. Therefore, HR continued to rise throughout exercise for a longer duration before VT constraint occurred compared to the ventilation phenotype. As such, the combined phenotype achieved a higher HRpeak, Wmax, VO2peak and VEpeak versus ventilatory phenotype, and similar peak values to the cardiovascular phenotype.  Although these findings which are presented in Study #1 (Chapter 2), clearly demonstrate that the three phenotypes are associated with distinct ventilatory, cardiovascular and metabolic responses to incremental exercise, it is logical that a range of exercise responses exist even within phenotypes. For example, although all patients with a cardiovascular phenotype had a significant ventilatory reserve at peak exercise, VEpeak\/MVC and MVC - VEpeak ranged between 42-84% and 12-78L\/min, respectively. Similarly, even though a large cardiac reserve was observed in the ventilatory phenotype, HRpeak and cardiac reserve (i.e., HRmax - HRpeak) ranged between 60-88%pred and 19-65 beats per minute, respectively. These ranges demonstrate that even within the same phenotype, certain patients may lie in closer proximity to the boundary of a ventilatory or cardiovascular limitation.  As such, exercise limitation in COPD is likely a continuum (Figure 5.1) with the phenotypes 120 representing specific physiological transitions that demarcate when the pulmonary and\/or cardiovascular system significantly contributes to the attainment of VO2peak.    Figure 5.1. A theoretically model of the exercise limitation continuum in patients with COPD. The grey and white shaded areas signify the ventilatory or cardiovascular contributions to exercise limitation, respectively. Across all three phenotypes, both the pulmonary and cardiovascular systems contribution to exercise limitation but to varying degrees, likely mediated by skeletal muscle function. Reduced skeletal muscle function and\/or dysfunction may be associated with an increase in ventilatory contributions to exercise limitation. However, with maintained or improved skeletal muscle function, a cardiovascular contribution may contribute to exercise limitation to a greater degree. On the far left of the continuum, patients with very severe disease (i.e., FEV1 <1.0L) are unlikely to ever be able to reach a cardiovascular limitation even if drive to breathe is reduced because ventilatory capacity is so severely limited. In patients with a cardiovascular phenotype who are positioned on the far right of the continuum, physiological alterations to the pulmonary system would be expected to occur to a greater degree than would be associated with aging (i.e., small airway dysfunction and gas exchange abnormalities). Despite being well below the boundary of a ventilatory limitation to exercise, these pulmonary alterations would still contribute to the incremental exercise responses observed.           121 This concept of a continuum of exercise limitations offers a novel integrative paradigm to understand factors\/interventions that may modify disease state in COPD. As proposed in Study #1 (Chapter 2), one key regulating factor that likely determines the contribution of a ventilatory and\/or cardiovascular limitation and thus phenotype classification is skeletal muscle function. In patients with skeletal muscle deconditioning and\/or dysfunction, drive to breathe is increased for a given workload due to increased CO2 and H+ production from a greater reliance on anaerobic glycolysis and increased activation of type III & IV afferents.48,131,170,310 As such, dynamic hyperinflation resulting in VT constraint is likely to be accelerated leading to a ventilatory limitation at a lower workload before HRmax is reached. This scenario would result in a leftward regression on the continuum and the transition from a combined phenotype to a ventilatory phenotype.  On the contrary, it is interesting to consider that if the ventilatory drive to breathe is reduced through optimal exercise conditioning and medical therapy, dynamic hyperinflation leading to the development of VT constraint and dyspnea are likely to be delayed.  As such, certain patients with a ventilatory phenotype may be able to reach a higher HRpeak during incremental exercise and patients who are close to being able to obtain HRmax may thus be able to transition to a combined phenotype. Similarly, an individual with a combined phenotype may attenuate ventilatory drive enough through conditioning that HR becomes the primary limitation to exercise, thus transition to a cardiovascular phenotype. If possible, this transition is likely to be accompanied by improvements in cardiopulmonary fitness (i.e., VO2peak) and reduced exertional symptoms at submaximal workloads, which have important implications for quality of life and prognosis in COPD.297     122 As FEV1 alone remains a poor predictor of VO2peak in COPD, previous studies have sought to identify alternative variables associated with a reduced VO2peak in this patient population. A large retrospective study performed by Neder et al., investigated whether mechanical inspiratory constraint (defined as EILV\/TLC \u226590%) and\/or poor ventilatory efficiency (VE\/VCO2 nadir >34) were more strongly associated with a reduction in VO2peak than breathing reserve (i.e., VEpeak <20% of MVC) in COPD.121 It was observed that the combination of mechanical inspiratory constraint and ventilatory inefficiency, independent of breathing reserve was highly predictive of a reduced VO2peak (i.e., VO2peak <1.04L\/min (the sample\u2019s median) and <50%pred) across the entire disease severity spectrum in COPD.121   The data presented in Study #1 (Chapter 2) of this dissertation are in contrast to the findings of Neder et al., as VO2peak was significantly greater in the cardiovascular and combined phenotypes compared to the ventilatory phenotype. These results demonstrate that the identification of a ventilatory limitation in the presence of a cardiac reserve is associated with a reduced VO2peak; however, a cardiovascular limitation is associated with a higher VO2peak, independent of the presence of a ventilatory limitation.  A higher VO2peak occurred in patients with a cardiovascular contribution to exercise limitation despite ~75% of patients with a cardiovascular or combined phenotype reaching an EILV\/TLC \u226588% and ~40% having a VE\/VCO2 nadir >34.  In these patients, HRmax was reached due to the continued ability to increase VE thus resulting in a higher VEpeak, Wmax and VO2peak. This was opposed to the ventilatory phenotype, in which 100% and 61% of patients reached EILV\/TLC \u226588% or VE\/VCO2 nadir >34, respectively, but significant dynamic hyperinflation occurred such that VT became constrained and ventilatory reserve was dramatically reduced. As such, exercise was terminated at a lower VEpeak and before HRmax was reached indicating a true ventilatory limitation to exercise. Therefore, as EILV reached \u226588% of TLC in a high percentage of all patients and the degree of ventilatory inefficiency was highly variable between phenotypes, 123 classifying patients phenotypically may be a more robust means to identify patients with a reduced VO2peak.  In addition to these observations, when VO2peak was categorized based on quartiles the lowest quartile of patients consisted exclusively of those with a ventilatory or combined phenotype with a mean VO2peak of ~12ml\/kg\/min (~55%pred) (Figure 5.2). This was in contrast to the highest quartile, in which patients with a cardiovascular or combined phenotype represented ~88% of the sample and mean VO2peak was ~23ml\/kg\/min (~95%pred) demonstrating that the majority of these patients had a normal age-and-sex matched VO2peak (i.e., 83% had a VO2peak \u226580%pred). A VO2peak <5 METs (<17.5ml\/kg\/min) has been established as one of the strongest predictors of both all-cause and cardiovascular mortality, independent of health or chronic condition.210 In individuals with COPD, a VO2peak <17.5ml\/kg\/min has been associated with a relative risk of death from any cause of 1.0-2.7 compared to patients with a VO2peak >8 METs (>28ml\/kg\/min).210 Furthermore, a VO2peak <14ml\/kg\/min is considered an important prognostic indicator in other clinical populations such as heart failure.311 As such, the observation that 30\/46 patients with a ventilatory phenotype versus 3\/16 patients with a cardiovascular phenotype fell within the lower two quartiles for VO2peak, suggests that the identification of patients according to exercise limitation phenotype is clinically relevant and may be of prognostic importance.   124  Figure 5.2. The relative proportion of phenotypes associated with each VO2peak quartile. VO2peak %predicted calculated using the FRIEND database.312  5.3 Application of the Exercise Limitation Phenotypes to Exercise Prescription  Data presented in Study #2 (Chapter 3) revealed that CP occurred at a higher absolute and relative workload in patients with a cardiovascular or combined phenotype compared to the ventilatory phenotype. This demonstrated that the boundary which separates sustainable intensities (i.e., workloads < CP) that can be maintained for a relatively long duration (i.e., >30-minutes) from intensities that are predictably limited (i.e., workloads >CP) was higher in patients who have a cardiovascular contribution to exercise limitation, independent of the presence of a ventilatory limitation. This difference was due to the underlying physiological exercise responses associated with each phenotype. In the cardiovascular phenotype, CP appears to be determined by changes in aerobic-anaerobic energy contributions and the accumulation\/clearance of fatigue-related metabolites as all patients maintained a ventilatory reserve and almost all reached HRmax at Tlim during the constant load trials, despite exercising at higher workloads compared to the ventilatory phenotype.  In contrast, CP in the ventilatory phenotype was governed by a mechanical constraint to VT expansion due to greater static and dynamic hyperinflation that resulted in an attenuated increase in VT and significantly lower 125 HR at Tlim compared to the cardiovascular phenotype. In the combined phenotype, CP was determined by a combination of both VT constraint and the limits of aerobic metabolism.  The phenotype specific differences in the workload associated with CP and the physiological factors that govern this boundary of sustainability, have important implications for exercise prescription. This was revealed by the application of the CP concept to predict the maximum intensity that could be sustained continuously for 30-minutes (i.e., CP30). CP30 occurred at a higher workload that was associated with a greater metabolic load in both the cardiovascular and combined phenotypes versus the ventilatory phenotype (~55-60%Wmax versus ~45%Wmax). This demonstrated that patients with a cardiovascular contribution to exercise limitation can sustain a greater level of metabolic perturbation while exercising for the same duration as those with a classic ventilatory phenotype. This finding speaks to the critical importance of prescribing exercise intensity according to the phenotype of exercise limitation as it illustrates that the prescription of a generic exercise intensity, even if individualized (i.e., 60%Wmax) will result in a range of exercise durations achieved between patients due to differences in the power-duration relationship between different exercise limitation phenotypes.  In an ideal setting, exercise prescription would be completely individualized to each patient based on their unique exercise limitation, physiological responses and individual goals. However, in pulmonary rehabilitation through necessity, exercise prescription is commonly performed on a trial-and-error basis (particularly at the beginning of a training program) where intensity is generically prescribed and then adjusted according to the patient\u2019s exertional symptoms. While it is acknowledged that it is not practical for clinicians to perform multiple constant load trials to determine each patient\u2019s CP, based on the findings of Study #2 classifying patients phenotypically in addition to evaluating their physiological responses to incremental exercise may offer a more accurate means to optimize exercise prescription.  A key finding from Study #2 was that on average, CP30 occurred at ~20-30% below AT in the 126 ventilatory phenotype, ~10% below AT in the combined phenotype and ~10% above AT in the cardiovascular phenotype. Therefore, by performing an incremental CPET and simply identifying the exercise limitation phenotype and AT, practitioners can easily estimate CP30 as an initial starting intensity for an aerobic training program.  Small adjustments in intensity may be required to ensure all individuals can sustain exercise for 30-minutes; however, these adjustments will be minor compared to the trial-and-error approach to exercise prescription. It is interesting to note that although AT could be identified in all patients regardless of phenotype (determined by ventilatory equivalents and the modified Beaver plot), the RCP could not be clearly defined in any patients with a ventilatory phenotype. However, the RCP was identified in a larger portion of patients with a combined phenotype (~65%) and generally occurred shortly before test termination as on average, patients went ~15 watts past the RCP. In contrast, RCP was identified in all patients with a cardiovascular phenotype. When AT, RCP, CP, CP30 and MDSS are plotted and compared across phenotypes (Figure 5.3), it can be observed that with the transition from a ventilatory to combined phenotype and combine to cardiovascular phenotype, CP and therefore CP30 shift closer towards the AT. Considering that exercise tolerance in the ventilatory phenotype was curtailed due to a rise in lung volumes leading to a mechanical constraint to ventilation, it is logical that CP30 would be positioned well-below AT where metabolism is predominantly aerobic and the influence of metabolites that contribute to an increased drive to breathe (e.g., H+, CO2, K+) are tightly regulated.  Although statistically, there were no differences between absolute CP30 and MDSS in either the ventilatory or combined phenotypes (p>0.05 for both comparisons), MDSS occurred on average ~5 watts lower in both phenotypes. Therefore, while CP30 represents the maximal workload that can be maintained for 30-minutes, if the goal of an exercise session is to maintain dyspnea at a steady-state the intensity will likely need to be reduced by ~5 watts below CP30. As such, MDSS may be an ideal initial intensity for patients with a ventilatory phenotype who may be anxious due to the anticipation of exertional dyspnea during exercise.    127 However, in the cardiovascular phenotype MDSS occurred at a significantly lower absolute (~10 watts) and relative intensity (~15%Wmax) compared to CP30 (p=0.03 for both comparisons). Interestingly, three patients with a cardiovascular phenotype who were unable to complete 30-minutes at CP30, successfully completed 30-minutes at MDSS. In these individuals, MDSS was position below AT as opposed to CP30 which was either at or above AT. Therefore, the ability of these individuals to complete 30-minutes at MDSS was likely due to a reduction in the accumulation of fatigue-related metabolic by-products due to anaerobic glycolysis.   128    Figure 5.3. The different exercise intensity domains in the (A) ventilatory phenotype, (B) combined phenotype, and (C) cardiovascular phenotype expressed relative to the maximum workload achieved during the incremental CPET. The average relative workloads corresponding to CP30 and MDSS are also presented. The average relative workload corresponding to the three constant load trials as well as the respective average durations achieved during each trial are also presented. The RCP was clearly identified according to the specified methodology in ventilatory n=0\/10, combined n=7\/11, and cardiovascular n=9\/9. MDSS determined from patients who achieved 30 minutes of continuous exercise in ventilatory n=6, combined n=6, and cardiovascular n=5.129  Exercise intensity domains are typically demarcated by the workloads correspond to the AT, RCP, and CP.313\u2013315 Since AT, RCP, CP, CP30 and MDSS differ between the exercise limitation phenotypes, it is possible to differentiate specific \u201czones of training\u201d for each phenotype to further enhance exercise prescription in COPD. By incorporating the workloads and durations achieved during the three constant load trials used to construct the power-duration curves (also described in Figure 5.3), it is also possible to estimate the duration that each intensity domain may be tolerated for. A theoretically decision tree algorithm for aerobic exercise prescription in patients with COPD is presented in Figure 5.4. As the RCP could not be determined in any patients with a ventilatory phenotype, a 4-zone model was identified using the intensities corresponding to CP, CP30 and AT. However, as RCP was identified in the majority of patients with a cardiovascular or combined phenotype, a 4-zone model was determined using CP, CP30, AT and RCP. While the exact physiological adaptations derived from exercise performed in each of the different zones is unknown, general adaptations can be postulated based on the corresponding estimated intensities and durations. For example, exercise performed in the low to moderate intensity zones (~40-65%Wmax) that can be maintained for a longer duration (~30-45minutes) may confer cardiovascular adaptations such as enhanced plasma volume expansion and structural remodeling of the heart and blood vessels.316,317 Whereas exercise performed in the moderate to high intensity zones (\u226570%Wmax) that can be maintained for a shorter duration (~15-20 minutes, or 1-3 minute intervals) and is thus associated with a greater metabolic stimulus, may result in physiological adaptations related to changes in enzymatic adaptations and mitochondria biogenesis.318,319130  Figure 5.4. A theoretically decision tree algorithm for aerobic exercise prescription based on a four-zone model in COPD patients with different exercise limitation phenotypes. Incremental CPET performed on a cycle ergometer using a 10W\/min protocol. AT and RCP are determined from ventilatory equivalents and the modified Beaver plot. Intensity ranges represent the estimated relative workloads calculated from the interquartile ranges associated with each threshold within each phenotype. Exercise duration is predicted from CP30 and the average duration achieved during the constant load trials used to construct the power-duration relationships. Abbreviations: CPET, cardiopulmonary exercise test; VEpeak, peak minute ventilation; MVC, maximum voluntary ventilation estimated as FEV1*35; HRpeak, peak heart rate; HRmax, age-predicted maximal heart rate calculated as 220-age; CP30, critical power-30 representing the highest workload that can be maintained continuously for 30-minutes; AT, anaerobic threshold; RCP, respiratory compensation point. 131  5.4 Strengths and Limitations  The strengths of this dissertation lie in the novelty of concepts and practical application of the research findings. The criteria used to determine a ventilatory or cardiovascular limitation provides a relatively straightforward methodology to estimate exercise limitation without any additional complex physiological testing or analysis and are accepted by the current ATS\/ACCP exercise testing guidelines.62 As such, clinicians and practitioners may be more inclined to incorporate the phenotyping of exercise limitations in patients with COPD into clinical practice. Phenotyping patients according to exercise limitation appears to be an important element to designing and optimizing appropriate exercise prescriptions for patients with COPD. By using the same assessment (i.e., an increment CPET) that would have been previously performed to prescribe a generic exercise prescription (i.e., 60%Wmax for 20-minutes), patients can be classified phenotypically and CP30 can be prescribed with considerable confidence by accurately identifying AT. Then, intensity can be easily increased in small increments (e.g., 2-5 watts for the ventilatory phenotype, 5 watts for the combined phenotype, and 5-10 watts for the cardiovascular phenotype) to continually apply progressive overload as the patient adapts and progresses during training.  There are a number of limitations to the studies performed in this dissertation. 1) The use of different reference equations or more advanced techniques (i.e., the VECAP method,299 invasive measures of Q) to determine ventilatory capacity and maximal cardiac output may result in the slight re-classification of exercise limitation phenotype for a few patients. However, as we are proposing that exercise limitations lie on a continuum, this would result in patients moving either slightly right or left along the continuum. As such, the phenotypes would still represent important physiological transition points. 2) All exercise testing was performed on a cycle ergometer, and thus the physiological exercise responses may have differed had the incremental CPET and CLTs been performed on a treadmill 132 instead. As treadmill exercise is associated with a greater metabolic cost,320 VE would have likely been augmented potentially resulting in the attainment of a ventilatory limitation earlier during exercise in certain patients. However, as cycling is the most common mode of exercise used in patients with COPD to minimize risk of injury and to more easily quantify workload, it is a more relevant and applicable mode of exercise than the treadmill. 3) No verification trial was performed at CP in Study #2 (Chapter 3). This would have enhanced our understanding of the physiological responses associated with CP and what this threshold indicates in COPD. 4) Having a larger sample size in Study #3 (Chapter 4 )may have provided more clarity regarding the application of MDSS to patients with a cardiovascular or combined phenotype, particularly in those who report dyspnea as the primary limiting symptom at peak exercise in these phenotypes. Finally, 5) the contributory role of skeletal muscle (dys)function to the ventilatory response during exercise could not be directly evaluated as we did not perform direct measured of skeletal muscle function. However, the role of skeletal muscle function in the drive to breathe during exercise in COPD has been well established as per previous well-executed invasive and pharmaceutical studies.131,170   5.5 Future Directions  While we have identified three distinct phenotypes of exercise limitation in COPD, there are likely numerous sub-phenotypes that exist. In order to increase the external validity of these findings and to obtain a more comprehensive understanding of exercise limitations in the COPD population as a whole, evaluating the role of co-morbidities such as hypoxemia, obesity, and muscle atrophy on the phenotyping of exercise limitations is necessary. Additionally, investigating the effects of skeletal muscle function on phenotype\/ sub-phenotype classification is essential to understanding exercise limitations in this patient population. Manipulating the physiological variables that govern CP and assessing the effect on the power-duration relationship would provide further evidence to support that a 133 cardiovascular phenotype truly exists in COPD. For example, if CP is indeed limited by convective O2 delivery and the accumulation of fatigue-related metabolites, breathing a hyperoxic gas should result in an increase in CP by allowing oxidative metabolism to contribute to ATP production for a longer duration delaying the reliance on anaerobic glycolysis while still maintaining a ventilatory reserve at exercise limitation. Finally, whether an exercise training program prescribed according to the phenotype specific CP30 & MDSS reduces the number of non-responders to exercise training and thereby results in superior physiological benefits and clinical outcomes compared to the current standard of care, is an important question that warrants further investigation.   5.6 Overall Conclusion   Three phenotypes of exercise imitation were identified in patients with COPD that were associated with distinct pulmonary, cardiovascular and metabolic exercise responses during incremental CPET. These exercise limitation phenotypes were not solely dependent upon the severity of airflow obstruction. When exercise tolerance was investigated across phenotypes using the power-duration relationship, CP and CP30 (the highest workload that could be maintained for 30-minutes) occurred at a higher workload in patients with a cardiovascular contribution to exercise limitation compared to patients with a classic ventilatory phenotype. In the cardiovascular and combined phenotypes, exercise at CP30 was associated with a higher [BLa] however EELV remained lower compared to the ventilatory phenotype. Additionally, a novel index of sustainable exercise intensity, MDSS (maximal dyspnea steady-state), was utilized to predict the workload at which dyspnea would remain at a steady-state during continuous exercise. MDSS was found to be more applicable to patients with a ventilatory phenotype versus those with a cardiovascular or combined phenotype demonstrating that dyspnea is not be the primary limiting symptom in all patients with a cardiovascular contribution to exercise limitation. Collectively, the novel findings of this 134 dissertation demonstrate the importance of identifying phenotypes of exercise limitation in COPD as they may provide novel prognostic utility independent of disease severity, aid in the optimization of exercise prescription and can be used to develop innovative methods to individualize intensity and duration for aerobic training in patients.  Further work is necessary to delineate sub-phenotypes of exercise limitations in COPD to continue optimizing exercise prescription so that all patients may achieve significant benefits following exercise training and pulmonary rehabilitation.                     135 Bibliography  1.  Vogelmeier CF, Criner GJ, Martinez FJ, et al. Global strategy for the diagnosis, management, and prevention of chronic obstructive lung disease 2017 report. Am J Respir Crit Care Med 2017;195(5):557\u201382.  2.  Barnes PJ. Inflammatory mechanisms in patients with chronic obstructive pulmonary disease. J Allergy Clin Immunol 2016;138(1):16\u201327.  3.  Postma DS, Bush A, Van Den Berge M. Risk factors and early origins of chronic obstructive pulmonary disease. Lancet 2015;385(9971):899\u2013909.  4.  Naghavi M, Abajobir AA, Abbafati C, et al. Global, regional, and national age-sex specific mortality for 264 causes of death, 1980\u20132016: a systematic analysis for the Global Burden of Disease Study 2016. Lancet 2017;390:1151\u2013210.  5.  Leung C, Bourbeau J, Sin DD, et al. The prevalence of chronic obstructive pulmonary disease (COPD) and the heterogeneity of risk factors in the Canadian population: Results from the Canadian obstructive lung disease (COLD) study. Int J COPD 2021;16:305\u201320.  6.  Khakban A, Sin DD, FitzGerald JM, et al. The projected epidemic of chronic obstructive pulmonary disease hospitalizations over the next 15 years a population-based perspective. Am J Respir Crit Care Med. 2017;195(3):287\u201391.  7.  Khakban A, Sin DD, FitzGerald JM, et al. Ten-year trends in direct costs of COPD: A population-based study. Chest 2015;148(3):640\u20136.  8.  Burrows B, Hopkins J, Fletcher C, Heard B, Jones N, Wootliff J. The emphysematous and bronchial types of chronic airways obstruction. Lancet 1966;287:830\u20135.  9.  van der Vaart H, Postma DS, Timens W, et al. Acute effects of cigarette smoking on inflammation in healthy intermittent smokers. Respir Res 2005;6(22):[doi: 10.1186\/1465-9921-6-22].  10.  Lapperre TS, Postma DS, Gosman MME, et al. Relation between duration of smoking cessation and bronchial inflammation in COPD. Thorax 2006;61(2):115\u201321.  11.  Hogg JC. Pathophysiology of airflow limitation in chronic obstructive pulmonary disease. Lancet 2004;364:709\u201321.  12.  Coxson HO, Rogers RM, Whittall KP, et al. A quantification of the lung surface area in emphysema using computed tomography. Am J Respir Crit Care Med 1999;159:851\u20136.  13.  Retamales I, Elliott M, Meshi B, et al. Amplification of inflammation in emphysema and its association with laten adenoviral infection. Am J Respir Crit Care Med 2001;164:469\u201373.  14.  Kim WD, Eidelman DH, Izquierdo JL, Ghezzo H, Saetta MP, Cosio MG. Centrilobular and panlobular emphysema in smokers. Am Rev Respir Dis 1991;144:1385\u201390.  136 15.  Hogg J, Macklem P, Thurlbeck W. Site and nature of airway obstruction in chronic obstructive lung disease. N Engl J Med 1968;278(25):1355\u201360.  16.  McDonough JE, Yuan R, Suzuki M, et al. Small-airway obstruction and emphysema in chronic obstructive pulmonary disease. N Engl J Med 2011;365:1567\u201375.  17.  Mead J, Turner JM, Macklem PT, Little JB. Significance of the relationship between lung recoil and maximum expiratory flow. J Appl Physiol 1967;22:95\u2013108.  18.  Miravitlles M, Dirksen A, Ferrarotti I, et al. European Respiratory Society statement: Diagnosis and treatment of pulmonary disease in \u03b11-antitrypsin deficiency. Eur Respir J 2017;50(5):1700610 [https:\/\/doi.org\/10.1183\/13993003.00610-20.  19.  Stuart-Harris C, Crofton J, Gilson J, et al. Definition and classification of chronic bronchitis for clinical and epidemiological purposes. Lancet 1965;285:775\u20139.  20.  Mullen B, Wright JL, Wiggs BR, Pare PD, Hogg J. Reassessment of inflammation of airways in chronic bronchitis. BMJ 1985;291:1235\u20139.  21.  Kanner RE, Anthonisen NR, Connett J. Lower respiratory illnesses promote FEV1 decline in current smokers but not ex-smokers with mild chroni obstructive pulmonary disease. Am J Respir Crit Care Med 2001;164:358\u201364.  22.  Vestbo J, Prescott E, Lange P. Association of chronic mucus hypersecretion with FEV1 and chronic obstructive pulmonary disease morbidity. J Respir Crit Care Med 1996;153:1530\u20135.  23.  Miller MR, Quanjer PH, Swanney MP, Ruppel G, Enright PL. Interpreting lung function data using 80% predicted and fixed thresholds misclassifies more than 20% of patients. Chest 2011;139(1):52\u20139.  24.  Culver BH, Graham BL, Coates AL, et al. Recommendations for a standardized pulmonary function report: an official American Thoracic Society technical statement. Am J Respir Crit Care Med 2017;196(11):1463\u201372.  25.  Quanjer PH, Stanojevic S, Cole TJ, et al. Multi-ethnic reference values for spirometry for the 3-95-yr age range: The global lung function 2012 equations. Eur Respir J 2012;40(6):1324\u201343.  26.  Stanojevic S, Graham BL, Cooper BG, et al. Official ERS technical standards: Global Lung Function Initiative reference values for the carbon monoxide transfer factor for Caucasians. Eur Respir J 2017;50:1700010 [doi.org\/10.1183\/13993003.00010-20.  27.  Cosio M, Ghezzo H, Hogg J, et al. The relations between structural changes in small airways and pulmonary function tests. N Engl J Med 1977;298:1277\u201381.  28.  Corbin RP, Loveland M, Martin RR, Macklem PT. A four-year follow-up study of lung mechanics in smokers. Am Rev Respir Dis 1979;120:293\u2013304.  29.  Deesomchok A, Webb KA, Forkert L, et al. Lung hyperinflation and its reversibility in patients with airway obstruction of varying severity. COPD J Chronic Obstr Pulm Dis 137 2010;7(6):428\u201337.  30.  Morrison NJ, Abboud RT, Ramadan F, et al. Comparison of single breath carbon monoxide diffusing capacity and pressure-volume curves in detecting emphysema. Am Rev Respir Dis 1989;139:1179\u201387.  31.  Wagner PD, Dantzker DR, Dueck R, Clausen JL, West JB. Ventilation-perfusion inequality in chronic obstructive pulmonary disease. J Clin Invest 1977;59:203\u201316.  32.  Rodr\u00edguez-Roisin R, Drakulovic M, Rodr\u00edguez DA, Roca J, Barber\u00e0 JA, Wagner PD. Ventilation-perfusion imbalance and chronic obstructive pulmonary disease staging severity. J Appl Physiol 2009;106:1902\u20138.  33.  Van Eeden SF, Sin DD. Chronic obstructive pulmonary disease: A chronic systemic inflammatory disease. Respiration 2008;75(2):224\u201338.  34.  Anthonisen NR, Skeans MA, Wise RA, Manfreda J, Kanner RE, Connett JE. The effects of a smoking cessation intervention on 14.5-year mortality. Ann 2005;142:233\u20139.  35.  Sin DD, Wu L, Man P. The relationship between reduced lung function and cardiovascular mortality. Chest 2005;127:1952\u20139.  36.  Reardon JZ, Lareau SC, ZuWallack R. Functional status and quality of life in chronic obstructive pulmonary disease. Am J Med 2006;119(10A):S32\u20137.  37.  Mesquita R, Spina G, Pitta F, et al. Physical activity patterns and clusters in 1001 patients with COPD. Chron Respir Dis 2017;14(3):256\u201369.  38.  Polkey M, Moxhma J. Attacking the disease spiral in chronic obstructive pulmonary disease. Clin Med 2006;6(2):190\u20136.  39.  Anthonisen NR, Connett JE, Kiley JP, et al. Effects of smoking intervention and the use of an inhalde anticholinergic bronchodilator on the rate of decline in FEV1. JAMA 1994;272(19):1497\u2013505.  40.  O\u2019Donnell DE, Sciurba F, Celli B, et al. Effect of fluticasone propionate\/salmeterol on lung hyperinflation and exercise endurance in COPD. Chest 2006;130(3):647\u201356.  41.  Calverley PMA, Anderson JA, Celli B, et al. Salmeterol and fluticasone propionate and survival in chronic obstructive pulmonary disease. N Engl J Med 2007;356:775\u201389.  42.  Decramer M, Celli B, Kesten S, Lystig T, Mehra S, Tashkin DP. Effect of tiotropium on outcomes in patients with moderate chronic obstructive pulmonary disease (UPLIFT): a prespecified subgroup analysis of a randomised controlled trial. Lancet 2009;374:1171\u20138.  43.  Nici L, Aaron SD, Alexander PE, et al. Pharmacologic management of chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2020;201(9):E56\u201369.  44.  Spruit MA, Singh SJ, Garvey C, et al. An official American Thoracic Society\/European 138 Respiratory Society statement: key concepts and advances in pulmonary rehabilitation. Am J Respir Crit Care Med 2013;188(8):e13\u201364.  45.  Ries AL, Kaplan RM, Limberg TM, Prewitt LM. Effects of pulmonary rehabilitation on physiologic and psychosocial outcomes in patients with chronic obstructive pulmonary disease. Ann Intern Med 1995;122:823\u201332.  46.  McCarthy B, Casey D, Devane D, Murphy K, Murphy E, Lacasse Y. Pulmonary rehabilitation for chronic obstructive pulmonary disease (Review). Cochran Database Syst Rev 2015;2(CD003793):[doi: 10.1002\/14651858.CD003793.pub3].  47.  Casaburi R, Patessio A, Ioli F, Zanaboni S, Donner CF, Wasserman K. Reductions in exercise lactic acidosis and ventilation as a result of exercise training in patients with obstructive lung disease. Am Rev Respir Dis 1991;143(1):9\u201318.  48.  Porszasz J, Emtner M, Goto S, Somfay A, Whipp BJ, Casaburi R. Exercise training decreases ventilatory requirements and exercise-induced hyperinflation at submaximal intensities in patients with COPD. Chest 2005;128(4):2025\u201334.  49.  Maltais F, Leblanc P, Simard C, et al. Skeletal muscle adaptation to endurance training in patients with chronic obstructive pulmonary disease. Am J Respir Crit Care Med 1996;154:442\u20137.  50.  Troosters T, Gosselink R, Decramer M. Exercise training in COPD: How to distinguish responders from nonresponders. J Cardiopulm Rehabil Prev 2001;21(1):10\u20137.  51.  Spruit M, Augustin I, Vanfleteren L, et al. Differential response to pulmonary rehabilitation in COPD: multidimensional profiling. Eur Respir J 2015;46(6):1538\u201340.  52.  De Torres JP, Pinto-Plata V, Ingenito E, et al. Power of outcome measurements to detect clinically significant changes in pulmonary rehabilitation of patients with COPD. Chest 2002;121(4):1092\u20138.  53.  Garrod R, Marshall J, Barley E, Jones PW. Predictors of success and failure in pulmonary rehabilitation. Eur Respir J 2006;27(4):788\u201394.  54.  Vanfleteren MJ, Koopman M, Spruit MA, et al. Effectiveness of pulmonary rehabilitation in patients with chronic obstructive pulmonary disease with different degrees of static lung hyperinflation. Arch Phys Med Rehabil 2018;99(11):2279\u201386.  55.  Beauchamp MK, Janaudis-Ferreira T, Goldstein RS, Brooks D. Optimal duration of pulmonary rehabilitation for individuals with chronic obstructive pulmonary disease - a systematic review. Chron Respir Dis 2011;8(2):129\u201340.  56.  Huynh VC, Fuhr DP, Byers BW, Selzler AM, Moore LE, Stickland MK. Influence of Disease Severity and Exercise Limitation on Exercise Training Intensity and Load and Health Benefits From Pulmonary Rehabilitation in Patients With COPD: an exploratory study. J Cardiopulm Rehabil Prev 2018;38(5):320\u20136.  57.  Ward TJC, Plumptre CD, Dolmage TE, et al. Change in V\u0307O2peak in response to aerobic exercise training and the relationship with exercise prescription in people with 139 COPD: a systematic review and meta-analysis. Chest 2020;158:131\u201344.  58.  Plankeel JF, McMullen B, MacIntyre NR. Exercise Outcomes After Pulmonary Rehabilitation Depend on the Initial Mechanism of Exercise Limitation Among Non-Oxygen-Dependent COPD Patients. Chest [Internet] 2005;127:110\u20136. Available from: www.chestjournal.org 59.  Eves ND, Sandmeyer LC, Wong EY, et al. Helium-hyperoxia: A novel intervention to improve the benefits of pulmonary rehabilitation for patients with COPD. Chest 2009;135(3):609\u201318.  60.  Perry S, Koelwyn G, Wong L, Davidson W, Eves N. Helium-hyperoxia: Alleviating respiratory limitation to improve pulmonary rehabilitation. Int J Respir Care 2011;188(8):e13\u201364.  61.  Howley E, Bassett D, Welch H. Criteria for maximal oxygen uptake: Review and commentary. Med Sci Sport Exerc 1995;27(9):1292\u2013301.  62.  American Thoracic Society & American College of Chest Physicians. ATS\/ACCP Statement on cardiopulmonary exercise testing. Am J Respir Crit Care Med 2003;167(2):211\u201377.  63.  Astrand P-O, Cuddy TE, Saltin B, Stenberg J. Cardiac output during submaximal and maximal work. J Appl Physiol 1964;19(2):268\u201374.  64.  Higginbotham MB, Morris KG, Williams RS, Mchale PA, Coleman RE, Cobb FR. Regulation of Stroke Volume during Submaximal and Maximal Upright Exercise in Normal Man. Circ Res 1986;58(2):281\u201391.  65.  Grimby G, Nilsson NJ, Saltin B. Cardiac output during submaximal and maximal exercise in active middle-aged athletes [Internet]. 1966. Available from: www.physiology.org\/journal\/jappl 66.  Plotnick GD, Becker LC, Fisher ML, et al. Use of the Frank-Starling mechanism during submaximal versus maximal upright exercise. Am J Physiol - Hear Circ Physiol 1986;251(20):H1101\u20135.  67.  Steingart RM, Wexler J, Slagle S, Scheuer J. Radionuclide ventriculographic responses to graded supine and upright exercise: critical role of the Frank-Starling mechanism at submaximal exercise. Am J Cardiol 1984;1671\u20137.  68.  Secher NH, Clausen JP, Klausen K, Noer I, Trap-Jensen J. Central and regional circulatory effects of adding arm exercise to leg exercise. Acta Physiol Scand 1977;100:288\u201397.  69.  Richter EA, Kiens B, Hargreaves M, Kjaer M. Effect of arm-cranking on leg blood flow and noradrenaline spillover during leg exercise in man. Acta Physiol Scand 1992;144:9\u201314.  70.  Savard GK, Richter EA, Strange S, Kiens B, Christensen NJ, Saltin B. Norepinephrine spillover from skeletal muscle during exercise in humans: role of muscle mass. Am J 140 Physiol - Hear Circ Physiol 1989;257(6):H1812\u20138.  71.  Andersen P, Adams RP, Sjogaard G, Thorboe A, Saltin B. Dynamic knee extension as model for study of isolated exercising muscle in humans. J Appl Physiol 1985;59(5):1647\u201353.  72.  Bebout DE, Hogan MC, Hempleman SC, Wagner PD. Effects of training and immobilization on VO2 and DO2 in dog gastrocnemius muscle in situ. J Appl Physiol 1993;74:1697\u2013703.  73.  Hogan MC, Bebout DE, Wagner PD. Effect of hemoglobin concentration on maximal 02 uptake in canine gastrocnemius muscle in situ. J Appl Physiol 1991;70:1105\u201312.  74.  Groebe K, Thews G. Calculated intra-and extracellular PO2 gradients in heavily working red muscle. Hear Circ Physiol 1990;259:H84\u201392.  75.  Gayeski TEJ, Honig CR. Intracellular PO2 in long axis of individual fibers in working dog gracilis muscle. Hear Circ Physiol 1988;254:H1179\u201386.  76.  Hogan MC, Bebout DE, Wagner PD. Effect of increased Hb-02 affinity on VO2 max at constant 02 delivery in dog muscle in situ. J Appl Physiol 1991;70:2656\u201362.  77.  Wagner PD. Determinants of maximal oxygen transport and utilization. Annu Rev Physiol 1996;58:21\u201350.  78.  Gledhill N, Cox D, Jamnik R. Endurance athlete\u2019s stroke volume does not plateau: major advantage is diastolic function. Med Sci Sports Exerc 1994;26:1116\u201321.  79.  Warburton D, Haykowsky M, Quinney A, Blackmore D, Teo K, Humen D. Myocardial responses to incremental exercise in endurance trained athletes: influence of heart rate, contractility and the Frank-Starling effect. Exp Physiol 2002;87:613\u201322.  80.  Ogawa T, Spina RJ, Martin W, et al. Effects of aging, sex, and physical training on cardiovascular responses to exercise. Circulation 1992;86:494\u2013503.  81.  Wiebe C, Gledhill N, Jamnik V, Ferguson S. Exercise cardiac function in young through elderly endurance trained women. Med Sci Sports Exerc 1999;31:684\u201391.  82.  Jensen-Urstad M, Bouvier F, Nejat M, Saltin B, Brodin L. Left ventricular function in endurance runners during exercise. Acta Physiol Scand 1998;164:167\u201372.  83.  Gandevia B, Hugh-Jones P. Terminology for measurements of ventilatory capacity: A report to the thoracic society. Thorax 1957;12(4):290\u20133.  84.  Whipp BJ, Wasserman K. Alveolar-arterial gas tension differences during graded exercise. J Appl Physiol 1969;27:361\u20135.  85.  Gallagher CG, Brown E, Younes M. Breathing pattern during maximal exercise and during submaximal exercise with hypercapnia. J Appl Physiol 1987;63(1):238\u201344.  86.  McParland C, Mink J, Gallagher CG. Respiratory adaptations to dead space loading during maximal incremental exercise. J Appl Physiol 1991;70(1):55\u201362.  141 87.  O\u2019Donnell D, Hong HH, Webb KA. Respiratory sensation during chest wall restriction and dead space loading in exercising men. J Appl Physiol 2000;88:1859\u201369.  88.  Jensen D, O\u2019Donnell DE, Li R, Luo YM. Effects of dead space loading on neuro-muscular and neuro-ventilatory coupling of the respiratory system during exercise in healthy adults: implications for dyspnea and exercise tolerance. Respir Physiol Neurobiol 2011;179(2\u20133):219\u201326.  89.  Henke KG, Sharratt M, Pegelow D, Dempsey JA. Regulation of end-expiratory lung volume during exercise. J Appl Physiol 1988;64(1):135\u201346.  90.  Younes M, Kivinen G. Respiratory mechanics and breathing pattern during and following maximal exercise. J Appl Physiol 1984;57(6):1773\u201382.  91.  Fleg JL, Schulman S, O\u2019connor F, et al. Effects of acute B-adrenergic receptor blockade on age-assocaited changes in cardiovascular performance during dynamic exercise. Circulation 1994;90:2333\u201341.  92.  Hagberg JM, Allen WK, Seals DR, Hurley BF, Ehsani A, Holloszy J 0. A hemodynamic comparison of young and older endurance athletes during exercise. J Appl Physiol 1985;58:2041\u20136.  93.  Rivera AM, Pels III AE, Sady SP, Sady MA, Cullinane EM, Thompson PD. Physioloigcal factors associated with the lower maximal oxygen consumption of master runners. J Appl Physiol 1989;66:949\u201354.  94.  Fleg JL, Schulman SP, O\u2019Connor F, et al. Cardiovascular responses to exhaustive upright cycle exercise in highly trained older men. J Appl Physiol 1994;77:1500\u20136.  95.  Roman MA, Rossiter HB, Casaburi R. Exercise, ageing and the lung. Eur. Respir. J. 2016;48(5):1471\u201386.  96.  Knudson RJ, Lebowitz M, Holberg CJ, Burrows B. Changes in the normal maximal expiratory flow-volume curve with growth and aging. Am Rev Respir Dis 1983;127:725\u201334.  97.  DeLorey DS, Babb TG. Progressive Mechanical Ventilatory Constraints with Aging. Am J Respir Crit Care Med 1999;160:169\u201377.  98.  Johnson BD, Reddan WG, Pegelow DF, Seow KC, Dempsey JA. Flow limitation and regulation of functional residual capacity during exercise in a physically active aging population. Am Rev Respir Dis 1991;143:960\u20137.  99.  Johnson B, Badr M, Dempsey J. Impact of the aging pulmonary system on the response to exercise. Clin Chest Med 1994;15:229\u201346.  100.  Georges R, Saumon G, Loiseau A. The relationship of age to pulmonary membrane conductance and capillary blood volume. Am Rev Respir Dis 1978;117:1069\u201378.  101.  Frontera WR, Suh D, Krivickas LS, Hughes VA, Goldstein R, Roubenoff R. Skeletal muscle fiber quality in older men and women. Am J Physiol Cell Physiol  142 2000;279:C611\u20138.  102.  Frontera WR, Hughes VA, Fielding RA, Fiatarone MA, Evans WJ, Roubenoff R. Aging of skeletal muscle: a 12-yr longitudinal study. J Appl Physiol 2000;88:1321\u20136.  103.  Betik AC, Thomas MM, Wright KJ, Riel CD, Hepple RT. Exercise training from late middle age until senescence does not attenuate the declines in skeletal muscle aerobic function. Am J Physiol Regul Integr Comp Physiol 2009;297:744\u201355.  104.  Santanasto AJ, Glynn NW, Jubrias SA, et al. Skeletal muscle mitochondrial function and fatiguability in older adults. J Gerontol A Biol Sci Med Sci 2015;70(11):1379\u201385.  105.  Hawkins SA, Wiswell RA. Rate and mechanism of maximal oxygen consumption decline with aging implications for exercise training. Sport Med 2003;33(12):877\u201388.  106.  Molgat-Seon Y, Dominelli PB, Ramsook AH, et al. The effects of age and sex on mechanical ventilatory constraint and dyspnea during exercise in healthy humans. J Appl Physiol 2018;124:1092\u2013106.  107.  Ofir D, Laveneziana P, Webb KA, Lam Y-M, O\u2019Donnell D. Sex differences in the perceived intensity of breathlessness during exercise with advancing age. J Appl Physiol 2008;104:1583\u201393.  108.  Wilkie SS, Guenette JA, Dominelli PB, Sheel AW. Effects of an aging pulmonary system on expiratory flow limitation and dyspnoea during exercise in healthy women. Eur J Appl Physiol 2012;112(6):2195\u2013204.  109.  McClaran SR, Babcock MA, Pegelow DF, Reddan WG, Dempsey JA. Longitudinal effects of aging on lung function at rest and exercise in healthy active fit elderly adults. J Appl Physiol 1995;78(5):1957\u201368.  110.  O\u2019Donnell DE, Guenette JA, Maltais F, Webb KA. Decline of resting inspiratory capacity in COPD: The impact on breathing pattern, dyspnea, and ventilatory capacity during exercise. Chest 2012;141(3):753\u201362.  111.  O\u2019Donnell DE, Lam M, Webb KA. Spirometric correlates of improvement in exercise performance after anticholinergic therapy in chronic obstructive pulmonary disease. Am J Respir Crit Care Med 1999;160:542\u20139.  112.  O\u2019Donnell DE, Revill SM, Webb KA. Dynamic hyperinflation and exercise intolerance in chronic obstrutive pulmonary disease. Am J Respir Crit Care Med 2001;164:770\u20137.  113.  Ofir D, Laveneziana P, Webb KA, Lam YM, O\u2019Donnell DE. Mechanisms of dyspnea during cycle exercise in symptomatic patients with GOLD stage I chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2008;177(6):622\u20139.  114.  Maltais F, LeBlanc P, Whittom F, et al. Oxidative enzyme activities of the vastus lateralis muscle and the functional status in patients with COPD. Thorax 2000;55(10):848\u201353.  115.  Maltais F, Jobin J, Sullivan MJ, et al. Metabolic and hemodynamic responses of lower 143 limb during exercise in patients with COPD. J Appl Physiol 1998;84:1573\u201380.  116.  Killian KJ, Leblanc P, Martin DH, Summers E, Jones NL, Campbell EJM. Exercise capacity and ventilatory, circulatory, and symptom limitation in patients with chronic airflow limitation. Am Rev Respir Dis 1992;146:935\u201340.  117.  Faisal A, Alghamdi BJ, Ciavaglia CE, et al. Common mechanisms of dyspnea in chronic interstitial and obstructive lung disorders. Am J Respir Crit Care Med 2016;193(3):299\u2013309.  118.  Elbehairy AF, Ciavaglia CE, Webb KA, et al. Pulmonary gas exchange abnormalities in mild chronic obstructive pulmonary disease: Implications for dyspnea and exercise intolerance. Am J Respir Crit Care Med 2015;191(12):1384\u201394.  119.  Chin RC, Guenette JA, Cheng S, et al. Does the respiratory system limit exercise in mild chronic obstructive pulmonary disease? Am J Respir Crit Care Med 2013;187(12):1315\u201323.  120.  Guenette JA, Chin RC, Cheng S, et al. Mechanisms of exercise intolerance in Global Initiative for Chronic Obstructive Lung Disease grade 1 COPD. Eur Respir J 2014;44(5):1177\u201387.  121.  Neder JA, Berton DC, Marillier M, Bernard AC, O\u2019Donnell DE. Inspiratory Constraints and Ventilatory Inefficiency Are Superior to Breathing Reserve in the Assessment of Exertional Dyspnea in COPD. COPD J Chronic Obstr Pulm Dis 2019;16(2):174\u201381.  122.  Thirapatarapong W, Armstrong HF, Thomashow BM, Bartels MN. Differences in gas exchange between severities of chronic obstructive pulmonary disease. Respir Physiol Neurobiol 2013;186(1):81\u20136.  123.  Nery LE, Wasserman K, French W, Oren A, Davis JA. Contrasting cardiovascular and respiratory responses to exercise in mitral valve and chronic obstructive pulmonary diseases. Chest 1983;83(3):446\u201353.  124.  Barbera JA, Roca J, Ramirez J, Wagner PD, Ussetti P, Rodriguez-Roisin R. Gas exchange during exercise in mild chronic obstructive pulmonary disease. Am Rev Respir Dis 1991;144:520\u20135.  125.  Dantzker D, D\u2019Alonzo G. The effect of exercise on pulmonary gas exchange in patients with severe chronic obstructive pulmonary disease. Am Rev Respir Dis 1986;134:1135\u20139.  126.  Whittom F, Jobin J, Simard P-M, et al. Histochemical and morphological characteristics of the vastus lateralis muscle in patients with chronic obstructive pulmonary disease. Med Sci Sports Exerc 1998;30(10):1467\u201374.  127.  Saey D, Michaud A, Couillard A, et al. Contractile fatigue, muscle morphometry, and blood lactate in chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2005;171(10):1109\u201315.  128.  Oelberg DA, Medoff BD, Markowitz DH, Pappagianopoulos PP, Ginns LC, Systrom 144 DM. Systemic oxygen extraction during incremental exercise in patients with severe chronic obstructive pulmonary disease. Eur J Appl Physiol 1988;78:201\u20137.  129.  Simon M, Leblanc P, Jobin J, Desmeules M, Sullivan MJ, Maltais F. Limitation of lower limb VO2 during cycling exercise in COPD patients. J Appl Physiol 2001;90:1013\u20139.  130.  Maltais F, Simard A-A, Simard C, Jobin J, Desgagnes P, LeBlanc P. Oxidative capacity of the skeletal muscle and lactic acid kinetics during exercise in normal subjects and in patients with COPD. Am J Respir Crit Care Med 1996;153:288\u201393.  131.  Gagnon P, Bussi\u00e8res JS, Ribeiro F, et al. Influences of spinal anesthesia on exercise tolerance in patients with chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2012;186(7):606\u201315.  132.  Sue DY, Wasserman K, Moricca RB, Casaburi R. Metabolic acidosis during exercise in patients with chronic obstructive pulmonary disease. Chest 1988;94(5):931\u20138.  133.  Neder JA, Arbex FF, Alencar MCN, et al. Exercise ventilatory inefficiency in mild to end-stage COPD. Eur Respir J 2015;45(2):377\u201387.  134.  Wasserman K, Whipp B, Koyal S, Beaver W. Anaerobic threshold and respiratory gas exchange during exercise. J Appl Physiol 1973;35(2):236\u201343.  135.  Dempsey JA, Smith CA. Pathophysiology of human ventilatory control. Eur Respir J 2014;44(2):495\u2013512.  136.  Amann M, Blain GM, Proctor LT, Sebranek JJ, Pegelow DF, Dempsey JA. Group III and IV muscle afferents contribute to ventilatory and cardiovascular response to rhythmic exercise in humans. J Appl Physiol 2010;109(4):966\u201376.  137.  Yan S, Kaminski D, Sliwinski P. Reliability of inspiratory capacity for estimating end-expiratory lung volume changes during exercise in patients with chronic obstructive pulmonary disease. Am J Respir Crit Care Med 1997;156:55\u20139.  138.  Laveneziana P, Webb KA, Ora J, Wadell K, O\u2019Donnell DE. Evolution of dyspnea during exercise in chronic obstructive pulmonary disease impact of critical volume constraints. Am J Respir Crit Care Med 2011;184(12):1367\u201373.  139.  O\u2019Donnell DE, Hamilton AL, Webb KA. Sensory-mechanical relationships during high-intensity, constant-work-rate exercise in COPD. J Appl Physiol 2006;101(4):1025\u201335.  140.  O\u2019Donnell DE, Elbehairy AF, Webb KA, Neder JA. The link between reduced inspiratory capacity and exercise intolerance in chronic obstructive pulmonary disease. Ann Am Thorac Soc 2017;14(July):S30\u20139.  141.  Sliwinski P, Kaminski D, Zielinski J, Yan S. Partitioning of the elastic work of inspiration in patients with COPD during exercise. Eur Respir J 1998;11(2):416\u201321.  142.  Sinderby C, Spahija J, Breck J, et al. Diaphragm activation during exercise in chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2001;163:1637\u201341.  145 143.  O\u2019Donnell DE, Bertley JC, Chau LKL, Webb KA. Qualitative aspects of exertional breathlessness in chronic airflow limitation. Am J Respir Crit Care Med 1997;155:109\u201315.  144.  Martinez FJ, Couser JI, Celli BR. Factors influencing ventilatory muscle recruitment in patients with chronic airdlow obstruction. Am Rev Respir Dis 1990;142:275\u201382.  145.  Yan S, Kaminski D, Sliwinski P. Inspiratory muscle mechanics of patients with chronic obstructive pulmonary disease during incremental exercise. Am J Respir Crit Care Med 1997;156:807\u201313.  146.  Eves ND, Petersen SR, Haykowsky MJ, Wong EY, Jones RL. Helium-hyperoxia, exercise, and respiratory mechanics in chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2006;174(7):763\u201371.  147.  Levison H, Cherniack R. Ventilatory cost of exercise in chronic obstructive pulmonary disease. J Appl Physiol 1968;25:21\u20137.  148.  O\u2019Donnell DE, Laveneziana P, Ora J, Webb KA, Lam YM, Ofir D. Evaluation of acute bronchodilator reversibility in patients with symptoms of GOLD stage l COPD. Thorax 2009;64(3):216\u201323.  149.  Potter WA, Olafsson S, Hyatt RE. Ventilatory mechanics and expiratory flow limitation during exercise in patients with obstructive lung disease. J Clin Invest 1971;50:910\u20139.  150.  Luo YM, Li RF, Jolley C, et al. Neural respiratory drive in patients with COPD during exercise tests. Respiration 2011;81(4):294\u2013301.  151.  Jolley C, Luo Y, Steier J, Rafferty G, Polkey M, Moxham J. Neural respiratory drive and breathlessness in COPD. Eur. Respir. J. 2015;45(2):301\u20134.  152.  Jolley CJ, Luo YM, Steier J, et al. Neural respiratory drive in healthy subjects and in COPD. Eur Respir J 2009;33(2):289\u201397.  153.  Swallow EB, Reyes D, Hopkinson NS, et al. Quadriceps strength predicts mortality in patients with moderate to severe chronic obstructive pulmonary disease. Thorax 2007;62(2):115\u201320.  154.  Coronell C, Orozco-Levi M, M\u00e9ndez R, Ram\u00edrez-Sarmiento A, G\u00e1ldiz JB, Gea J. Relevance of assessing quadriceps endurance in patients with COPD. Eur Respir J 2004;24(1):129\u201336.  155.  Decramer M, Gosselink R, Troosters T, Verschueren M, Evers G. Muscle weakness is related to utilization of healty care resources in COPD patients. Eur Respir J 1997;10(2):417\u201323.  156.  Willlams TJ, Patterson GA, Mcclean PA, Zamel N, Maurer JR. Maximal exercise testing in single and double lung transplant recipients. Am Rev Respir Dis 1992;145:101\u20135.  157.  Saey D, Debigar\u00e9 R, LeBlanc P, et al. Contractile leg fatigue after cycle exercise: a factor limiting exercise in patients with chronic obstructive pulmonary disease. Am J 146 Respir Crit Care Med 2003;168(4):425\u201330.  158.  Marklund S, Bui KL, Nyberg A. Measuring and monitoring skeletal muscle function in COPD: Current perspectives. Int J COPD 2019;14:1825\u201338.  159.  Gosker HR, Zeegers MP, Wouters EFM, Schols AMWJ. Muscle fibre type shifting in the vastus lateralis of patients with COPD is associated with disease severity: A systematic review and meta-analysis. Thorax 2007;62(11):944\u20139.  160.  Eliason G, Abdel-Halim SM, Piehl-Aulin K, Kadi F. Alterations in the muscle-to-capillary interface in patients with different degrees of chronic obstructive pulmonary disease. Respir Res 2010;11:97 [doi: 10.1186\/1465-9921-11-97].  161.  Gifford JR, Trinity JD, Kwon O-S, et al. Altered skeletal muscle mitochondrial phenotype in COPD: disease vs. disuse. J Appl Physiol 2018;124:1045\u201353.  162.  Puente-Maestu L, P\u00e9rez-Parra J, Godoy R, et al. Abnormal mitochondrial function in locomotor and respiratory muscles of COPD patients. Eur Respir J 2009;33(5):1045\u201352.  163.  Bernard S, Blanc PLE, Whittom F, et al. Peripheral muscle weakness in patients with chronic obstructive pulmonary disease. Am J Respir Crit Care Med 1998;158:629\u201334.  164.  Man WD-C, Soliman G, Nikoletou D, Harris ML, Rafferty GF, Mustfa N. Non-volitional assessment of skeletal muscle strength in patients with chronic obstructive pulmonary disease. Thorax 2003;58:665\u20139.  165.  Seymour JM, Ward K, Sidhu PS, et al. Ultrasound measurement of rectus femoris cross-sectional area and the relationship with quadriceps strength in COPD. Thorax 2009;64(5):418\u201323.  166.  Debigar\u00e9 R, C\u00f4t\u00e9 CH, Hould FS, LeBlanc P, Maltais F. In vitro and in vivo contractile properties of the vastus lateralis muscle in males with COPD. Eur Respir J 2003;21(2):273\u20138.  167.  Seymour JM, Spruit MA, Hopkinson NS, et al. The prevalence of quadriceps weakness in COPD and the relationship with disease severity. Eur Respir J 2010;36(1):81\u20138.  168.  Jakobsson P, Jorfeldt L, Henriksson J. Metabolic enzyme activity in the quadriceps femoris muscle in patients with severe chronic obstructive pulmonary disease. J Respir Crit Care Med 1995;151:374\u20137.  169.  Green HJ, Bombardier E, Burnett M, et al. Organization of metabolic pathways in vastus lateralis of patients with chronic obstructive pulmonary disease. Am J Physiol Regul Integr Comp Physiol 2008;295:935\u201341.  170.  Saey D, Lemire BB, Gagnon P, et al. Quadriceps metabolism during constant workrate cycling exercise in chronic obstructive pulmonary disease. J Appl Physiol 2011;110(1):116\u201324.  171.  Puente-Maestu L, Tejedor A, L\u00e1zaro A, et al. Site of mitochondrial reactive oxygen 147 species production in skeletal muscle of chronic obstructive pulmonary disease and its relationship with exercise oxidative stress. Am J Respir Cell Mol Biol 2012;47(3):358\u201362.  172.  Rabinovich RA, Bastos R, Ardite E, et al. Mitochondrial dysfunction in COPD patients with low body mass index. Eur Respir J 2007;29(4):643\u201350.  173.  Picard M, Godin R, Sinnreich M, et al. The mitochondrial phenotype of peripheral muscle in chronic obstructive pulmonary disease: Disuse or dysfunction? Am J Respir Crit Care Med 2008;178(10):1040\u20137.  174.  Br\u00f8nstad E, Rognmo O, Tjonna AE, et al. High-intensity knee extensor training restores skeletal muscle function in COPD patients. Eur Respir J 2012;40(5):1130\u20136.  175.  Decramer M, Bock V De, Dom R. Functional and histological picture of steroid-induced myopathy in chronic obstructive pulmonary disease. Am J Respir Crit Care Med 1996;153:1958\u201364.  176.  Costes F, Gosker H, Feasson L, et al. Impaired exercise training-induced muscle fiber hypertrophy and Akt\/mTOR pathway activation in hypoxemic patients with COPD. J Appl Physiol 2015;118(8):1040\u20139.  177.  Richardson RS, Sheldon J, Poole DC, Hopkins SR, Ries AL, Wagner PD. Evidence of skeletal muscle metabolic reserve during whole body exercise in patients with chronic obstructive pulmonary disease. Am J Respir Crit Care Med 1999;159:881\u20135.  178.  Oelberg DA, Kacmarek RM, Pappagianopoulos PP, Ginns LC, Systrom DM. Ventilatory and Cardiovascular Responses to Inspired He-O 2 during Exercise in Chronic Obstructive Pulmonary Disease. Am J Respir Crit Care Med 1998;158:1876\u201382.  179.  Broxterman RM, Hoff J, Wagner PD, Richardson RS. Determinants of the diminished exercise capacity in patients with chronic obstructive pulmonary disease: looking beyond the lungs. J Physiol 2020;598(3):599\u2013610.  180.  Light R, Mintz H, Linden G, Brown S. Hemodynamics of patients with severe chronic obstructive pulmonary disease during progressive upright exercise. Am Rev Respir Dis 1984;130:391\u20135.  181.  Morrison D, Adcock K, Michael Collins C, Goldman S, Caldwell J, Schwarz MI. Right ventricular dysfunction and the exercise limitation of chronic obstructive pulmonary disease. J Am Coll Cardiol 1987;9(6):1219\u201329.  182.  Slutsky R, Hooper W, Ackerman W, et al. Evaluation of left ventricular function in chronic obstructive pulmonary disease by exercise gated equilibrium radionuclide angiography. Am Hear J 1981;101(4):414\u201320.  183.  Stewart RI, Lewis CM. Cardiac output during exercise in patients with COPD. Chest 1986;89(2):199\u2013205.  184.  Babb TG, Viggiano R, Hurley B, Staats B, Rodarte JR. Effect of mild-to-moderate airflow limitation on exercise capacity. J Appl Physiol 1991;70(1):223\u201330.  148 185.  Matthews JI, Bush BA, Ewald FW. Exercise responses during incremental and high intensity and low intensity steady state exercise in patients with obstructive lung disease and normal control subjects. Chest 1989;96(1):11\u20137.  186.  Berger H, Matthay R, Davies R, Zaret B, Gottschalk A. Comparison of exercise right ventricular performance in chronic obstructive pulmonary disease and coronary artery disease: noninvasive assessment by quantitative radionuclide angiocardiography. Invest Radiol 1979;14(5):342\u201353.  187.  Matthay RA, Berger HJ, Davies RA, et al. Right and left ventricular exercise performance in chronic obstructive pulmonary disease: Radionuclide assessment. Ann Intern Med 1980;93:234\u20139.  188.  Vassaux C, Torre-Bouscoulet L, Zeineldine S, et al. Effects of hyperinflation on the oxygen pulse as a marker of cardiac performance in COPD. Eur Respir J 2008;32(5):1275\u201382.  189.  Tzani P, Aiello M, Elia D, et al. Dynamic hyperinflation is associated with a poor cardiovascular response to exercise in COPD patients. Respir Res 2011;12:150[doi: 10.1186\/1465-9921-12-150].  190.  Ramponi S, Tzani P, Aiello M, Marangio E, Clini E, Chetta A. Pulmonary rehabilitation improves cardiovascular response to exercise in COPD. Respiration 2013;86(1):17\u201324.  191.  Montes De Oca M, Rassulo J, Bartolome CR. Respiratory muscle and cardiopulmonary function during exercise in very severe COPD. Am J Respir Crit Care Med 1996;154:1284\u20139.  192.  Schrijen F V, Henriquez A, Carton D, Delormet N, Butler J. Pulmonary vascular resistance rises with lung volume on exercise in obstructed airflow disease. Clin Physiol 1989;9:143\u201350.  193.  Butler J, Schrijen F, Henriquez A, Polu J-M, Albert RK. Cause of the raised wedge pressure on exercise in chronic obstructive pulmonary disease. Am Rev Respir Dis 1988;138:350\u20134.  194.  Magee F, Wright JL, Hogg JC. Pulmonary vascular structure and function in chronic obstructive pulmonary disease. Thorax 1988;43:183\u20139.  195.  Agusti AGN, Barbera JA, Roca J, Wagner PD, Guitart R, Rodriguez-Roisin R. Hypoxic pulmonary vasoconstriction and gas exchange during exercise in chronic obstructive pulmonary disease. Chest 1990;97(2):268\u201375.  196.  Holverda S, Rietema H, Westerhof N, et al. Stroke volume increase to exercise in chronic obstructive pulmonary disease is limited by increased pulmonary artery pressure. Heart 2009;95(2):137\u201341.  197.  Boerrigter BG, Bogaard HJ, Trip P, et al. Ventilatory and cardiocirculatory exercise profiles in COPD: the role of pulmonary hypertension. Chest 2012;142(5):1166\u201374.  149 198.  Mahler D, Brent B, Loke J, Zaret B, Matthay R. Right ventricular performance and central circulatory hemodynamics during upright exercise in patients with chronic obstructive pulmonary disease. Am Rev Respir Dis 1984;130:722\u20139.  199.  Hilde JM, Skj\u00f8rten I, Hansteen V, et al. Haemodynamic responses to exercise in patients with COPD. Eur Respir J 2013;41(5):1031\u201341.  200.  Mitchell JR, Whitelaw WA, Sas R, Smith ER, Tyberg J V., Belenkie I. RV filling modulates LV function by direct ventricular interaction during mechanical ventilation. Am J Physiol - Hear Circ Physiol 2005;289:H549\u201357.  201.  Rozenbaum Z, Ben-Gal Y, Kapusta L, et al. Combined echocardiographic and cardiopulmonary exercise to assess determinants of exercise limitation in chronic obstructive pulmonary disease. J Am Soc Echocardiogr 2021;34(2):146\u201355.  202.  John M, Lange A, Hoernig S, Witt C, Anker SD. Prevalence of anemia in chronic obstructive pulmonary disease: Comparison to other chronic diseases. Int J Cardiol 2006;111(3):365\u201370.  203.  Patel MS, Mckie E, Steiner MC, Pascoe SJ, Polkey MI. Anaemia and iron dysregulation: Untapped therapeutic targets in chronic lung disease? BMJ Open Respir Res 2019;6(1):e000454 [doi:10.1136\/bmjresp-2019-000454].  204.  Boutou AK, Stanopoulos I, Pitsiou GG, et al. Anemia of chronic disease in chronic obstructive pulmonary disease: A case-control study of cardiopulmonary exercise responses. Respiration 2011;82(3):237\u201345.  205.  Ferrari M, Manea L, Anton K, et al. Anemia and hemoglobin serum levels are associated with exercise capacity and quality of life in chronic obstructive pulmonary disease. BMC Pulm Med 2015;15(1).  206.  Chiappa GR, Queiroga F, Meda E, et al. Heliox improves oxygen delivery and utilization during dynamic exercise in patients with chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2009;179(11):1004\u201310.  207.  Louvaris Z, Zakynthinos S, Aliverti A, et al. Heliox increases quadriceps muscle oxygen delivery during exercise in COPD patients with and without dynamic hyperinflation. J Appl Physiol 2012;113(7):1012\u201323.  208.  Amann M, Regan MS, Kobitary M, et al. Impact of pulmonary system limitations on locomotor muscle fatigue in patients with COPD. Am J Physiol Regul Integr Comp Physiol 2010;299:314\u201324.  209.  Oga T, Nishimura K, Tsukino M, Sato S, Hajiro T. Analysis of the factors related to mortality in chronic obstructive pulmonary disease: role of exercise capacity and health status. Am J Respir Crit Care Med 2003;167(4):544\u20139.  210.  Myers J, Prakash M, Froelicher V, Partington S, Atwood JE. Exercise capacity and mortality among men referred for exercise testing. N Engl J Med 2002;346(11):793\u2013801.  150 211.  Neder JA, Alharbi A, Berton DC, et al. Exercise ventilatory inefficiency adds to lung function in predicting mortality in COPD. COPD J Chronic Obstr Pulm Dis 2016;13(4):416\u201324.  212.  Yoshimura K, Maekura R, Hiraga T, et al. Identification of three exercise-induced mortality risk factors in patients with COPD. COPD J Chronic Obstr Pulm Dis 2014;11(6):615\u201326.  213.  Guenette JA, Webb KA, O\u2019Donnell DE. Does dynamic hyperinflation contribute to dyspnoea during exercise in patients with COPD? Eur Respir J 2012;40(2):322\u20139.  214.  Whipp B, Huntsman D, Stoner N, Lamarra N, Wasserman K. A constant which determines the duration of tolerance to high-intensity work. Fed Proc 1982;41:1591.  215.  Moritani T, Nagata A, Devries HA, et al. Critical power as a measure of physical work capacity and anaerobic threshold. Ergonomics 1981;24(5):339\u201350.  216.  Poole DC, Ward SA, Gardner GW, Whipp BJ. Metabolic and respiratory profile of the upper limit for prolonged exercise in man. Ergonomics 1988;31(9):1265\u201379.  217.  Jenkins DG, Quigley BM. The y-intercept of the critical power function as a measure of anaerobic work capacity. Ergonomics 1991;34(1):13\u201322.  218.  Fukuba Y, Miura A, Endoi M, et al. The curvature constant parameter of the power-duration curve for varied-power exercise. Med Sci Sports Exerc 2003;35(8):1413\u20138.  219.  Monod H, Scherrer J. The work capacity of a synergic muscular group. Ergonomics 1965;8(3):329\u201338.  220.  Poole DC, Ward SA, Whipp BJ. The effects of training on the metabolic and respiratory profile of high-intensity cycle ergometer exercise. Eur J Appl Physiol 1990;59:421\u20139.  221.  Jones AM, Wilkerson DP, DiMenna F, Fulford J, Poole DC. Muscle metabolic responses to exercise above and below the \u201ccritical power\u201d assessed using 31P-MRS. Am J Physiol Regul Integr Comp Physiol 2008;294:R585\u201393.  222.  Vanhatalo A, Black MI, DiMenna FJ, et al. The mechanistic bases of the power\u2013time relationship: muscle metabolic responses and relationships to muscle fibre type. J Physiol 2016;594(15):4407\u201323.  223.  Jones AM, Burnley M, Black MI, Poole DC, Vanhatalo A. The maximal metabolic steady state: redefining the \u2018gold standard.\u2019 Physiol Rep 2019;7(10):e14098.  224.  Ferguson C, Whipp BJ, Cathcart AJ, Rossiter HB, Turner AP, Ward SA. Effects of prior very-heavy intensity exercise on indices of aerobic function and high-intensity exercise tolerance. J Appl Physiol 2007;103:812\u201322.  225.  Murgatroyd SR, Ferguson C, Ward SA, Whipp BJ, Rossiter HB. Pulmonary O2 uptake kinetics as a determinant of high-intensity exercise tolerance in humans. J Appl Physiol 2011;110(6):1598\u2013606.  151 226.  Miura A, Sato HH, Sato HH, Whipp BJW, Fukuba Y. The effect of glycogen depletion on the curvature constant parameter of the power-duration curve for cycle ergometry. Ergonomics 2000;43(1):133\u201341.  227.  Green S, Dawson BT, Goodman C, Carey MF. Y-intercept of the maximal work-duration relationship and anaerobic capacity in cyclists. Eur J Appl Physiol 1994;69:550\u20136.  228.  Deb SK, Gough LA, Sparks SA, McNaughton LR. Determinants of curvature constant (W\u2019) of the power duration relationship under normoxia and hypoxia: the effect of pre-exercise alkalosis. Eur J Appl Physiol 2017;117(5):901\u201312.  229.  Hill DW. The critical power concept: A review. Sport Med 1993;16(4):237\u201354.  230.  Gaesser G, Carnevale T, Garfinkel A, Walter D, Womack C. Estimation of critial power with nonlinear and linear models. Med Sci Sports Exerc 1995;27(10):1430\u20138.  231.  Bull AJ, Housh TJ, Johnson GO, Perry SR. Effect of mathematical modeling on the estimation of critical power. Med Sci Sport Exerc 2000;32(2):526\u201330.  232.  Bergstrom HC, Housh TJ, Zuniga JM, et al. Differences among estimates of critial power and anaerobic work capacity derived from five mathematicl models and the three-minute all-out test. J strength Cond Res 2014;28(3):592\u2013600.  233.  Muniz-Pumares D, Karsten B, Triska C, Glaister M. Methodological approached and related challenged associated with the determination of critial power and curvature constant. J strength Cond Res 2018;33(2):584\u201396.  234.  Housh DJ, Housh TJ, Bauge SM. A methodological consideration for the determination of critical power and anaerobic work capacity. Res Q Exerc Sport 1990;61(4):406\u20139.  235.  Mattioni Maturana F, Fontana FY, Pogliaghi S, Passfield L, Murias JM. Critical power: How different protocols and models affect its determination. J Sci Med Sport 2018;21(7):742\u20137.  236.  Vanhatalo A, Jones AM, Burnley M. Application of Critical Power in Sport. Int J Sports Physiol Perform 2011;6:128\u201336.  237.  Hughson RL, Orok CJ, Staudt LE. A high velocity treadmill running test to assess endurance running potential. Int J Sport Med Sport Med 1984;5:23\u20135.  238.  Housh DJ, Housh TJ, Bauge SM. The accuracy of the critical power test for predicting time to exhaustion during cycle ergometry. Ergonomics 1989;32(8):997\u20131004.  239.  Chorley A, Lamb KL. The application of critical power, the work capacity above critical power (W\u2032), and its reconstitution: A narrative review of current evidence and implications for cycling training prescription. Sports 2020;8(9):123.  240.  Skiba PF, Clarke D, Vanhatalo A, Jones AM. Validation of a novel intermittent W\u2032 model for cycling using field data. Int J Sports Physiol Perform 2014;9(6):900\u20134.  152 241.  Jenkins DG, Quigley BM. Blood lactate in trained cyclists during cycle ergometry at critical power. Eur J Appl Physiol 1990;61:278\u201383.  242.  McLellan T, Cheung K. A comparative evaluation of the individual anaerobic threshold and the critial power. Med Sci Sports Exerc 1992;24(5):543\u201350.  243.  Overend TJ, Cunningham DA, Paterson DH, Smith WDF. Physiological responses of young and elderly men to prolonged exercise at critical power. Eur J Appl Physiol 1992;64:187\u201393.  244.  Keir DA, Fontana FY, Robertson TC, et al. Exercise intensity thresholds: Identifying the boundaries of sustainable performance. Med Sci Sports Exerc 2015;47(9):1932\u201340.  245.  Burnley M, Doust JH, Vanhatalo A. A 3-min all-out test to determine peak oxygen uptake and the maximal steady state. Med Sci Sports Exerc 2006;38(11):1995\u20132003.  246.  Vanhatalo A, Fulford J, Dimenna FJ, Jones AM. Influence of hyperoxia on muscle metabolic responses and the power-duration relationship during severe-intensity exercise in humans: A 31P magnetic resonance spectroscopy study. Exp Physiol 2010;95(4):528\u201340.  247.  Jones AM, Grassi B, Christensen PM, Krustrup P, Bangsbo J, Poole DC. Slow component of V\u0307o2 kinetics: Mechanistic bases and practical applications. Med Sci Sports Exerc 2011;43(11):2046\u201362.  248.  Barstow TJ, Jones AM, Nguyen PH, Casaburi R. Influence of muscle fiber type and pedal frequency on oxygen uptake kinetics of heavy exercise. J Appl Physiol 1996;81(4):1642\u201350.  249.  Carter H, Pringle JSM, Boobis L, Jones AM, Doust JH. Muscle glycogen depletion alters oxygen uptake kinetics during heavy exercise. Med Sci Sports Exerc 2004;36(6):965\u201372.  250.  Endo MY, Kobayakawa M, Kinugasa R, et al. Thigh muscle activation distribution and pulmonary VO2 kinetics during moderate, heavy, and very heavy intensity cycling exercise in humans. Am J Physiol Regul Integr Comp Physiol 2007;293:812\u201320.  251.  Muniz-Pumares D, Karsten B, Triska C, Glaister M. Methodological approached and related challenges associated with the determination of critical power and curvature constant. J Strength Cond Res [Internet] 2019;32:584\u201396. Available from: www.nsca.com 252.  Bishop D, Jenkins DC, Howard A, Jenkins DG. The Critical Power Function is Dependent on the Duration of the Predictive Exercise Tests Chosen. Int J Sport Med 1998;19:125\u20139.  253.  Jenkins D, Kretek K, Bishop D. The duration of predicting trials influences time to fatigue at critical power. J Sci Med Sport 1998;1(4):213\u20138.  254.  Vautier JF, Vandewalle H, Arabi H, Monod H. Critical power as an endurance index. Appl Ergon 1995;26(2):117\u201321.  153 255.  Vandewalle H, Kapitaniak B, Griin S, et al. Comparison between a 30-s all-out test and a time-work test on a cycle ergometer. Eur J Appl Physiol 1989;58:375\u201381.  256.  Clingeleffer A, Mc Naughton LR, Davoren B. The use of critical power as a determinant for establishing the onset of blood lactate accumulation. Eur J Appl Physiol 1994;68:182\u20137.  257.  Billat V, Renoux J, Pinoteau J, Petit B, Koralsztein JP. Times to exhaustion at 90, 100, and 105% of velocity at VO2max (maximal aerobic speed) and critical speed in elite longdistance runners. Arch Physiol Biochem 1995;103(2):129\u201335.  258.  Vandewalle H, Vautier J, Kachouri M, Lechevalier J, Mondon H. Work-exhaustion time relationships and the critical power concept: a critical review. J Sport Med Phys Fit 1997;37:89\u2013102.  259.  Gal\u00e1n-Rioja M\u00c1, Gonz\u00e1lez-Moh\u00edno F, Poole DC, Gonz\u00e1lez-Rav\u00e9 JM. Relative proximity of critical power and metablic\/ventilatory thresholds: systematic review and meta-analysis. Sport Med 2020;50(10):1771\u201383.  260.  Caen K, Vermeire K, Bourgois JG, Boone J. Exercise thresholds on trial: are they really equivalent? Med Sci Sports Exerc 2018;50(6):1277\u201384.  261.  Dekerle J, Baron B, Dupont L, Vanvelcenaher J, Pelayo P. Maximal lactate steady state, respiratory compensation threshold and critical power. Eur J Appl Physiol 2003;89(3\u20134):281\u20138.  262.  Broxterman RM, Ade CJ, Craig JC, Wilcox SL, Schlup SJ, Barstow TJ. The relationship between critical speed and the respiratory compensation point: Coincidence or equivalence. Eur J Sport Sci 2015;15(7):631\u20139.  263.  Jenkins DG, Quigley BM. Blood lactate in trained cyclists during cycle ergometry at critical power. Eur J Appl Physiol 1990;61:278\u201383.  264.  Neder JA, Jones PW, Nery LE, Whipp BJ. The effect of age on the power\/duration relationship and the intensity-domain limits in sedentary men. Eur J Appl Physiol 2000;82:326\u201332.  265.  Puente-Maestu L, SantaCruz A, Vargas T, Mart\u00ednez-Abad Y, Whipp BJ. Effects of training on the tolerance to high-intensity exercise in patients with severe COPD. Respiration 2003;70(4):367\u201370.  266.  Malaguti C, Nery LE, Corso SD, et al. Alternative strategies for exercise critical power estimation in patients with COPD. Eur J Appl Physiol 2006;96(1):59\u201365.  267.  Porszasz J, Rambod M, Van der Vaart H, et al. Sinusoidal high-intensity exercise does not elicit ventilatory limitation in chronic obstructive pulmonary disease. Exp Physiol 2013;98(6):1102\u201314.  268.  Van Der Vaart H, Murgatroyd SR, Rossiter HB, Chen C, Casaburi R, Porszasz J. Selecting constant work rates for endurance testing in COPD: the role of the power-duration relationship. COPD J Chronic Obstr Pulm Dis 2014;11(3):267\u201376.  154 269.  Neder JA, Jones PW, Nery LE, Whipp BJ. Determinants of the exercise endurance capacity in patients with chronic obstructive pulmonary disease: The power-duration relationship. Am J Respir Crit Care Med 2000;162:497\u2013504.  270.  Billat VL, Sirvent P, Py G, Koralsztein J-P, Mercier J. The concept of maximal lactate steady state: A bridge between biochemistry, physiology and sport science. Sport Med 2003;33(6):407\u201326.  271.  Beneke R. Methodological aspects of maximal lactate steady state-implications for performance testing. Eur J Appl Physiol 2003;89(1):95\u20139.  272.  Bergstrom HC, Housh TJ, Cochrane KC, et al. Application of the critical heart model to treadmill running. J strength Cond Res 2015;29(8):2237\u201348.  273.  Mielke M, Housh TJ, Hendrix CR, et al. A test for determining critical heart rate using the critical power model. J strength Cond Res 2011;25(2):504\u201310.  274.  Malek MH, Housh TJ, Schmidt RJ, Coburn JW, Beck TW. Proposed tests for measuring the running velocity at the oxygen consumption and heart rate thresholds for treadmill exercise. J strength Cond Res [Internet] 2005;19(4):847\u201352. Available from: http:\/\/journals.lww.com\/nsca-jscr 275.  Nakamura F, Okuno N, Perandini L, et al. Critical power can be estimated from nonexhaustive tests based on rating of percieved exertion responses. J strength Cond Res 2008;22(3):937\u201343.  276.  Devries HA, Moritani T, Nagata A, Magnussen KE. The relation between critical power and neuromuscular fatigue as estimated from electromyographic data. Ergonomics 1982;25(9):783\u201391.  277.  Puente-Maestu L, Villar F, De Miguel J, et al. Clinical relevance of constant power exercise duration changes in COPD. Eur Respir J 2008;34(2):340\u20135.  278.  Vivodtzev I, Gagnon P, Pepin V, et al. Physiological correlates of endurance time variability during constant-workrate cycling exercise in patients with copd. PLoS One 2011;6(2):e17007 [doi:10.1371\/journal.pone.0017007].  279.  Horowitz MB, Littenberg B, Mahler DA. Dyspnea ratings for prescribing exercise intensity in patients with COPD. Chest 1996;109(5):1169\u201375.  280.  Horowitz MB, Mahler DA. Dyspnea ratings for prescription of cross-modal exercise in patients with COPD. Chest 1998;113(1):60\u20134.  281.  Agusti A, Calverley PMA, Celli B, et al. Characterisation of COPD heterogeneity in the ECLIPSE cohort. Respir Res 2010;11(1):122.  282.  Casanova C, De Torres JP, Aguirre-Ja\u00edme A, et al. The progression of chronic obstructive pulmonary disease is heterogeneous: The experience of the BODE cohort. Am J Respir Crit Care Med 2011;184(9):1015\u201321.  283.  Raffestin B, Escourrou P, Legrand A, Duroux P, Lockhart A. Circulatory transport of 155 oxygen in patients with chronic airflow obstruction exercising maximally. Am Rev Respir Dis 1982;125(4 I):426\u201331.  284.  Gelinas JC, Lewis NC, Harper MI, et al. Aerobic exercise training does not alter vascular structure and function in chronic obstructive pulmonary disease. Exp Physiol 2017;102(11):1548\u201360.  285.  Davidson WJ, Verity WS, Traves SL, Leigh R, Ford GT, Eves ND. Effect of incremental exercise on airway and systemic inflammation in patients with COPD. J Appl Physiol 2012;112(12):2049\u201356.  286.  Wanger J, Clausen JL, Coates A, et al. Standardisation of the measurement of lung volumes. Eur Respir J 2005;26(3):511\u201322.  287.  Graham BL, Brusasco V, Burgos F, et al. 2017 ERS\/ATS standards for single-breath carbon monoxide uptake in the lung. Eur Respir J 2017;49(1):160016.  288.  Graham BL, Steenbruggen I, Barjaktarevic IZ, et al. Standardization of spirometry 2019 update an official American Thoracic Society and European Respiratory Society technical statement. Am J Respir Crit Care Med 2019;200(8):E70\u201388.  289.  Borg G. Psychophysical bases of perceived exertion. Med Sci Sports Exerc 1982;14(5):377\u201381.  290.  Beaver WL, Wasserman K, Whipp BJ. A new method for detecting anaerobic threshold by gas exchange. J Appl Physiol 1986;60(6):2020\u20137.  291.  Wasserman K, Whipp BJ, Koyal SN, Beaver WL. Anaerobic threshold and respiratory gas exchange during exercise. J Appl Physiol [Internet] 1973;35(2):236\u201343. Available from: www.physiology.org\/journal\/jappl 292.  Blackie SP, Fairbarn MS, McElvaney NG, Wilcox PG, Morrison NJ, Pardy RL. Normal values and ranges for ventilation and breathing pattern at maximal exercise. Chest 1991;100(1):136\u201342.  293.  Lewthwaite H, Benedetti A, Stickland MK, et al. Normative Peak Cardiopulmonary Exercise Test Responses in Canadian Adults Aged \u226540 Years. Chest 2020;158(6):2532\u201345.  294.  Inbar O, Oren A, Scheinowitz M, Rotstein A, Dlin R, Casaburi R. Normal cardiopulmonary responses during incremental exercise in 20- to 70-yr-old men. Med Sci Sports Exerc 1994;26(5):538\u201346.  295.  Alberto Neder J, Nery LE, Peres C, et al. Reference values for dynamic responses to incremental cycle ergometry in males and females aged 20 to 80. Am J Respir Crit Care Med 2001;164(8 I):1481\u20136.  296.  Whipp BJ, Higgenbotham MB, Cobb FC. Estimating exercise stroke volume from asymptotic oxygen pulse in humans. J Appl Physiol 1996;81(6):2674\u20139.  297.  Cote CG, Pinto-Plata VM, Marin JM, Nekach H, Dordelly LJ, Celli BR. The modified 156 BODE index: Validation with mortality in COPD. Eur Respir J 2008;32(5):1269\u201374.  298.  Bruce RM, Turner A, White MJ. Ventilatory responses to muscle metaboreflex activation in chronic obstructive pulmonary disease. J Physiol 2016;594(20):6025\u201335.  299.  Johnson BD, Weisman IM, Zeballos ; R Jorge, Beck KC. Emerging concepts in the evaluation of ventilatory limitation during exercise: The exercise tidal flow-volume loop. Chest 1999;116:488\u2013503.  300.  Tanaka H, Monahan KD, Seals DR. Age-predicted maximal heart rate revisited. J Am Coll Cardiol 2001;37(1):153\u20136.  301.  O\u2019Donnell DE, Lam M, Webb KA. Measurement of symptoms, lung hyperinflation, and endurance during exercise in chronic obstructive pulmonary disease. Am J Respir Crit Care Med 1998;158:1557\u201365.  302.  McLellan T, Cheung K. A comparative evaluation of the individual anaerobic thershold and the critical power. Med Sci Sports Exerc 1992;24(5):543\u201350.  303.  Brickley G, Doust J, Williams CA. Physiological responses during exercise to exhaustion at critical power. Eur J Appl Physiol 2002;88:146\u201351.  304.  O\u2019Donnell DE, Voduc N, Fitzpatrick M, Webb KA. Effect of salmeterol on the ventilatory response to exercise in chronic obstructive pulmonary disease. Eur Respir J 2004;24(1):86\u201394.  305.  Nery LE, Wasserman K, Andrews JD, Huntsman DJ, Hansen JE, Whipp BJ. Ventilatory and gas exchange kinetics during exercise in chronic airways obstruction. J Appl Physiol 1982;53(6):1594\u2013602.  306.  Puente-Maestu L, Palange P, Casaburi R, et al. Use of exercise testing in the evaluation of interventional efficacy: An official ERS statement. Eur Respir J 2016;47(2):429\u201360.  307.  Dempsey JA, Smith CA. Pathophysiology of human ventilatory control: Number 6 in the series \u201cphysiology in respiratory medicine.\u201d Eur. Respir. J. 2014;44(2):495\u2013512.  308.  Maltais F, Leblanc P, Jobin J, et al. Intensity of training and physiologic adaptation in patients with chronic obstructive pulmonary disease. Am J Respir Crit Care Moo 1997;155:555\u201361.  309.  Guenette JA, Jensen D, Webb KA, Ofir D, Raghavan N, O\u2019Donnell DE. Sex differences in exertional dyspnea in patients with mild COPD: Physiological mechanisms. Respir Physiol Neurobiol 2011;177(3):218\u201327.  310.  Casaburi R, Patessio A, Ioli F, Zanaboni S, Donner CF, Wasserman K. Reductions in exercise lactic acidosis and ventilation as a result of exercise training in patients with obstructive lung disease. AM REV RESPIR DIS 1991;143(1):9\u201318.  311.  Mancini DM, Eisen H, Kussmaul W, Mull R, Edmonds LH, Wilson JR. Value of peak exercise oxygen consumption for optimal timing of cardiac transplantation in 157 ambulatory patients with heart failure. Circulation 1991;83(3):778\u201386.  312.  Kaminsky LA, Imboden MT, Arena R, Myers J. Reference standards for cardiorespiratory fitness measured with cardiopulmonary exercise testing using cycle ergometry: data from the fitness registry and the importance of exercise national database (FRIEND) registry. Mayo Clin Proc 2017;92(2):228\u201333.  313.  Jones AM, Vanhatalo A, Burnley M, Morton RH, Poole DC. Critical power: Implications for determination of VO2max and exercise tolerance. Med Sci Sports Exerc 2010;42(10):1876\u201390.  314.  Seiler S, T\u00f8nnessen E. Intervals, thresholds, and long slow distance: the role of intensity and duration in endurance training. Sportscience 2009;13:32\u201353.  315.  Sasso JP, Eves ND, Christensen JF, Koelwyn GJ, Scott J, Jones LW. A framework for prescription in exercise-oncology research. J Cachexia Sarcopenia Muscle 2015;6:115\u201324.  316.  Arbab-Zadeh A, Perhonen M, Howden E, et al. Cardiac remodeling in response to 1 year of intensive endurance training. Circulation 2014;130(24):2152\u201361.  317.  Laughlin MH, Roseguini B. Mechanisms for exercise training-induced increases in skeletal muscle blood flow capacity: differences with interval sprint training versus aerobic endurance training. J Physiol Pharmacol 2008;59(7):71\u201388.  318.  Little JP, Safdar A, Wilkin GP, Tarnopolsky MA, Gibala MJ. A practical model of low-volume high-intensity interval training induces mitochondrial biogenesis in human skeletal muscle: potential mechanisms. J Physiol 2010;588(6):1011\u201322.  319.  Hood MS, Little JP, Tarnopolsky MA, Myslik F, G\u00edbala MJ, Joseph Gibala M. Low-volume interval training improves muscle oxidative capacity in sedentary adults. Med Sei Sport Exerc 2011;43(10):1849\u201356.  320.  Palange P, Forte S, Onorati P, Manfredi F, Serra P, Carlone S. Ventilatory and metabolic adaptations to walking and cycling in patients with COPD. J Appl Physiol 2000;88:1715\u201320.     158 Appendix Table A.1. MDSS Subgroup Analysis: Characteristics and Incremental CPET Responses Variable Ventilatory (n=6) Combined (n=7) Cardiovascular (n=6) Subgroup  (n=4)  P Male:Female 4:2 2:5 2:4 4:0  - Age (years) 63 \u00b1 9 72 \u00b1 6 70 \u00b1 8 66 \u00b1 7 0.21 Height (m) 1.70 \u00b1 0.09 1.65 \u00b1 0.11 1.64 \u00b1 0.09 1.74 \u00b1 0.05 0.26 BMI (kg\/m2) 24.2 \u00b1 2.7 26.9 \u00b1 3.3 25.2 \u00b1 4.0 28.8 \u00b1 3.1 0.17 FEV1 (L) 1.41 \u00b1 0.51\u00a7 1.65 \u00b1 0.44 2.19 \u00b1 0.45 2.54 \u00b1 0.85 0.02 FEV1 (%pred) 46 \u00b1 10*\u2020\u00a7 71 \u00b1 14 90 \u00b1 13 78 \u00b1 25 <0.01 Range in FEV1 (%pred) 31-56 50-90 79-106 49-111 - FEV1\/FVC 35 \u00b1 8*\u2020 54 \u00b1 6 61 \u00b1 8 51 \u00b1 16 <0.01 RV (L) 3.47 \u00b1 0.98 2.57 \u00b1 0.66 2.72 \u00b1 0.44 3.14 \u00b1 0.91 0.18 TLC (L) 7.46 \u00b1 1.30 5.65 \u00b1 1.28\u00a7 6.31 \u00b1 0.79 8.04 \u00b1 1.15 0.01 RV\/TLC (%) 46 \u00b1 6 46 \u00b1 7 43 \u00b1 6 39 \u00b1 6 0.33 DLCO\/VA (%) 66 \u00b1 11 81 \u00b1 11 80 \u00b1 20 87 \u00b1 33 0.32 VEpeak (L\/min) 46.1 \u00b1 17.0 58.0 \u00b1 16.6 57.6 \u00b1 10.8 68.6 \u00b1 5.1 0.15 VEpeak (%pred MVC) 93 \u00b1 9* 95 \u00b1 10 70 \u00b1 8 85 \u00b1 29 0.03 VE reserve (L\/min) 2.4 \u00b1 3.7* 2.7 \u00b1 5.4 26.1 \u00b1 13.4 19 \u00b1 29 0.02 HRpeak (beats\/min) 125 \u00b1 14 140 \u00b1 8 147 \u00b1 13 147 \u00b1 20 0.07 HRpeak (%pred) 80 \u00b1 10*\u2020\u00a7 95 \u00b1 6 97 \u00b1 7 95 \u00b1 10 <0.01 Cardiac reserve (beats\/min) 33 \u00b1 17*\u2020\u00a7 8 \u00b1 9 4 \u00b1 11 7 \u00b1 15 <0.01 Workload (watts) 75 \u00b1 26 91 \u00b1 39 102 \u00b1 34 130 \u00b1 41 0.20 VO2peak (ml\/kg\/min) 16.5 \u00b1 5.3 18.1 \u00b1 5.0 21.1 \u00b1 3.8 21 \u00b1 6 0.32 VO2peak (L\/min) 1.17 \u00b1 0.47 1.51 \u00b1 0.50 1.60 \u00b1 0.50 1.88 \u00b1 0.51 0.26 VCO2 (L\/min) 1.18 \u00b1 0.54 1.64 \u00b1 0.56 1.71 \u00b1 0.50 2.08 \u00b1 0.48 0.09 RER 1.00 \u00b1 0.16 1.06 \u00b1 0.09 1.07 \u00b1 0.07 1.12 \u00b1 0.04 0.33 VE\/VCO2 nadir 39 \u00b1 5 34 \u00b1 3 35 \u00b1 7 33 \u00b1 7 0.30 PETO2 (mmHg) 105.6 \u00b1 7.8 108.1 \u00b1 3.8 108.9 \u00b1 7.2 110.0 \u00b1 6.1 0.21 PETCO2 (mmHg) 34.4 \u00b1 3.6 35.0 \u00b1 2.8 34.8 \u00b1 6.2 34.8 \u00b1 5.0 0.38 VD\/VT  0.29 \u00b1 0.06\u00a7 0.23 \u00b1 0.05\u00a7 0.17 \u00b1 0.06 0.14 \u00b1 0.09 <0.01 Tidal volume (L) 1.44 \u00b1 0.35\u00a7 1.54 \u00b1 0.61 1.77 \u00b1 0.30 2.47 \u00b1 0.79 0.03 \u0394 Tidal Volume (L) 0.46 \u00b1 0.34\u00a7 0.73 \u00b1 0.54 0.98 \u00b1 0.22 1.55 \u00b1 0.71 0.01 \u0394 IC (L) -0.72 \u00b1 0.29 -0.44 \u00b1 0.20 -0.52 \u00b1 0.24 -0.42 \u00b1 0.57 0.39 SpO2 (%) 93 \u00b1 2 95 \u00b1 3 96 \u00b1 3 94 \u00b1 1 0.09 Blood Lactate (mmol\/L) 2.8 \u00b1 0.9\u00a7 4.4 \u00b1 2.3 5.0 \u00b1 1.0 6.4 \u00b1 1.3 0.02 Dyspnea (Borg Units) 5.3 \u00b1 1.4 5.4 \u00b1 2.0 4.4 \u00b1 2.1 5.5 \u00b1 1.9 0.84 Leg fatigue (Borg Units) 4.2 \u00b1 2.6 5.6 \u00b1 2.9 5.6 \u00b1 1.9 6.5 \u00b1 1.9 0.43 Dyspnea\/LF\/both (%) 67\/33\/0 43\/43\/14 17\/66\/17 0\/75\/25 -  Abbreviations: BMI, body mass index; FEV1, forced expiratory volume in 1 second; FEV1\/FVC, ratio of forced expiratory volume in 1 second to forced vital capacity; RV, residual volume; TLC, total lung capacity; DLCO\/VA, the ratio of diffusion of carbon monoxide to alveolar ventilation; VEpeak, peak minute ventilation; MVC, estimated maximum ventilatory capacity; HRpeak, peak heart rate; VO2peak, peak oxygen consumption; VCO2, volume of carbon dioxide produced; RER, respiratory exchange ratio; VE\/VCO2, ratio of minute ventilation to carbon dioxide produced; PETO2, partial pressure of end-tidal oxygen; PETCO2, partial pressure of end-tidal carbon dioxide; VD\/VT, deadspace to tidal volume ratio; \u0394 IC, change in inspiratory capacity from rest to peak; SpO2, oxyhemoglobin saturation;  \u0394SpO2, change in oxyhemoglobin saturation from rest to peak. Comparisons: *p=0.05, ventilatory vs. cardiovascular. \u2020p=0.05, ventilatory vs. combined. \u2021p=0.05, combined vs. cardiovascular. \u00a7p=0.05, subgroup of patients in whom MDSS could not be calculate vs. phenotype.   159 Table A.2. MDSS Subgroup Analysis: Constant Load Trial Exercise Responses Variable  Ventilatory (n=6) Combined (n=7) Cardiovascular (n=6) Subgroup (n=4)  P Constant Load Trial #1      Workload (watts) 38 \u00b1 16 56 \u00b1 26 72 \u00b1 27  81 \u00b1 31 0.06 Workload (%Wmax) 50 \u00b1 6* 59 \u00b1 7 69 \u00b1 3 63 \u00b1 5 <0.01 Duration (sec) 940 \u00b1 475 1102 \u00b1 391 617 \u00b1 293 696 \u00b1 283 0.13 Heart rate (beats\/min) 110 \u00b1 12*\u2020 129 \u00b1 17 134 \u00b1 12 137 \u00b1 19 0.01 VE (L\/min) 36.4 \u00b1 12.3 43.4 \u00b1 16.3 47.9 \u00b1 7.2 54.5 \u00b1 1.8 0.14 VE reserve (L\/min) 12.1 \u00b1 5.8* 12.6 \u00b1 5.4 28.3 \u00b1 8.3 34.6 \u00b1 30.2 0.03 IRV (L) 0.74 \u00b1 0.36 0.52 \u00b1 0.26 0.84 \u00b1 0.52 1.10 \u00b1 0.31 0.13 \u0394 Tidal volume (L) 0.36 \u00b1 0.15 0.53 \u00b1 0.59 0.82 \u00b1 0.34 1.25 \u00b1 0.76 0.05 \u0394 IC (L) -0.47 \u00b1 0.13 -0.40 \u00b1 0.18 -0.33 \u00b1 0.11 -0.41 \u00b1 0.89 0.31 RER  0.91 \u00b1 0.09 0.96 \u00b1 0.06 1.00 \u00b1 0.02 0.99 \u00b1 0.01 0.13 Blood Lactate (mmol\/L) 2.1 \u00b1 0.9\u00a7 3.5 \u00b1 2.1 4.3 \u00b1 0.9 5.1 \u00b1 0.7 <0.01 SpO2 (%) 93 \u00b1 2 93 \u00b1 2 95 \u00b1 4 95 \u00b1 2 0.34 Dyspnea (Borg unit) 5.5 \u00b1 1.6 4.9 \u00b1 1.8 6.2 \u00b1 2.8 4.5 \u00b1 1.9 0.58 Leg Fatigue (Borg unit) 5.0 \u00b1 2.4 4.7 \u00b1 3.0 7.2 \u00b1 2.6 5.0 \u00b1 1.4 0.33 Dyspnea\/LF\/both (%) 50\/33\/17 43\/43\/14 67\/33\/0 0\/25\/75 - Constant Load Trial #2      Workload (watts) 43 \u00b1 16 63 \u00b1 29 78 \u00b1 29 89 \u00b1 33 0.05 Workload (%Wmax) 57 \u00b1 4*\u2020\u00a7 67 \u00b1 8\u2021 76 \u00b1 3 70 \u00b1 5 <0.01 Duration (sec) 653 \u00b1 507 962 \u00b1 610 369 \u00b1 179 691 \u00b1 358 0.22 Heart rate (beats\/min) 106 \u00b1 12*\u2020\u00a7 132 \u00b1 16 140 \u00b19 138 \u00b1 19 <0.01 VE (L\/min) 37.8 \u00b1 13.3 45.2 \u00b1 15.4 52.2 \u00b1 9.0 56.3 \u00b1 4.3 0.10 VE reserve (L\/min) 10.6 \u00b1 4.3 10.6 \u00b1 5.9 24.0 \u00b1 10.0 31.8 \u00b1 27.5 0.05 IRV (L) 0.69 \u00b1 0.39 0.50 \u00b1 0.35 0.71 \u00b1 0.40 0.72 \u00b1 0.40 0.54 \u0394 Tidal volume (L) 0.36 \u00b1 0.18\u00a7 0.60 \u00b1 0.53 0.94 \u00b1 0.33 1.48 \u00b1 0.82 0.01 \u0394 IC (L) -0.49 \u00b1 0.36 0.44 \u00b1 0.16 -0.28 \u00b1 0.15 -0.65 \u00b1 0.56 0.34 RER  0.92 \u00b1 0.12 0.99 \u00b1 0.06 1.03 \u00b1 0.06 1.03 \u00b1 0.05 0.08 Blood Lactate (mmol\/L) 2.5 \u00b1 1.1\u00a7 4.0 \u00b1 2.2 4.3 \u00b1 0.8 5.8 \u00b1 1.1 0.02 SpO2 (%) 93 \u00b1 2 93 \u00b1 2 96 \u00b1 3 96 \u00b1 2 0.07 Dyspnea (Borg unit) 4.8 \u00b1 1.2 4.7 \u00b1 2.2 5.3 \u00b1 2.7 4.5 \u00b1 1.3 0.98 Leg Fatigue (Borg unit) 4.5 \u00b1 1.8 5.0 \u00b12.8 6.3 \u00b1 3.4 5.0 \u00b1 0.8 0.64 Dyspnea\/LF\/both 50\/33\/17 28\/43\/28 67\/33\/0 0\/75\/25 -      160 Table A.2. MDSS Subgroup Analysis: Constant Load Trial Exercise Responses (Cont\u2019d) Constant Load Trial #3      Workload (watts) 51 \u00b1 19 70 \u00b1 31 85 \u00b1 32 99 \u00b1 37 0.10 Workload (%Wmax) 67 \u00b1 3*\u2020 76 \u00b1 8 82 \u00b1 3 77 \u00b1 6 <0.01 Duration (sec) 419 \u00b1 401 380 \u00b1 187 271 \u00b1 96 492 \u00b1 293 0.42 Heart rate (beats\/min) 115 \u00b1 4*\u2020\u00a7 138 \u00b1 13 139 \u00b1 14 140 \u00b1 21 0.01 VE (L\/min) 38.5 \u00b1 13.4 50.2 \u00b1 16.7 51.3 \u00b1 11.1 59.5 \u00b1 4.8 0.12 VE reserve (L\/min) 9.9 \u00b1 4.2 5.8 \u00b1 5.8\u2021 24.9 \u00b1 8.3 28.6 \u00b1 27.1 <0.01 IRV (L) 0.65 \u00b1 0.24 0.45 \u00b1 0.25\u00a7 0.61 \u00b1 0.26 1.03 \u00b1 0.25 0.04 \u0394 Tidal volume (L) 0.44 \u00b1 0.30\u00a7 0.63 \u00b1 0.66 1.06 \u00b1 0.18 1.45 \u00b1 0.81 0.03 \u0394 IC (L) -0.67 \u00b1 0.24\u2020 0.33 \u00b1 0.20 -0.46 \u00b1 0.20 -0.23 \u00b1 0.67 0.13 RER 0.95 \u00b1 0.13 1.03 \u00b1 0.51 1.07 \u00b1 0.08 1.06 \u00b1 0.03 0.17 Blood Lactate (mmol\/L) 2.5 \u00b1 1.4 4.5 \u00b1 2.3 4.4 \u00b1 1.6 5.4 \u00b1 2.2 0.13 SpO2 (%) 93 \u00b1 2 94 \u00b1 3 96 \u00b1 2 95 \u00b1 2 0.08 Dyspnea (Borg unit) 5.3 \u00b1 1.4 5.1 \u00b1 2.4 4.8 \u00b1 1.8 3.5 \u00b1 1.7 0.49 Leg Fatigue (Borg unit) 4.5 \u00b1 1.4 5.1 \u00b1 2.7 6.7 \u00b1 2.7 4.5 \u00b1 1.3 0.35 Dyspnea\/LF\/both 100\/0\/0 14\/29\/57 17\/33\/50 25\/25\/50 -  Abbreviations: VO2, volume of oxygen consumed; VE, minute ventilation; IRV, inspiratory reserve volume; \u0394Tidal volume, the change in tidal volume from rest to end-exercise; \u0394IC, the change in inspiratory capacity from rest to end-exercise; SpO2, oxyhemoglobin saturation; LF, leg fatigue; Both, participants stated both dyspnea and leg fatigue as the limiting symptom at end-exercise. Comparisons: *p=0.05, ventilatory vs. cardiovascular. \u2020p=0.05, ventilatory vs. combined. \u2021p=0.05, combined vs. cardiovascular. \u00a7p=0.05, subgroup of patients in whom MDSS could not be calculate vs. phenotype ","type":"literal","lang":"en"}],"http:\/\/www.europeana.eu\/schemas\/edm\/hasType":[{"value":"Thesis\/Dissertation","type":"literal","lang":"en"}],"http:\/\/vivoweb.org\/ontology\/core#dateIssued":[{"value":"2021-09","type":"literal","lang":"en"}],"http:\/\/www.europeana.eu\/schemas\/edm\/isShownAt":[{"value":"10.14288\/1.0398500","type":"literal","lang":"en"}],"http:\/\/purl.org\/dc\/terms\/language":[{"value":"eng","type":"literal","lang":"en"}],"https:\/\/open.library.ubc.ca\/terms#degreeDiscipline":[{"value":"Kinesiology","type":"literal","lang":"en"}],"http:\/\/www.europeana.eu\/schemas\/edm\/provider":[{"value":"Vancouver : University of British Columbia Library","type":"literal","lang":"en"}],"http:\/\/purl.org\/dc\/terms\/publisher":[{"value":"University of British Columbia","type":"literal","lang":"en"}],"http:\/\/purl.org\/dc\/terms\/rights":[{"value":"Attribution-NonCommercial-NoDerivatives 4.0 International","type":"literal","lang":"*"}],"https:\/\/open.library.ubc.ca\/terms#rightsURI":[{"value":"http:\/\/creativecommons.org\/licenses\/by-nc-nd\/4.0\/","type":"literal","lang":"*"}],"https:\/\/open.library.ubc.ca\/terms#scholarLevel":[{"value":"Graduate","type":"literal","lang":"en"}],"http:\/\/purl.org\/dc\/terms\/contributor":[{"value":"Eves, Neil","type":"literal","lang":"en"}],"http:\/\/purl.org\/dc\/terms\/title":[{"value":"Phenotyping exercise limitation in chronic obstructive pulmonary disease to enhance exercise prescription","type":"literal","lang":"en"}],"http:\/\/purl.org\/dc\/terms\/type":[{"value":"Text","type":"literal","lang":"en"}],"https:\/\/open.library.ubc.ca\/terms#identifierURI":[{"value":"http:\/\/hdl.handle.net\/2429\/78770","type":"literal","lang":"en"}]}}