Copyright is owned by the Author of the thesis. Permission is given for a copy to be downloaded by an individual for the purpose of research and private study only. The thesis may not be reproduced elsewhere without the permission of the Author. Clinical Utility of Near-infrared Spectroscopy in Skeletal Muscle A thesis presented in partial fulfillment of the requirements for the degree of Doctor of Philosophy (Sport and Exercise Science) At Massey University, Wellington, New Zealand Adam Alexander Lucero 2021 Abstract Introduction: Near infrared spectroscopy (NIRS) provides for non-invasive assessment of resting skeletal muscle hemodynamic and respiratory responses. However, the interday reliability of skeletal muscle blood flow (mBF) and oxygen consumption (mV̇O2) responses to stressors such as exercise in both healthy and clinical populations has not been established. Moreover, direct comparison of differing NIRS technologies is absent. The purpose of this thesis is three-fold 1) establish a standard protocol for the assessment of resting and exercise skeletal muscle hemodynamics for healthy and clinical populations, 2) compare the reliability of NIRS outcomes in continuous wave (cw) NIRS to the more robust frequency domain (fd) NIRS technology, and 3) assess validity against in-vitro skeletal muscle metabolic parameters. Methods: In the first study, a standard protocol developed to measure mBF, mV̇O2, and perfusion ([tHb]) in the vastus lateralis (VL) at rest and up to 30% of maximum voluntary contraction and mV̇O2 recovery rate constant (k) as an index of muscle oxidative capacity was conducted in twelve healthy adults and was repeated twice within 10 days to establish repeatability. The NIRS measures were conducted using a cw-NIRS device. Secondly, this protocol was repeated in 10, healthy males and 10 non-insulin dependent sedentary males with T2D for characterisation and comparison of outcomes derived from cw-NIRS versus fd-NIRS. Thirdly, cw-NIRS and whole-body oxygen consumption (V̇O2) were measured in 24 men with T2D while performing incremental ramp cycle exercise to volitional exhaustion; in addition, biopsies of the VL were collected. Results: In study 1, mBF and mV̇O2 proportionally increased with intensity (0.55 to 7.68 ml∙min-1∙100ml-1 and 0.05 to 1.86 mlO2∙min-1∙100g-1, respectively) up to 25% MVC where it began to plateau at 30% MVC. For studies 1 and 2, a mBF/mV̇O2 ratio of ~5 was consistent for all exercise stages. For both Healthy and T2D groups, patterns of change and values for mBF and mV̇O2 during exercise were not substantially different ii between devices and were moderate to highly reproducible (ICC: 0.72-0.98). The mean typical error for exercise mBF and mV̇O2 with 90% Confidence Intervals was 0.41 (0.31- 0.59) and 0.38 (0.29-0.55) for ND and T2D, respectively. Substantial differences were seen in ND and T2D, respectively, between CW- and FD-NIRS values for perfusion. Thirdly, the [HHb] primary phase during dynamic exercise was substantially correlated to V̇O2peak while the secondary phase was substantially correlated to measures of mitochondrial function. Conclusion: NIRS can reliably assess mBF and mV̇O2 responses at rest and during low- moderate exercise. The popular cw-NIRS device performed comparably to the more robust fd-NIRS when assessing mBF and mVO2 in both healthy and T2D populations, but cw-NIRS tended to overestimate perfusion, likely due to assumptions of constant scattering. Finally, combining NIRS with external respiration during continuous exercise has potential in investigating barriers to glucose disposal and exercise tolerance in T2D. Taken together, NIRS is a valid tool for applications in research, clinical diagnosis, and therapeutic assessment of skeletal muscle hemodynamics, microvascular, and respiratory plasticity. iii Acknowledgments It takes a village to raise a child, and an army to graduate a PhD candidate. I have met so many wonderful people along my journey through graduate school, many of which played pivotal roles in encouraging me to keep going and pushing me forward with all their might, not only within the University but outside of it as well. I would like to thank my Supervisors David Rowlands, Lee Stoner, and James Faulkner for continually supporting and being patient even as the years dragged on. Each of them imparted their career, research, and life advice guiding me through my growing pains. Outside of my supervisors, help and advice from other professors in the building Sarah Schultz, Sally Lark, Michelle Thunders, Jim Jones, Rachel Page, Jim Clark, to name just a few, helped round my perspective and skills. Beyond Massey University professors like Mike Hamlin, Peter Wagner, Robert Ternansky, John Stanskas, Sheri Lillard, Diana Avila all inspired me with their passion for science and research embedding within me a want to discover. I am very thankful for all the help and support from lab technicians, research assistants, fellow graduate students, and study participants. We spent many many hours in lab together and I will always revere those times and one of the best chapters of my life. Special thanks to Chris Harris, Glen Bell, Terrance Ryan, and Osvaldo Enriquez for your contributions to the custom equipment and analysis software I had to build for this thesis. Finally I would like to thank my awesome, awesome friends and family for their constant love and support. To my friends for celebrating with me when the results were good, and joining my pity parties when p was in a galaxy far far away from 0.05. To my parents Adan and Roxanna Lucero for teaching me the importance of education from day 1, and sacrificing themselves to help me through to the end. And to my husband Logan Smith for carrying me to the finish line. iv Publications Publications arising from This Thesis 1. Lucero, A. A., Addae, G., Lawrence, W., Neway, B., Credeur, D. P., Faulkner, J., ... & Stoner, L. (2018). Reliability of muscle blood flow and oxygen consumption response from exercise using near‐infrared spectroscopy. Experimental physiology, 103(1), 90-100. Other Publications During Thesis 1. Lucero, A. A., Lambrick, D. M., Faulkner, J. A., Fryer, S., Tarrant, M. A., Poudevigne, M., ... & Stoner, L. (2014). Modifiable cardiovascular disease risk factors among indigenous populations. Advances in Preventive Medicine, 2014.. 2. Gaffney K, Lucero AA, Macartney-Coxson D, Clapham J, Whitfield P, Palmer B, Wakefield S, Faulkner J, Stoner L, Rowlands DS. Effects of Whey Protein on Skeletal Muscle Microvascular and Mitochondrial Plasticity Following 10-Weeks of Exercise Training in Men with Type-2 Diabetes. Applied Physiology, Nutrition, and Metabolism. 2021 Feb 16(ja). 3. Trexler ET, Keith DS, Lucero AA, Stoner L, Schwartz TA, Persky AM, Ryan ED, Smith-Ryan AE. Effects of citrulline malate and beetroot juice supplementation on energy metabolism and blood flow during submaximal resistance exercise. Journal of dietary supplements. 2020 Nov 1;17(6):698-717. 4. Black MJ, Lucero AA, Fink PW, Stoner L, Shultz SP, Lark SD, Rowlands DS. The effects of uniquely-processed titanium on balance and walking performance in healthy older adults. Journal of functional biomaterials. 2018 Jun;9(2):39. 5. Gaffney KA, Lucero AA, Stoner L, Faulkner J, Whitfield P, Krebs J, Rowlands DS. Nil Whey Protein Effect on Glycemic Control after Intense Mixed-Mode v Training in Type 2 Diabetes. Medicine and science in sports and exercise. 2018 Jan 1;50(1):11-7. 6. Gaffney KA, Lucero AA, Stoner L, Faulkner J, Whitfield P, Krebs J & Rowlands DS (2018). Nil Whey Protein Effect on Glycemic Control after Intense Mixed- Mode Training in Type 2 Diabetes. Medicine and science in sports and exercise 50, 11-17. 7. Lizamore CA, Stoner L, Lucas S, Lucero AA & Hamlin MJ (2015). Does arterial health affect VO2peak and muscle oxygenation in a sedentary cohort? Medicine and science in sports and exercise 47, 272-279. 8. Witter T, Poudevigne M, Lambrick DM, Faulkner J, Lucero AA, Page R, Perry III LG, Tarrant MA & Stoner L (2015). A conceptual framework for managing modifiable risk factors for cardiovascular diseases in Fiji. Perspectives in public health 135, 75-84. 9. Fryer S, Stoner L, Lucero AA, Witter T, Scarrott C, Dickson T, Cole M & Draper N (2015). Haemodynamic kinetics and intermittent finger flexor performance in rock climbers. International journal of sports medicine 36, 137-142. 10. Fryer S, Stoner L, Scarrott C, Lucero AA, Witter T, Love R, Dickson T & Draper N (2015). Forearm oxygenation and blood flow kinetics during a sustained contraction in multiple ability groups of rock climbers. Journal of sports sciences 33, 518-526. 11. Fryer S, Stoner L, Scarrott C, Lucero AA, Witter T, Love R & Draper N (2014). Oxygen Uptake Or Delivery, Which Is The Limiting Factor For Intermittent Forearm Contractions In Rock-climbers? In Medicine and science in sports and exercise, vol. 46, pp. 756-757. 12. Gaffney K, Lucero AA & Stoner L (2014). Obesity is driving the cardiovascular disease epidemic: however, should obesity be classified as a disease? Journal of atherosclerosis and thrombosis 21, 77-78. vi 13. Stoner L, Lucero AA, Palmer BR, Jones LM, Young JM & Faulkner J (2013). Inflammatory biomarkers for predicting cardiovascular disease. Clinical biochemistry 46, 1353-1371. Refereed Conference Presentations 1. Non-invasive interest group poster award for ACSM conference (USA 2016) 2. Best poster award at SESNZ/MedSci conference (NZ 2015) 3. Co-lead NIRS muscle measures symposium at SESNZ conference, Wellington NZ (2014) vii Table of Contents Table of Contents Abstract................................................................................................................................ii Acknowledgments..............................................................................................................iv Publications..........................................................................................................................v Publications arising from This Thesis.............................................................................v Other Publications During Thesis...................................................................................v Refereed Conference Presentations..............................................................................vii Table of Contents..............................................................................................................viii List of Figures...................................................................................................................xiv List of Tables..................................................................................................................xviii List of Abbreviations........................................................................................................xix 1- Introduction.....................................................................................................................1 Theoretical background...................................................................................................1 Purpose statement & research significance.....................................................................3 Thesis organization.........................................................................................................4 References.......................................................................................................................7 2 - Experimental and Methodological Considerations.........................................................9 Introduction to topic........................................................................................................9 External influences on NIRS signal.............................................................................11 Probe preparation......................................................................................................11 Experimental set-up..................................................................................................12 Internal influences on NIRS signal...........................................................................14 Signal Processing and Analysis.....................................................................................15 viii Conclusion.....................................................................................................................18 References.....................................................................................................................18 3 - Technical Review..........................................................................................................20 Introduction...................................................................................................................20 Theory...........................................................................................................................21 Principles of Spectroscopy.......................................................................................21 Spectroscopy in Tissue.............................................................................................22 Multiple Chromophores in Tissue............................................................................24 Non-Linear Light Travel...........................................................................................24 Non-Homogenous Medium......................................................................................25 Instrumentation.........................................................................................................26 Continuous-wave......................................................................................................27 Time-Resolved NIRS...............................................................................................28 Frequency-Domain NIRS.........................................................................................29 Multi-Channel Devices.............................................................................................30 NIRS Metrics and Methodology..............................................................................31 Tissue Oxygen Saturation and Perfusion..................................................................31 Skeletal Muscle Blood Flow....................................................................................32 Indocyanine Green Method using NIRS..............................................................32 Venous Occlusion Method using NIRS...............................................................33 Skeletal Muscle Oxygen Consumption....................................................................35 Ischaemic Calibration..........................................................................................36 Correction for Blood-Volume Change.................................................................37 Mitochondrial Oxidative Respiratory Capacity........................................................38 Reoxygenation Rate..................................................................................................39 Oxygn delivery to utilization during exercise..........................................................40 ix Technical Limitations....................................................................................................42 Differential Pathlength Factor..................................................................................42 Haemoglobin vs. Myoglobin....................................................................................45 Skeletal mBF and mV̇O2..........................................................................................45 NIRS in Clinical Research............................................................................................47 References.....................................................................................................................49 4 - Reliability of Muscle Blood Flow and Oxygen Consumption Response from Exercise Using Near-infrared Spectroscopy.....................................................................................58 New Findings................................................................................................................59 Abstract.........................................................................................................................60 Introduction...................................................................................................................61 Methods.........................................................................................................................62 Ethical Approval.......................................................................................................62 Experimental Procedures..........................................................................................63 Near-infrared Spectroscopy......................................................................................65 Local Skeletal Muscle Blood Flow..........................................................................65 Skeletal Muscle Oxygen Consumption....................................................................66 Local Skeletal Muscle Perfusion Change and Tissue Saturation Index...................67 Electromyography, Whole Body Oxygen Consumption & Heart Rate....................68 Statistical Analysis....................................................................................................69 Results...........................................................................................................................70 Discussion.....................................................................................................................75 Limitations and Future Direction.............................................................................77 Conclusion.....................................................................................................................78 References.....................................................................................................................79 x 5 - Skeletal Muscle Microvascular and Respiratory Responses to Exercise in Healthy and T2D Adults: A Comparison and Reproducibility Study of NIRS Technologies . . .84 Abstract.........................................................................................................................85 Introduction...................................................................................................................86 Methods.........................................................................................................................87 Participants...............................................................................................................87 Experimental procedures..........................................................................................88 Near-infrared spectroscopy.......................................................................................91 Local skeletal muscle blood flow.............................................................................92 Skeletal muscle oxygen consumption......................................................................93 Skeletal muscle perfusion index and tissue saturation of oxygen............................93 Skeletal muscle blood-volume change during exercise............................................94 Muscle oxygen recovery capacity............................................................................94 Statistics....................................................................................................................96 Results......................................................................................................................97 Skeletal muscle blood flow and oxygen consumption...........................................101 Perfusion index.......................................................................................................102 Tissue saturation of oxygen....................................................................................102 Exercise blood-volume change...............................................................................102 Muscle oxygen recovery capacity (MORC)...........................................................103 Discussion...................................................................................................................104 References...................................................................................................................109 6 - Deoxygenation kinetics During Ramp Cycle Exercise Correlate to Aerobic and Mitochondrial Capacity....................................................................................................115 Adam A. Lucero1, Kim Gaffney1, Lee Stoner1,3, James Faulkner1,2, David S. Rowlands1. ..........................................................................................................................................115 Abstract.......................................................................................................................116 xi Introduction.................................................................................................................117 Methods.......................................................................................................................118 Participants.............................................................................................................118 Experimental Procedure.........................................................................................119 Near infrared spectroscopy.....................................................................................120 Skeletal muscle biopsy...........................................................................................121 Cytochrome C Oxidase and Citrate Synthase activity...........................................121 Mitochondrial assessments.....................................................................................122 Data Analysis..........................................................................................................122 Statistics..................................................................................................................123 Results....................................................................................................................124 Discussion...............................................................................................................130 Limitations..............................................................................................................132 Future Directions....................................................................................................134 Conclusion..............................................................................................................134 References..............................................................................................................136 7 - Conclusion..................................................................................................................140 General Limitations & Recommendations.............................................................144 Conclusion..............................................................................................................149 References...................................................................................................................151 Appendix..........................................................................................................................154 A – Ethics....................................................................................................................154 Full Name of Applicant............................................................................155 General.....................................................................................................156 Project Details..........................................................................................157 xii Participants...............................................................................................161 Data Collection.........................................................................................164 B – Participation Sheet................................................................................................178 Committee Approval Statement....................................................................182 Compensation for Injury...............................................................................182 C – Participant Consent Form.....................................................................................183 D – Recruitment Letter to Nurses...............................................................................184 E – Screening Questionnaire.......................................................................................187 F – Statement of Contribution: DRC – 16..................................................................190 G – Example custom Python code for fd-NIRS and Analysis....................................190 H – Occlusion Device.................................................................................................208 Parts List.................................................................................................................208 Prototype.................................................................................................................209 xiii List of Figures 1. Figure 1.1. Organization of the thesis. 2. Figure 1.2. Progression of thesis. 3. Figure 2.1. Panel A shows a representative example of NIRS signals (µM) collected during experiments showing the raw t[Hb] (green) and [HHb] (blue) traces. The y-axis denotes relative [Hb] (uM) and the x-axis showing time (s). Values obtained from averaging the NIRS signal or collecting the rate of change during an occlusion event from the resting period (seconds 300-900) are defined as baseline and other values are evaluated as a change from rest. Zoom panels B and C show t[Hb] and [Hhb] signals under occlusion events during rest . Zoom panels D and E show t[Hb] and [HHb] signals occlusion events during exercise with three knee extensions on either side of occlusion. Black arrows denote the inflation and deflation of occlusion and the shaded grey area denotes the linear increase used to calculate rate of change. Experimental conditions must be tightly controlled so that changes seen in the NIRS signal traces can be said to confidently reflect haemodynamic changes in the tissue being measured. 4. Figure 2.2. Process of attaching NIRS probe to vastus lateralis. Panel A shows the marked probe location, B shows how the probe was securely taped to skin preventing any side to side or lifting movements during experiment, and C shows the photon shield cover placed over the probe. The blue rapid inflation cuff can be seen upstream of the NIRS probe as high as possible so as to prevent artefact motion from cuff inflation. 5. Figure 2.3. Custom built rapid occlusion device (Cuffinator). Compressed air would enter from the right and be regulated to occlusion pressure, which would rapidly inflate and deflate using green and red buttons, respectively. 6. Figure 2.4. Example experimental set up depicting the participant seated in the Biodex allowing for tight control of seating position and workload. 7. Figure 2.5. Example graph showing the ICC results obtained from processing the same NIRS data using the AC (red) or DC (blue) signal and if constant scattering (cs, solid) or occlusion scattering (os, hatched) affected the AC or DC signal for all NIRS parameters. 8. Figure 3.1. An overview of electromagnetic radiation absorption. Light is emitted form a source and passes through a solution. Upon striking the solution, photons matching the energy gap of molecules suspended within are absorbed resulting in an excitation for the molecule. By comparing the attenuation in the intensity of the transmitted light with its emitted intensity yields the molecular concentration. Figure borrowed from Wikimedia commons. 9. Figure 3.2. Absorption spectra of O2Hb, HHb, and cytochrome oxidase (Cytox). Image borrowed from (M. van Beekvelt et al., 2001) 10. Figure 3.3. Representation of path of NIRS light travel from source to detector. Image borrowed from (Marco Ferrari et al., 2004) xiv 11. Figure 3.4. A scattering medium where the incident light (Io) is scattered on cellular structures and organelles (represented by black dots). Light ray A is scattered traveling the pathlength correction factor times length (L). Light ray B is absorbed completely. Image borrowed from (M. van Beekvelt et al., 2001). 12. Figure 3.5. Schematic showing the three main types of NIRS instruments. d, source- detector separation ( in equations). Phase shift (ɸ) used to determine μa and μ’s. ⍴ Borrowed from (Thomas J. Barstow, 2019) 13. Figure 3.6. Schematic view of a TSI measurement. Light through tissue with three transmitters. Image borrowed from Artinis Manual. 14. Figure 3.7. Phase shift between the incident light (dashed line) and scattered light through tissues (solid lines) at 70 MHz of modulation frequency. Image borrowed from (Yamashita et al., 2013). 15. Figure 3.8. Example trace of NIRS [tHb] (green), [O2Hb] (red), and [HHb] (blue) in response to venous occlusion (V.O.). 16. Figure 3.9. Example trace of NIRS [tHb] (green), [O2Hb] (red), and [HHb] (blue) in response to arterial occlusion (A.O.). 17. Figure 3.10. Muscle oxygenated hemoglobin/myoglobin (O2Hb; as a percentage of the ischemic calibration) during rest, resting arterial occlusions, and a 15-s electrical stimulation exercise followed by a series of transient arterial occlusions after exercise. The final 3–5 min are an ischemic calibration used to determine a relative concentration. 18. Figure 3.11. Scattering calculated from the DC signal of a fd-NIRS device for 692 and 834 nm over an entire experiment (see chapter 5). Assumed constant scattering is overlaid on the calculated scattering signal. Red arrows indicate where occlusions are occurring. 19. Figure 3.12. Average scattering values for 692 (blue) and 834 (orange) nm under each occlusion during experiment. 20. Figure 3.13. Hb parameters calculated under held-constant scattering (darker line) and occlusion specific scattering (lighter line) for venous and arterial occlusion during resting (A, B) and exercise 1 (C, D). The dashed line indicates the occlusion specific scattering overlaid over the constant scattering. 21. Figure 4.1. Experimental Protocol. Representative example of NIRS signals (µM) collected from visits 2-4 showing the raw [tHb] (dark grey) and [HHb] (light grey) traces. The horizontal black lines above the x-axis denote the start and end of each intensity level (%MVC). Panel A shows the timeline (s) for one exercise intensity (5% MVC). After 3 min of knee extension exercise the cuff was rapidly inflated for 5-10 s for 4 venous occlusions (VO; 70-80 mmHg) and 2 arterial occlusions (AO; 250-300 mmHg) with 45 s of knee-extension exercise between occlusions for the assessment of mBF and mV̇O2, respectively. Zoom panels B and C show [tHb] and [HHb] signals for VO and AO, respectively, during rest. Zoom panels D and E show [tHb] and [HHb] signals for VO and AO, respectively, during exercise with three knee extensions on either side of xv occlusion. Black arrows denote the inflation and deflation of occlusion and the shaded gray area denotes the linear increase used in the assessment of mBF or mV̇O2 22. Figure 4.2. The responses of all NIRS parameters over all exercise intensities. to increasing exercise intensity. Panels show (A) mBF, (B) mV̇O2, (C) relative perfusion, and (D) TSI%. Data are means and bars standard deviation. Workloads substantially greater (smallest effect) than resting or the previous workload are denoted with an asterisk (*) or a triangle (Δ), respectively. Statistical likelihoods are given next to the symbol as possible (P, 50-74.9%), likely (L, 75-94.9%), very likely (VL, 95-99.49%) and most likely (ML, 99.5-100%). 23. Figure 4.3. The relationship between mV̇O2 and mBF over all exercise intensities. (A) mBF as a function of mV̇O2 with the regression line denoted by the dashed grey line given by the equation y = 8.071x + 0.5482; R2 = 0.9914. (B) mV̇O2/mBF ratio as a function of exercise intensity. Data are means and bars standard deviation. 24. Figure 4.4. Responses of EMG (top), V̇O2 (middle), and HR (bottom) to increasing exercise intensity. Data are means and bars standard deviation. Workloads substantially greater (smallest effect) than resting or the previous workload are denoted with an asterisk (*) or a triangle (Δ), respectively. Chances are given next to symbol as possible (P, 50-74.9%), likely (L, 75-94.9%), very likely (VL, 95-99.49%) and most likely (ML, 99.5-100%). 25. Figure 5.1. Experimental Protocol. The protocol is conducted using one NIRS device on one leg and is repeated on the other leg with the other NIRS device. The expanded portion of the timeline shows the occlusion timing for an example exercise stage. Abbreviations: MORC, Muscle Oxygen Recovery Capacity; VO, venous occlusion; AO, arterial occlusion; mBF, muscle blood flow; mVO2, muscle oxygen consumption. 26. Figure 5.2. Mean values ± standard deviations for all NIRS parameters for non-diabetic (ND; solid) and type 2 diabetic (T2D; hashed) populations from frequency-domain (FD; grey) and continuous-wave (CW; white) NIRS at rest, 5% and 15%. Mean values are staggered for clarity. The parameters are (A) skeletal muscle blood flow (mBF), (B) oxygen consumption (mV̇O2), (C) total hemoglobin concentration ([tHb]), (D) normalized [tHb], (E) tissue oxygen saturation (SaO2%/TSI%) and (F) exercise blood- volume change ([tHb] range). The symbol # denotes substantial differences between T2D and ND groups and * denotes substantial differences within-groups between-technology. Likelihood of substantial difference between values is denoted as #/*, possibly; ##/**, likely; ###/***, very likely; ####/****, most likely. 27. Figure 6.1. Representative plot for an individual of (A) second-by-second normalized %Δ[HHb] derived from NIRS assessment, and (b) breath-by-breath V̇O2 against absolute external power output during a cycling ramp incremental test. 28. Figure 6.2. Representative profile for O2 extraction (%Δ[HHb]) as a function of systemic V̇O2 for a representative participant. Model fit is a double-linear where the break point (BP) indicates the split of the two segments. xvi 29. Figure 6.2. Representative profile for O2 extraction (%Δ[HHb]) as a function of systemic V̇O2 for a representative participant. Model fit is a double-linear where the break point (BP) indicates the split of the two segments. 30. Figure 6.4. Linear regression for aerobic performance measures of peak power (PP), V̇O2peak, and V̇O2peak/kg vs primary component parameters m1 and b1. Grey shaded area is the 90% confidence interval. xvii List of Tables 1. Table 4.1 Mean values and standard deviations for participant characteristics. 2. Table 4.2. Reliability of mBF and mV̇O2 for rest and all exercise intensities. 3. Table 5.1. Mean values and standard deviations for participant characteristics. 4. Table 5.2. Equations and parameters for mBF and mVO2. 5. Table 5.3. ICC values for all NIRS parameters for non-diabetic (ND) and type 2 diabetic (T2D) groups from frequency-domain (FD) and continuous-wave (CW) NIRS. 6. Table 5.4. The muscle blood flow to oxygen consumption ratio (mBF/mV̇O2) ratio for rest and both exercise intensities for both non-diabetic (ND) and type 2 diabetic (T2D) groups. 7. Table 6.1 Mean values and standard deviations for participant characteristics. 8. Table 6.2. Parameter estimates for double-linear (%Δ[HHb]) model as a function of systemic VO2. 9. Table 6.3. Outcome parameters derived from skeletal muscle biopsy and NIRS evaluation during knee extension. xviii List of Abbreviations Abbreviation Definition AC Amplitude of Intensity Oscillations AO Arterial Occlusion ATT Adipose Tissue Thickness BFI Blood Flow Index BMI Body Mass Index CL Confidence Limits COX Cytochrome Oxidase CS Citrate Synthetase CV Coefficient of Variation Cw-NIRS continuous wave NIRS DC Average Light Intensity DO2 Delivery of Oxygen DPF Differential Path-length Factor ECG Electrocardiogram EMG Electromyography FBG Fasting Blood-Glucose Fd-NIRS frequency domain NIRS xix Hb Hemoglobin HBDif Hemoglobin Difference HHB Deoxygenated Hemoglobin HR Heart Rate ICC Intra-Class Correlation Coefficient ICG Indocyanine Green LT Lactate Threshold MB Myoglobin mBF Skeletal Muscle Blood Flow MORC Skeletal Muscle Oxygen Recovery Capacity MRT Mean Response Time MVC Maximum Voluntary Contraction mV̇O2 Skeletal Muscle Blood Flow NIRS Near-infrared Spectroscopy O2Hb Oxygenated Hemoglobin O2MB Oxygenated Myoglobin PAD Peripheral Arterial Disease PP Peak Power Qcap Capillary Blood Flow RF Rectus Femoris ROI Region of Interest xx xxi SD Standard Deviation STE Standardized Typical Error StO2 Tissue Oxygen Saturation T2D Type 2 Diabetes tc Time Constant TD Time Delay tHb Total Hemoglobin TOF Time of Flight TSI Tissue Saturation Index VL Vastus Lateralis VM Vastus Medialis VO Venous Occlusion V̇O2 Systemic/Whole Body Oxygen Consumption V̇O2max Maximal Whole-body Oxygen Consumption xxii 1- Introduction Theoretical background Skeletal muscles constitute 40-50% of total body weight enabling individuals to move, breathe, and do work. With respect to exercise, skeletal muscle must cope with a large range of activities from postural stability and basic locomotion to immediate explosive movements in response to unexpected environmental perturbation. Maintaining skeletal muscle function requires the integration of multiple protein, transporter, and organelle systems for dynamic control of energy metabolism. The energy systems are ultimately reliant upon oxygen-dependent respiration within the mitochondria, which adapt to providing energy for long periods of time at low to moderate output rates, to instantaneously increasing energy production more than 100-fold in response to explosive contractions (Westerblad et al., 2010). Skeletal muscle is also a primary tissue for nutrient disposal and utilization, coupled to its energy-strong role in locomotion and posture. Sustaining physical activity in skeletal muscle tissue is dependent upon the oxygen cascade system of processes including pulmonary ventilation, cardiac output, circulation, capillary microperfusion, uptake, transfer and respiratory energy production. Dynamic regulation of these processes is critical in upholding the physiological integrity of the body during sustained activity and fulfilling requirements of daily living. Dysregulation of several of the circulatory, capillary endothelial and basement membrane, and metabolic properties within skeletal muscle have been associated with metabolic diseases including type II diabetes (Bauer et al., 2007a; Hayden et al., 2005), generally attributed to the adverse effects of hyperglycemia and oxidative stress on vascular biology (Avogaro et al., 2011). Near-Infrared Spectroscopy (NIRS) has been used to non-invasively investigate local muscle oxidative metabolism at rest, and recently during exercise (M. Ferrari et al., 2011; M. Van Beekvelt et al., 2002). With high temporal resolution and non-invasive properties, NIRS can be used within a wider range of exercise modalities than invasive 1 counterparts. Much research using NIRS has been focused on brain microperfusion, however there is growing popularity of NIRS research on skeletal muscle. Coupling NIRS with rapid arterial and venous occlusions has permitted the easy measurement of skeletal muscle blood flow (mBF) and oxygen consumption (mV̇O2) (Cross & Sabapathy, 2017a; T. Ryan et al., 2012; M. van Beekvelt et al., 2001). By conducting a set number of rapid arterial occasions after a brief bout of exercise, the resultant mV̇O2 over time can be plotted resulting in a curve showing the post exercise mV̇O2 recovery rate. This method provides an assay of the muscle oxygen respiratory capacity (MORC), which is a hybrid index incorporating delivery, deoxygenation, and mitochondrial respiratory functional kinetics (T. Ryan et al., 2012). These measures can be combined with perfusion and saturation of oxygen (SaO2) derived from the NIRS signal to obtain an integrated picture of the combined systems maintaining skeletal muscle microvascular function. In a novel approach described in this thesis, I compared resting measures to measures taken at incremental levels of exercise intensity for investigation of the dynamic coupling of O2 delivery to O2 uptake adapting from rest to exercise. I show that NIRS can be used to investigate skeletal muscle health in metabolically diseased populations at rest, in association with insulin stimulation, and during controlled muscle contraction, and guide in treatment and intervention options. Since the pioneering work of Jobsis in the 1980s (Jobsis, 1977), various NIRS technologies have been developed to measure the concentration of oxygenated and deoxygenated hemoglobin (O2Hb and HHb) in skeletal muscle and brain tissue. Continuous-wave NIRS (cw-NIRS) has emerged as one of the preferred technologies due to its ease of use, affordability, and portability. However, cw-NIRS utilizes assumptions that may not hold under certain situations or for specific populations. Frequency-domain NIRS utilizes more advanced methods to determine hemoglobin concentrations, however it is more complex, more expensive, and less portable than cw-NIRS. Comparing and contrasting metrics derived with fd-NIRS to those of cw-NIRS will further our understanding of the assumptions of cw-NIRS allowing for confidence in its use under various modalities for differing cohorts. 2 Purpose statement & research significance The purpose of the current thesis is to evaluate the reproducibility and clinical utility of cw-NIRS for the purposes of evaluating skeletal muscle performance in health and metabolic disease tissue (type-2 diabetes, T2DM). I achieved this by: 1) Designing a comprehensive test to evaluate mBF, mVO2, perfusion, saturation, and MORC at rest and during exercise and establish the day-to-day reliability of this test in healthy individuals and compare and contrast the NIRS derived parameters to the same parameters derived from other methodologies in the literature. 2) Repeat this protocol in T2D individuals with the addition of a benchmark NIRS technology frequency-domain NIRS to test if the assumptions made in cw-NIRS are valid at exercise, and if they are affected by the diseased state of the muscle. 3) Investigate combining NIRS with systemic oxygen uptake to further examine the coupling of O2 delivery and consumption in T2Ds during incremental ramp cycle exercise and correlate the modeled profile to metrics derived from in vitro skeletal muscle biopsy to help explain the increases from rest to exercise. Skeletal muscle is the largest insulin sensitive tissue in the human body and is an important site for glucose handling and disposal. In T2D, skeletal muscle is characterised by decreased muscle mass (Guerrero et al., 2016), decreased blood flow, particularly at the microvascular level (L.S et al., 2008), and decreased mitochondrial respiratory capacity (Kelley et al., 2002). Chronic exercise has been shown to improve these abnormalities in skeletal muscle and improve metabolic function in T2D (Colberg et al., 2002)(Stanford et al., 2014). However, there is currently no consensus on a critical point of impairment or restoration in insulin resistant muscle. A comprehensive exploration of skeletal muscle adaptation in response to exercise therapy is required to establish important targets for rehabilitation. This thesis will investigate the ability of NIRS to meet this need, at least from the perspective of microvascular haemodynamics. 3 Thesis organization This thesis comprises 7 chapters with four primary sections. Chapters 4-6 are written as stand-alone chapters that incorporate standard paper format (abstract, introduction, methodology, results, discussion) and are specific to the aims of that chapter. The thesis is organized according to the flowcharts below (Figures 1.1 and 1.2). The Thesis is introduced in chapter 1 with chapters 2 and 3 providing a review of NIRS use in skeletal muscle and methodological considerations. The primary research studies are described in chapters 4-6. The study described in chapter 4 describes the reliability of muscle blood flow and oxygen consumption response from exercise using NIRS and has been published in the Journal of Experimental Physiology. The study described in chapter 5 provides a novel and important description of the comparative reproducibility and utility of cw-NIRS validated against fd-NIRS in healthy and T2Ds. This study is currently under review in the Journal of Applied Physiology. Finally, the study described in chapter 6 provides a novel characterization of the temporal profile of [HHb] during incremental ramp cycle exercise in T2Ds and a comparison of modeling kinetic parameters to skeletal muscle biopsied tissue, affirming interpretations of previous research of modeling the dynamic coupling of O2 delivery, [HHb] deflection thresholds, and utilization during incremental exercise to exhaustion. 4 Figure 1.1. Organization of the thesis. 5 Figure 1.2. Progression of thesis. 6 References Avogaro, A., Albiero, M., Menegazzo, L., Kreutzenberg, S. de, & Fadini, G. P. (2011). Endothelial Dysfunction in Diabetes. Diabetes Care, 34(Supplement 2), S285–S290. https://doi.org/10.2337/DC11-S239 Bauer, T. A., Reusch, J. E. B., Levi, M., & Regensteiner, J. G. (2007). Skeletal Muscle Deoxygenation After the Onset of Moderate Exercise Suggests Slowed Microvascular Blood Flow Kinetics in Type 2 Diabetes. Diabetes Care, 30(11), 2880–2885. https://doi.org/10.2337/dc07-0843 Colberg, S. R., Stansberry, K. B., McNitt, P. M., & Vinik, A. I. (2002). Chronic exercise is associated with enhanced cutaneous blood flow in Type 2 diabetes. Journal of Diabetes and Its Complications, 16(2), 139–145. https://doi.org/10.1016/S1056- 8727(01)00222-7 Cross, T. J., & Sabapathy, S. (2017). The impact of venous occlusion per se on forearm muscle blood flow: implications for the near-infrared spectroscopy venous occlusion technique. Clinical Physiology and Functional Imaging, 37(3), 293–298. https://doi.org/10.1111/cpf.12301 Ferrari, M., Muthalib, M., & Quaresima, V. (2011). The use of near-infrared spectroscopy in understanding skeletal muscle physiology: recent developments. Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences, 369(1955), 4577–4590. https://doi.org/10.1098/rsta.2011.0230 Guerrero, N., Bunout, D., Hirsch, S., Barrera, G., Leiva, L., Henríquez, S., & De la Maza, M. P. (2016). Premature loss of muscle mass and function in type 2 diabetes. Diabetes Research and Clinical Practice, 117, 32–38. https://doi.org/10.1016/J.DIABRES.2016.04.011 Hayden, M. R., Sowers, J. R., & Tyagi, S. C. (2005). The central role of vascular extracellular matrix and basement membrane remodeling in metabolic syndrome and type 2 diabetes: the matrix preloaded. Cardiovascular Diabetology 2005 4:1, 4(1), 1–20. https://doi.org/10.1186/1475-2840-4-9 Jobsis, F. (1977). Noninvasive, infrared monitoring of cerebral and myocardial oxygen sufficiency and circulatory parameters. Science, 198(4323), 1264–1267. https://doi.org/10.1126/SCIENCE.929199 7 Kelley, D. E., He, J., Menshikova, E. V., & Ritov, V. B. (2002). Dysfunction of Mitochondria in Human Skeletal Muscle in Type 2 Diabetes. Diabetes, 51(10), 2944–2950. https://doi.org/10.2337/DIABETES.51.10.2944 Ryan, T. E., Erickson, M. L., Brizendine, J. T., Young, H.-J., & McCully, K. K. (2012). Noninvasive evaluation of skeletal muscle mitochondrial capacity with near-infrared spectroscopy: correcting for blood volume changes. Journal of Applied Physiology, 113(2), 175–183. https://doi.org/10.1152/japplphysiol.00319.2012 S, L., S, G., PL, H., & JC, B. (2008). Reduced leg blood flow during submaximal exercise in type 2 diabetes. Medicine and Science in Sports and Exercise, 40(4), 612–617. https://doi.org/10.1249/MSS.0B013E318161AA99 Stanford, K. I., & Goodyear, L. J. (2014). Exercise and type 2 diabetes: molecular mechanisms regulating glucose uptake in skeletal muscle. Https://Doi.Org/10.1152/Advan.00080.2014, 38(4), 308–314. https://doi.org/10.1152/ADVAN.00080.2014 Van Beekvelt, M. C. P., Van Engelen, B. G. M., Wevers, R. A., & Colier, W. N. J. M. (2002). In vivo quantitative near-infrared spectroscopy in skeletal muscle during incremental isometric handgrip exercise. Clinical Physiology and Functional Imaging, 22(3), 210–217. https://doi.org/10.1046/j.1475-097X.2002.00420.x van Beekvelt, M., Colier, W. N. J. M., Wevers, R. O. N. A., Engelen, B. G. M. V. A. N., Wevers, R. A., & Perfor-, B. G. M. V. E. (2001). Performance of near-infrared spectroscopy in measuring local O 2 consumption and blood flow in skeletal muscle. 511–519. Westerblad, H., Bruton, J. D., & Katz, A. (2010). Skeletal muscle: Energy metabolism, fiber types, fatigue and adaptability. Experimental Cell Research, 316(18), 3093–3099. https://doi.org/10.1016/J.YEXCR.2010.05.019 8 2 - Experimental and Methodological Considerations Introduction to topic Since its inception in 1980 by Jobsis (Jobsis, 1977), near-infrared spectroscopy (NIRS) has exponentially increasing in its ability to non-invasively determine concentrations of oxygenated and deoxygenated hemoglobin in tissues 2-4 cm under the skin. While the majority of the research and strides have been made in neural monitoring (Cour et al., 2018), much work has been done on NIRS use in skeletal muscle at rest and during exercise (Grassi & Quaresima, 2016a). Although NIRS has much potential and promise in clinical applications, there exists many obstacles and limitations in experimental standardizations, signal processing equation coefficients, analysis parameters and processes, and interpretation of results. Before NIRS can be beneficial clinically more understanding and research is needed to more thoroughly develop our light transport models in biological tissue and characterize NIRS signals in various populations of all ages. There exist three main considerations to NIRS that affect confidence in the accuracy of NIRS derived concentrations of oxy and deoxy haemoglobin: 1. External influences on NIRS signal 2. Internal influences on NIRS signal 3. Signal processing and analysis 9 An example NIRS output is given in figure 2.1. For NIRS devices that can calculate absolute [Hb] these values can be reported as an average for a given period. For other devices where the [Hb] is arbitrary, an average value from a resting baseline is obtained and other measures are reported as a change from that baseline. For measures such as blood flow and oxygen consumption, they are calculated from the slope of the NIRS signal giving a rate of change. Each of the three points mentioned above can contribute substantially to the outputted NIRS signal, leading to inaccurate results. Thorough and detailed understanding of how each of these points affects the accuracy of NIRS derived measures under differing experimental conditions and populations is the ultimate goal of the current thesis. Figure 2.1. Panel A shows a representative example of NIRS signals (µM) collected during experiments showing the raw t[Hb] (green) and [HHb] (blue) traces. The y-axis denotes relative [Hb] (uM) and the x-axis showing time (s). Values obtained from averaging the NIRS signal or collecting the rate of change during an occlusion event from the resting period (seconds 300-900) are defined as baseline and other values are evaluated as a change from rest. Zoom panels B and C show t[Hb] and [Hhb] signals under occlusion events during rest. Zoom panels D and E show t[Hb] and [HHb] signals 10 occlusion events during exercise with three knee extensions on either side of occlusion. Black arrows denote the inflation and deflation of occlusion and the shaded grey area denotes the linear increase used to calculate rate of change. Experimental conditions must be tightly controlled so that changes seen in the NIRS signal traces can be said to confidently reflect haemodynamic changes in the tissue being measured. External influences on NIRS signal Probe preparation Regardless of technology used, NIRS probes all have a light source that emits the near-infrared light that passes through the tissue to be detected by a sensor placed 2-5 cm linearly from the light source. Environmental light can also be detected by the sensor, and for this reason it is recommended that the probe be taped directly to the skin so as to have no gap between probe and skin, and no movement in position during the experiment. In addition, a dark bandage or wrapping is commonly placed over the NIRS probe to prevent ambient light from hitting the probe. During pilot trials I identified that the wrapping was pushing the NIRS probe into the skin during contraction movements changing the region of tissue being measured and therefore the NIRS signals being output (T. Hamaoka et al., 2011). For this reason I made a custom ‘photon shield’ that encapsulated the probe allowing it to move freely with the skin during contractions while blocking out all ambient light. Figure 2.2 shows how the probe location was marked, attaching the probe to the skin with double sided tape and securely adhering the edges using strips of cross and lengthwise tape. Finally the photon shield was placed over the probe leaving room on all sides. This process yielded the lowest signal-to-noise ratio during pilot testing. Finally, a rapid-inflation cuff (Hokanson SC 10D, D. E. Hokanson, Inc., Bellevue, WA) used with a custom-built rapid inflation device was used to manipulate blood flow for mBF and mVO2 measures. This cuff was placed upstream of the NIRS probe as high as possible putting as much space as possible between the cuff and probe because in some participants the act of cuff inflation could swell the thigh artificially moving the probe and changing the area being measured. The participant was 11 familiarized with the occlusions at each session to prevent involuntary jerking motions caused by the shock of the inflation. Figure 2.2. Process of attaching NIRS probe to vastus lateralis. Panel A shows the marked probe location, B shows how the probe was securely taped to skin preventing any side to side or lifting movements during experiment, and C shows the photon shield cover placed over the probe. The blue rapid inflation cuff can be seen upstream of the NIRS probe as high as possible so as to prevent artefact motion from cuff inflation. Figure 2.3. Custom built rapid occlusion device (Cuffinator). Compressed air would enter from the right and be regulated to occlusion pressure, which would rapidly inflate and deflate using green and red buttons, respectively. Experimental set-up Other external influences including seating position, height, angle of motion and workload needed to be standardized so as to keep the same conditions day to day. Studies using NIRS to assess mV̇O2 in response to exercise have shown that using resistance 12 bands or electrical stimulation yielded consistent results (Terence E Ryan et al., 2013). However, in the current study we measured blood flow in addition to mV̇O2 and found that mBF measures were much more sensitive. In order to ensure reliable results we needed to tightly control seating position and workload. Participants were seated on an isokinetic dynamometer in the same experimental position (Biodex Medical Systems, Inc. Shirly, NY, USA) reclined to 70° to obtain a 110° hip angle and a 90˚ knee extension angle, which was optimal for experimental measures (Grassi & Quaresima, 2016a). The settings were adjusted so that the axial portion of the knee aligned with the axis of rotation on the dynamometer. The non-working leg was suspended in a 150º knee-joint angle throughout to limit changes in blood distribution between legs. Isotonic knee extension exercise was chosen as the resistance arm applied a specific force (N·m) only at the 90º angle then released once the force was reach allowing the participant to freely swing their leg up the rest of the way standardizing the amount and duration force was applied on the working muscle. Other modes of exercise allowed participants to vary how much force they applied, changing the outcome variables. Figure 2.4 shows an example set-up. 13 Figure 2.4. Example experimental set up depicting the participant seated in the Biodex alowing for tight control of seating position and workload. Internal influences on NIRS signal Internal influences on NIRS signal include adipose tissue, skin temperature, skin blood flow, and probe placement limiting muscle cross talk. Skeletal muscle tissue makeup can change the measured [Hb] concentrations using NIRS, and it is a goal of this thesis to identify if skeletal muscle differences in healthy and T2D show any changes in measured NIRS parameters, therefore all other influences must be minimized. Adipose tissue thickness (ATT) is known to alter optical properties of NIRS (Pirovano et al., 2021), and in the current thesis all participants ATT was measured using B-mode ultrasound and participants were excluded if ATT exceeded 2 cm since this is greater than the penetration of the NIRS light. Monte-Carlo simulations and models have been suggested to correct NIRS signals for ATT (Soller et al., 2005), however this was not applied to the current study as we did not measure significant changes in ATT day-to-day and used ischaemic calibration to correct NIRS signals where applicable. It is known that skin blood flow increases which can be brought on by an increase in skin temperature can significantly contribute to the O2Hb signal, mainly affecting blood flow measures (Buono et al., 2005). To minimize this, The temperature in the lab was constantly maintained at 14 20.5 ˚C, SD 0.8, and the experiment was designed so that the mBF measures were done prior to the more intensive and fatigue inducing mV̇O2 and MORC measures, saving the incremental cycling for last since this measure only used the HHb signal which is not significantly altered by changes in skin blood flow. Finally, it was imperative that a standard way of marking the probe placement on each individual was employed so that the NIRS was placed over the belly of vastus lateralis (VL) muscle, maximizing signal response and minimizing cross talk between muscle groups, especially during contractions (Iwasaki & Okada, 2004). In addition, the same spot needed to be located day-to-day, even weeks later from the initial test as changing the NIRS probe position changes the underlying tissue being measured which will introduce changes in the NIRS signals not related to the experiment. A standard method to locate the belly of the VL on individuals for use in electromyography was employed, accompanied with a custom template that allowed for ventral or dorsal changes in probe placement so that the probe could be placed on the flattest portion of the VL belly during isometric contraction limiting probe movement during contraction. Signal Processing and Analysis In the case of both NIRS devices, continuous wave (cw-) and frequency domain (fd-) NIRS, the raw data received from the detector can be processed using various equations and constants. How each approach affects the [Hb] measures and when to apply what method is still not fully understood. For cw-NIRS we followed recommendations in the literature for choosing default parameters. In the case of the fd- NIRS I wrote custom analysis software processing the signals using many different combinations of input variables following the literature for recommendations (see figure 2.5). Developing custom analysis software in python was critical to determining the optimal approach for calculation of [Hb], apart from how to incorporate real-time measures of scattering. . Figure 3.11 shows that the scattering changes abruptly during occlusion and applying real time scattering seemed to affect the NIRS derived [Hb] because the sharp decrease in scattering was altering the concentration. Two methods of 15 applying scattering were trialed, the first was to calculate the degree of scattering in the individual after a period of rest and hold it constant throughout the experiment, or apply a scattering value measured during occlusion to each measure. Ultimately, I decided to apply a resting scattering value to the whole experiment as this would allow us to evaluate the effect of assuming literature-based scattering coefficient for a cohort vs calculating a scattering coefficient for each individual in an experiment. To determine reliability and reproducibility of the NIRS derived measures, methods developed by Will Hopkins were employed as they were developed specifically for exercise performance testing and accounted for the effects of habituation (such as familiarization, practice, motivation, fatigue, or even the training effect of a single test) and allowing for the determination of the smallest important mechanistic difference in measures. This was important to determine the acceptable variance between measures and between NIRS technologies so as to be confident that NIRS outcomes can be attributed to muscle mechanics and not measurement variability. 16 Figure 2.5. Example graph showing the ICC results obtained from processing the same NIRS data using the AC (red) or DC (blue) signal and if constant scattering (cs, solid) or occlusion scattering (os, hatched) affected the AC or DC signal for all NIRS parameters. 17 Conclusion Due to the sensitive and experimental nature of NIRS, significant work went into piloting different aspects of the experimental design in order to develop a standardized protocol that could be built upon to enhance the ability of NIRS to non-invasively investigate skeletal muscle hemodynamic function. Achieving day-to-day reliability and correcting for external factors are critical steps in the development of NIRS for practical use in exercise and clinical physiology. It is my hope that this thesis contributes to the overall goal of development of non-invasive methodologies for clinical utility. References Buono, M. J., Miller, P. W., Hom, C., Pozos, R. S., & Kolkhorst, F. W. (2005). Skin Blood Flow Affects In Vivo Near-Infrared Spectroscopy Measurements in Human Skeletal Muscle. The Japanese Journal of Physiology, 55(4), 241–244. https://doi.org/10.2170/jjphysiol.T649 Cour, A. la, Greisen, G., & Hyttel-Sørensen, S. (2018). In vivo validation of cerebral near-infrared spectroscopy: a review. Neurophotonics, 5(4), 040901. https://doi.org/10.1117/1.NPH.5.4.040901 Grassi, B., & Quaresima, V. (2016). Near-infrared spectroscopy and skeletal muscle oxidative function in vivo in health and disease: a review from an exercise physiology perspective. Journal of Biomedical Optics, 21(9), 91313. Hamaoka, T., McCully, K. K., Niwayama, M., & Chance, B. (2011). The use of muscle near-infrared spectroscopy in sport, health and medical sciences: recent developments. Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences, 369(1955), 4591–4604. https://doi.org/10.1098/rsta.2011.0298 Iwasaki, A., & Okada, E. (2004). Influence of cross talk on near-infrared oxygenation monitoring of muscle. Second Asian and Pacific Rim Symposium on Biophotonics - Proceedings, APBP 2004, 159–160. https://doi.org/10.1109/apbp.2004.1412330 Jobsis, F. (1977). Noninvasive, infrared monitoring of cerebral and myocardial oxygen sufficiency and circulatory parameters. Science, 198(4323), 1264–1267. https://doi.org/10.1126/SCIENCE.929199 18 Pirovano, I., Porcelli, S., Re, R., Spinelli, L., Contini, D., Marzorati, M., & Torricelli, A. (2021). Effect of adipose tissue thickness and tissue optical properties on the differential pathlength factor estimation for NIRS studies on human skeletal muscle. Biomedical Optics Express, 12(1), 571. https://doi.org/10.1364/boe.412447 Ryan, T. E., Southern, W. M., Reynolds, M. A., & McCully, K. K. (2013). A cross- validation of near-infrared spectroscopy measurements of skeletal muscle oxidative capacity with phosphorus magnetic resonance spectroscopy. Journal of Applied Physiology, 115(12), 1757–1766. https://doi.org/10.1152/japplphysiol.00835.2013 Soller, B. R., Landry, M. R., Soyemi, O. O., & Yang, Y. (2005). Influence of a fat layer on the near infrared spectra of human muscle: quantitative analysis based on two- layered Monte Carlo simulations and phantom experiments. Optics Express, Vol. 13, Issue 5, Pp. 1570-1579, 13(5), 1570–1579. https://doi.org/10.1364/OPEX.13.001573 19 3 - Technical Review Introduction The purpose of this review is to highlight technical considerations of Near infrared spectroscopy (NIRS) to assess skeletal muscle hemodynamic parameters. This technology utilizes the oxygen dependent absorption characteristics of infrared light in the 600 – 900 nm range to provide information about the changes in tissue oxygenation. A major benefit of NIRS is that its non-invasive allowing for use during exercise, and can be used outside of the laboratory for unprecedented investigation of matching oxygen delivery (DO2) to oxygen utilization (mVO2) at rest or during exercise. This tool has the potential to further our understanding of perfusion, blood flow, and skeletal muscle respiratory dynamics, in both health and pathophysiology. By the end of this review it is hoped that the reader will be familiar with the background and operation of NIRS, be aware of its current limitations from the perspective of skeletal muscle, and understand the gaps in knowledge that need to be characterized so NIRS can be utilized in a clinical setting. The ability of NIRS to differentiate between and measure concentrations of the oxygenated and deoxygenated forms of hemoglobin and myoglobin (O2Hb/Mb and HHb/Mb, respectively), which added together yields total hemoglobin concentration (tHb/Mb = O2Hb/Mb + Hhb/Mb) allows for measure of skeletal muscle blood flow (De Blasi et al., 1994; Nioka et al., 2006), muscle oxygenation, and muscle oxygen consumption (Takafumi Hamaoka et al., 1996; Malagoni et al., 2010; Takayuki Sako et al., 2001; M. van Beekvelt et al., 2001). More recently, repeated arterial occlusions have been used after exercise to assess the recovery of muscle oxygen consumption (mVO2) as an index of mitochondrial function (Motobe et al., 2004; T.E. Ryan et al., 2012). Finally, the profile of HHb during ramp incremental exercise is said to reflect the dynamic balance between oxygen delivery and utilization (Boone et al., 2009; Spencer et al., 2012). Taken together NIRS is a powerful, non-invasive tool for investigating skeletal muscle hemodynamics at rest, transition to, and during exercise. However, NIRS research is complicated by multiple technologies, devices withing technologies with slight variations in function, tissue makeup changing contribution of skeletal muscle to NIRS signal individual to individual and even measurement to measurement within an 20 individual (e.g., adipose tissue thickness, skin blood flow, melanin), and differing techniques for probe placement and experimental design make interpreting NIRS results within a study and comparing studies to one another difficult. This review will briefly cover NIRS theory, discuss different NIRS technologies and how they are employed, introduce common NIRS experimental methods, technical limitations to these methods, and finally the use of NIRS in clinical studies. Theory Principles of Spectroscopy The principles of NIRS are based on the principles of traditional spectroscopy, modified to work for biological tissue. Spectroscopy is a technique for measuring concentration of a molecular compound suspended in a homogenous solution. It is based on the Beer-Lambert law which expresses the linear relationship between the absorbance and concentration of a compound at a fixed wavelength, given by: Eq 3.1: ODλ = log(Io/I) = eλ * c * l Where ODλ is a dimensionless factor known as the optical density of the medium, Io is the initial light intensity, I is the measured light intensity at the detector, eλ the extinction coefficient of the chromophore (mM-1*cm-1), c is the concentration of the compound, and l is the distance between the point of light emission (source) and final measurement (detector). As light is emitted from the source and travels through solution, some of the light will be absorbed by the chromophore in the molecular compound. The more molecular compound in solution the more of the emitted light that is absorbed and the less light that makes it to the detector. Therefore, by knowing the incident light (Io), detected light (I), source-detector distance (l) and optical properties of the light and molecular compound of interest, the molecular concentration can be calculated. This method has some requirements, namely that the molecular compound of interest has a light absorbing 21 chromophore, the emitted light wavelength is specific for that chromophore, there are no other molecules in solution that have chromophores absorbing light at the same wavelength, and that the light is traveling linearly and through a homogenous, non scattering medium so that the difference in the light intensity from source to detector is completely due to absorption by the molecular compound of interest. Figure 3.1. An overview of electromagnetic radiation absorption. Light is emitted form a source and passes through a solution. Upon striking the solution, photons matching the energy gap of molecules suspended within are absorbed resulting in an excitation for the molecule. By comparing the attenuation in the intensity of the transmitted light with its emitted intensity yields the molecular concentration. Figure borrowed from Wikimedia commons. Spectroscopy in Tissue Biological tissue presents many challenges for traditional spectroscopy. The wavelength of light used must be able to penetrate several centimeters of epidermal, adipose, and skeletal muscle tissue while being specific to a molecular compound of interest within the tissue. Wavelengths of light below 300 nm and above 1000 nm are completely absorbed by water molecules and not usable. From 450 nm - 650 nm (visible spectrum) light is also completely absorbed by hemoglobin and other chromophores. In the near-infrared spectrum, specifically between 700-1000 nm absorption of light is sufficiently low enough for light to be detected across many centimeters of tissue, and 22 within range of absorption spectra peaks for oxygenated and deoxygenated forms of hemoglobin, the main oxygen transporter in the blood. Moreover, it is thought that the [Hb] measured by NIRS comes mainly from blood vessels <1mm in diameter as vessels larger than this completely absorb the NIR light (Experimental study of migration depth for the photons measured at sample surface). Apart from wavelength of light used, spectroscopic techniques in biological tissue present many challenges, including multiple absorbing chromophores in the solution being measured, non-linear light travel, and that it is a non-homogenous scattering medium. The attenuation of light from source to detector is not only due to absorption by molecule of interest, but also absorption from chromophores of fixed concentration, absorption from chromophores of variable concentration, and light scattering. Differing NIRS technologies and apparatuses have been developed to account for these challenges, which adds complexity in accuracy and interpretation of results. This review will discuss these challenges with respect to three of the more popular NIRS technologies. Figure 3.2. Absorption spectra of O2Hb, HHb, and cytochrome oxidase (COX). Image borrowed from (M. van Beekvelt et al., 2001) 23 Multiple Chromophores in Tissue In the near-infrared region there are three main absorbing oxygen dependent chromophores, hemoglobin, myoglobin, and cytochrome oxidase. Due to different absorption spectra, the concentration of the oxygenated form of hemoglobin (O2Hb) can be measured separately from the deoxygenated form (HHb) by using wavelengths ~830 and ~750 nm, respectively. The degree of contribution to the calculated concentration by Mb and Cyt Ox is still being researched, but cyt Ox is thought to contribute negligible amounts of absorption while Mb is thought to vary between 10-50% (Davis & Barstow, 2013). The contribution of Mb to the NIRS light attenuation is controversial, but is thought to be constant and therefore any change in concentration can be attributed to changes in O2Hb and HHb. Non-Linear Light Travel In traditional spectroscopy the solution is placed in a direct line between light emitter and detector. With NIRS this is not possible as the light would be completely absorbed by tissue and bone leaving no light to measure at the detector on the other side. Instead, the light emitting probe and detector are both facing downward securely attached to the skin 2-4.5 cm apart from each other (see figure 3.3). In this way, the light is emitted in all directions, and a percentage makes its way to the detector in a ‘banana’ shape that penetrates ½ the source-detector distance with light intensity extending on either side (Marco Ferrari et al., 2004). Figure 3.3. Representation of path of NIRS light travel from source to detector. Image borrowed from (Marco Ferrari et al., 2004) 24 Non-Homogenous Medium In addition to added travel distance due to banana shape, the differing refractive index of skin, adipose, tendon, and skeletal muscle tissue and their membranes and organelles means that the incident light does not travel in a perfect arch from source to detector but rather is scattered as it is forced to change direction and bounce off of obstacles as it travels through the cellular membrane structures of differing refractive indices. In the absence of scatter the total light absorption in the medium is a linear sum of each chromophore concentration, but in a scattering medium this linear summation is distorted by the differing optical pathlength of light travel, which may differ by wavelength and tissue makeup. The distance or length of light travel must be known to calculate molecular concentration and since the travel is non-linear and scattered, this distance must be calculated using a complex function of the tissue absorption and scattering coefficients (μs and μa), scattering phase function, and tissue measurement geometry. Measuring changes in concentration of O2Hb and HHb became possible with the development of techniques to measure optical pathlength in tissue, and combined with literature based or individually sampled hemoglobin concentration from blood samples allowed for estimates of absolute concentration. Figure 3.4. A scattering medium where the incident light (Io) is scattered on cellular structures and organelles (represented by black dots). Light ray A is scattered traveling the pathlength correction factor times length (L). Light ray B is absorbed completely. Image borrowed from (M. van Beekvelt et al., 2001). 25 Instrumentation To obtain an accurate high-resolution absolute [O2Hb] and [HHb] at rest and in response to exercise requires measuring more than just light attenuation at the tissue surface. Clever manipulations of light combined with highly sensitive light emitting and detecting equipment have allowed researchers to develop methods to quantify the degree of absorption and scattering of light (μs and μa) in the tissue being measured in real time. The three main technologies used to measure skeletal muscle oxygenation are continuous-wave (cw), time-resolved (TRS), and frequency-domain (fd; phase- modulated) NIRS (see figure 3.5). These technologies include utilization of a literature based correction factor as in continuous-wave NIRS, measuring the temporal dispersion of light from an ultrashort input light pulse, as in time-of-flight NIRS, or phase and modulation depth changes of light, as in frequency-domain NIRS. While these approaches have been successful in accurately determining hemoglobin concentrations and skeletal muscle optical properties, they are limited by the accuracy of the light transport models in inhomogeneous media, which is further complicated by sampling during exercise since the contraction of the skeletal muscle changes the volume and makeup of tissue being measured (Thomas J. Barstow, 2019). Figure 3.5. Schematic showing the three main types of NIRS instruments. d, source- detector separation ( in equations). Phase shift (ɸ) used to determine μa and μ’s. ⍴ Borrowed from (Thomas J. Barstow, 2019) 26 Continuous-wave The most common, simplest to understand, easiest to use, and most portable NIRS technology is cw spatially resolved spectroscopy. The main advantages of cw-NIRS over the other two technologies affecting cost and ease of use are the ability to use LEDs vs laser diodes and no need for regular calibration. This technology utilizes a modified version of the Beer-Lambert law: eq 3.1: ODλ = log(Io/I) = eλ * c * l * DPF Where DPF is a dimensionless factor that accounts for the light attenuation contributions of absorption and scattering so that heme concentrations can be calculated. The DPF literature values at a given wavelength were obtained by time-of-flight measurements, which uses an ultra-short laser pulse fired into the tissue that is detected by an ultrafast camera to track the time from light pulse emission to detection giving the time to travel through tissue. From this time travel measure, the distance traveled can be calculated using the speed of light. Dividing the distance travelled by the source-detector separation distance (ρ) yields the DPF (eq. 3.2). In addition, while the underlying theory is different, the DPF for a given wavelength can also be approximated as a function of ρ, μs and μa (eq. 3.3). eq 3.2: DPF (λ) dρ = c * t(≅ λ)n(λ) eq 3.3: DPF (λ) ≈ √3µ′s / 2√µa Where c is the speed of light in a vacuum, n(λ) is the tissue refractive index at wavelength, and t(λ) is the photon mean time-of-flight (Essenpreis et al., 1993). Assuming a constant DPF and optical properties of tissue allows for cheaper technology that is more portable and user friendly, however the main caveat is that absolute heme concentrations cannot be calculated, only changes. Moreover, the DPF is wavelength specific, affected by differing thicknesses and changing optical properties of the ATT and skeletal muscle layers, and has been shown to change under occlusion and skeletal muscle contraction (Marco Ferrari et al., 1992). Because cw-NIRS devices assume constant optical properties, they are limited to only measuring relative [Hb], and 27 can potentially introduce confounding results for interindividual analyses and in circumstances where the DPF value has changed, for example, change in adipose thickness or exercise (Pirovano et al., 2021). Spatially-resolved (SRS) NIRS is an expansion of cw-NIRS that uses multiple light emitters at set distances from the detector (Figure 3.6). The intensity of the light emitted from the transmitters is measured as a function of the source-detector distance. The shape of this function is related to μa, from which the absolute heme concentrations and the tissue saturation index (TSI%), a surrogate of StO2, can be calculated. Figure 3.6. Schematic view of a TSI measurement. Light through tissue with three transmitters. Image borrowed from Artinis Manual. Time-Resolved NIRS As mentioned earlier, TRS was used in time-of-fight experiments to determine common DPF values in various muscle groups of men and women of varying ages (Duncan et al., 1995). In TRS, temporal changes in the reflected light intensity are measured after irradiation of an ultra-fast picosecond pulse, yielding a distribution of the total pathlength 28 of a photon traveling in the scattering medium. This technology allows for determination of relative light intensity, mean optical path length, and the optical properties of tissue μs an μa. Figure 3.7. Phase shift between the incident light (dashed line) and scattered light through tissues (solid lines) at 70 MHz of modulation frequency. Image borrowed from (Yamashita et al., 2013). Frequency-Domain NIRS Phase-modulated or fd-NIRS is based on amplitude-modulated light sources (at a frequency of the order of 100 MHz or larger). The time delay is related to light scattering and absorption, including biological signals. By utilizing multiple light emitters, frequency-domain multi-distance spectroscopy yields the average value (DC), amplitude (AC) and phase (ɸ) of the modulated light intensity at several source-detector distances, which by using equations derived from diffusion theory allows for continuous measurement of μs an μa and thus absolute heme concentrations and StO2 (Fantini, S. Franceschini, 2002). 29 Multi-Channel Devices Conventional NIRS devices offer a single channel spectrophotometer, which provides a limited (~3 cm3) sample volume. This raises the question whether the measured value is representative for the whole muscle, especially in large muscles like the quadriceps or gastrocnemius muscle. Recent advances in NIRS technology have included the addition of multiple-source detector pairs to image skeletal muscle. This has been done to take advantage of classical studies that have shown regional differences in skeletal muscle oxygenation and metabolism in different locations within a muscle (Laughlin & Armstrong, 1982). By simultaneously collecting data from multiple skeletal muscle regions, these devices avoid the variability caused by position dependent differences in muscle oxygenation that plague all single location measurements. Imaging devices also allow the study of regional differences in how skeletal muscle responds to exercise. The challenge of NIR imaging systems is how to evaluate the much greater amounts of information that are collected. Kime et al (Kime et al., 2005) evaluated heterogeneity of muscle O2 dynamics in a single muscle (VL) during bicycle exercise using an eight-channel NIRS mapping system. The half-time recovery (see above) was progressively delayed from distal sites to proximal sites of VL. On the other hand, there were no differences between medial and lateral sides at the same transverse level. Takagi et al (Takagi et al., 2013) examined O2 saturation in several leg muscles with progressive intensity exercise. Twelve healthy young males performed 20 W/min ramp bicycle exercise until exhaustion. Muscle O2 saturation was monitored continuously at the belly of the VL, rectus femoris, vastus medialis, biceps femoris, gastrocnemius lateralis, gastrocnemius medialis, and tibialis anterior by NIRS. Deoxygenation patterns were considerably different between muscles during ramp cycling exercise. Multiple channel NIRS devices have the potential to evaluate regional differences in oxygen status and could play an important role in monitoring exercise prescription and clinical uses (e.g. application of multiple NIRS imaging device to the exercising muscle metabolism). 30 NIRS Metrics and Methodology Adequate delivery of oxygen (DO2) to meet metabolic demand (V̇O2) is critical for all skeletal muscle, at rest and during exercise. At the microvascular level, DO2 is represented by microvascular blood flow (mBF) and V̇O2 by skeletal muscle oxygen consumption (mV̇O2). Delivery of O2 to the terminal cytochrome oxidase can be described by integration of the percussive (equation 3.4) and diffusive (equation .5) equations of Fick’s Law: eq 3.4: V̇O2 = Q * (Ca - Cv) O2 eq 3.5: V̇O2 = DO2 * (PmvO2 - pmitoO2) Where Q is blood flow, Ca and Cv are concentrations in arterial and venous blood, respectively, DO2 is the diffusivity of O2, and PmvO2 and PmitoO2 are the partial pressures of O2 in the microvasculature and mitochondria, respectively. The power of NIRS lies in its ability to non-invasively determine [HbO2] and [HHb], which combined with clever methodology can provide insight into the relative balance of mBF and mVO2 under given circumstances. Tissue Oxygen Saturation and Perfusion Early detection of inadequate tissue perfusion and oxygenation is critical in hemodynamic monitoring of oxygen saturation in tissue. Pulse Oximetry is the most common technology used to continuously measure tissue oxygenation but is limited in where it can be used (earlobe, finger, toe). Using similar technology, NIRS has greater tissue penetration than pulse oximetry, can be used to measure oxygenation up to 3 inches in depth all over the body, and provides a global assessment of oxygenation in all vascular compartments (arterial, venous, and capillary) (Lima & Bakker, 2006). From the calculated [O2Hb] and [HHb], tissue oxygen saturation (StO2) has been widely used to assess skeletal muscle oxygen saturation. StO2 can be calculated from the ratio of increase in O2Hb to tHb (O2Hb + HHb). 31 Unlike fd-NIRS which can calculate absolute [O2Hb] and [HHb], cw-NIRS is limited to relative changes in [O2Hb] and [HHb] and cannot calculate absolute concentrations it cannot calculate StO2 in a traditional sense but can calculate an Index of saturation known as the Tissue Saturation Index (TSI). This relies on the use of two or more light emitters to estimate absolute [Hb] values using principles of spatially resolved spectroscopy, and therefore the TSI is an estimation of StO2 percentage. Perfusion can be assessed as the average [tHb] ([O2Hb] + [HHb]) for a given period, and changes in [tHb] from a baseline are indicative of perfusion changes in the tissue being measured. A study using cw-NIRS looking at the perfusion changes in the calf muscles of patients with T2D, PAD, or both T2D + PAD before and after plantar-flexion exercise (Mohler et al., 2006)found reduced capillary volume expansion in T2D patients with or without PAD, concluding that it may be due to impaired vasodilation secondary to endothelial dysfunction. Skeletal Muscle Blood Flow Indocyanine Green Method using NIRS Near-infrared spectroscopy (NIRS) emitting a wavelength specific to the tracer indocyanine green (ICG) has been validated for measurement of regional MBF during exercise (Bonde-Petersen et al., 1975). The passage of ICG through the tissue is monitored noninvasively by NIRS probes taped to the skin overlying the muscles of interest. The most commonly used algorithm to analyze the tracer requires simultaneous recording of the ICG concentration curve in arterial blood as the input function (Bonde- Petersen et al., 1975). Thus, although NIRS itself is noninvasive, the measurement requires arterial catheterisation and continuous withdrawal of blood through a photodensitometer for several seconds after injection. Blood flow to the gastrocnemius measured by NIRS-ICG provides comparable values to dye dilution in combination with MRI (Boushel et al., 2000). 32 Given the invasive nature of catheterisation, an alternative algorithm has been proposed to calculate tissue perfusion from the NIRS data, namely the blood flow index (BFI). This index is calculated by dividing the ICG peak concentration by the time for ICG to reach peak concentration. Kuebler et al. (Kuebler et al., 1998) validated the BFI for determining cerebral perfusion in pigs against simultaneously measured regional blood flow derived from the radioactive microsphere technique. Although the BFI does not provide absolute blood flow values, it has been shown to sensitively detect perfusion differences between cerebral hemispheres after an acute ischemic stroke (Kuebler et al., 1998), to be well reproducible, and to be suited to detect intra-individual changes in brain blood flow (Kuebler, 2008; Wagner et al., 2003). The approach has been validated against the standard NIRS-ICG (described above), which relies on the Fick principle, for the measurement of skeletal muscle blood flow during exercise (r = 0.98) (Habazettl et al., 2010). Moreover, the inter-observer variability, analysed by Bland-Altman plots, was considerably lower for BFI compared to arterial catheterisation. Venous Occlusion Method using NIRS The venous occlusion method has the distinct advantage over the ICG method in that it can be used non-invasively to determine muscle blood flow (mBF) by applying a similar occlusion technique as used in conventional venous strain-gauge plethysmography (De Blasi et al., 1994; M. van Beekvelt et al., 2001). NIRS-determined mBF by the venous occlusion method has been shown to agree with traditional measurements using plethysmographyand the Fick method (De Blasi et al., 1994; M. van Beekvelt et al., 2001). Whereas strain-gauge plethysmography cannot distinguish between the various tissues of the limb, NIRS measures blood volume changes directly in the muscle of interest. The reproducibility of forearm blood flow at rest using the venous occlusion method was reasonable reliability (CVs 20.4%, 30.3%) (M. van Beekvelt et al., 2001). This method works by using rapid occlusions (~0.5 s) to manipulate blood flow so that microvascular blood flow can be measured. Rapidly inflating a tourniquet upstream 33 of the NIRS probe to a sub-systolic pressure (60-80 mmHg), known as a venous occlusion (VO), which occludes venous outflow without impeding arterial inflow, thus, the subsequent increase in venous volume is proportional to arterial inflow (Van Beekvelt et al. 2001b). The slope of the [tHb] signal after venous occlusion (See figure 6) can be converted into a mBF measure of units of mL per min per 100 mL of blood (mBF (mL·min-1·100 mL-1) = [(∆t[Hb]∙60)/ (([Hb]∙1000)/4)∙1000]/10) using a total hemoglobin count of 7.5 and 8.5 mmol∙l-1for female and male participants, respectively (M. van Beekvelt et al., 2001). For more accurate calculation of mBF, the literature values can be replaced with actual individual [Hb] determined from a blood sample. After cuff inflation, there is a rapid, progressive fall in mBF (especially during exercise), likely due to an increase in venous backpressure, diminishing the arteriovenous pressure gradient and stimulating the venoarterial reflex causing vasoconstriction of precapillary vessels (Rathbun et al., 2008). As a consequence, inclusion of more than one cardiac beat has been shown to underestimate mBF (Cross & Sabapathy, 2017a). Figure 3.8. Example trace of NIRS [tHb] (green), [O2Hb] (red), and [HHb] (blue) in response to venous occlusion (VO). 34 Skeletal Muscle Oxygen Consumption The venous occlusion method can be used to determine mV̇O2 and mBF by applying the same technique used in conventional venous plethysmography (M. van Beekvelt et al., 2001). Since venous outflow is blocked and the increase in HHb is thought to be solely due to the O2 consumed. Venous occlusion is less inconvenient for the subject and can be repeated at short time intervals (Homma et al., 1996). However, venous occlusion is also more prone to variations in flow within the arm due to changes in blood pressure and local vasoreactivity, whereas these influences are negligible during arterial occlusion because of the closed compartment, temporarily cut off from centrally mediated variations. Van Beekvelt et al reported that mV̇O2 from venous occlusion appeared unreliable when repeated several times within one session (M. van Beekvelt et al., 2001). In another study, the same group we found a coefficient of variation of 30.4% (Van Beekvelt et al., 2001) for resting mV̇O2 during venous occlusion while the coefficient of variation was much lower (16.2%) for an arterial occlusion method. Increasing the inflation pressure from sub-systolic to supra-systolic (250-300 mmHg), known as an arterial occlusion (AO), which restricts both venous outflow and arterial inflow, effectively arresting blood flow, is the more reliable and preferred method to measure mV̇O2. Under this method, tHb stays constant, and an increase of [HHb] and simultaneous decrease in [O2Hb] as oxygen is released from hemoglobin and diffused into the surrounding muscle tissue (see figure 3.9) (M. van Beekvelt et al., 2001). Since the rate of increase in [HHb] is directly proportional to the rate of increase in [HHb] and the concentration of each is obtained independently, the signals are often subtracted from each other to get a hemoglobin difference [HbDiff] which increases signal-to-noise. The slope of the [HbDif] can then be converted into milliliters of O2 per min per 100 grams of tissue (mV̇O2 (mlO2∙min-1∙100g-1) = abs[(∆[HbDif/2]∙60)/ (10 ·1.04) ·4] ∙ 22.4/1000), assuming 22.4 L for the volume of gas (STPD) and 1.04 kg∙L-1 for muscle density (M. van Beekvelt et al., 2001). 35 Figure 3.9. Example trace of NIRS [tHb] (green), [O2Hb] (red), and [HHb] (blue) in response to arterial occlusion (A.O.). Arterial occlusion derived mV̇O2 values have been found to agree with the conventional Fick method (M. van Beekvelt et al., 2001; M. C. P. Van Beekvelt et al., 2002) and MRS (Takafumi Hamaoka et al., 1996; T Sako et al., 2001), which measures mV̇O2 indirectly from PCr kinetics. Van Beekvelt et al (M. Van Beekvelt et al., 2002) calculated CVs for mVO2 in the human flexor digitorum superficialis at rest and during rhythmic isometric exercise, and reported a CV of 17.6% at rest and 16.3% - 23.3% during the different intensities of exercise. Recently, Ryan et al (T.E. Ryan et al., 2012) validated a blood volume corrected technique for mVO2 (see below) measured at rest on the medial gastrocnemius and vastus lateralis. The between-day CV for mVO2 improved from 31.4% for uncorrected measurements to 2.4% for corrected measurements versus. Ischaemic Calibration Since cw-NIRS cannot measure absolute concentration of Hb, an ischaemic calibration must be performed to normalize the slope of the tHb signal to account for any differences in [tHb]. A simple and common method of calibrating the NIRS signal is to 36 use the range of muscle oxygenation caused by arterial occlusion followed by reactive hyperemia (B. Chance et al., 1992). The arterial occlusion method is based on the assumptions that an adequate duration of ischemia will result in the complete disappearance of O2Hb, and that the reactive hyperemia after occlusion will almost completely eliminate HHb. So while O2Hb and HHb in arbitrary units may vary between measurement sites and individuals, the occlusion calibration will account for these changes (Takafumi Hamaoka et al., 2007). Ischaemic calibrations are not required for frequency-domain or time-of-flight NIRS. Correction for Blood-Volume Change A number of researchers have suggested that during occlusions there is a blood volume change that could confound the slope measurements for oxygen consumption (M. van Beekvelt et al., 2001). Blood volume changes during occlusions can mask changes in the NIRS signals due to oxygen consumption. In order for the arterial occlusion method of measuring mV̇O2 to be valid, the NIRS signal should reflect a symmetrical change in O2Hb and HHb, and therefore no change in tHb (tHb = O2Hb + HHb) (T. E. Ryan et al., 2012). The dissociation of oxygen molecules from oxyhemoglobin/myoglobin, which supplies the oxygen required for oxidative phosphorylation, should be reflected in equal and opposite changes in the NIRS signals for O2Hb and HHb. The calculation of a blood volume correction factor (ß) is based on the assumption that during an arterial occlusion, changes in O2Hb and HHb occur with a 1:1 ratio that represents mitochondrial oxygen consumption only (no arterial oxygen delivery or venous return of deoxygenated blood). The change in NIRS signals (both HHb and O2Hb) during an arterial occlusion occurs due to a metabolic consumption of oxygen and a blood volume flux from the redistribution of heme between high pressure arteries/arterioles and low-pressure veins/venules. Therefore, accurate quantification of mV̇O2 requires either 1) no change in blood volume, or 2) the removal of blood volume change from the NIRS signal (i.e., mV̇O2 = ∆NIRS Signal - ∆blood volume). To correct NIRS signals for changes in blood volume, the blood change must be proportioned into 37 oxygenated and deoxygenated sources. The equation below describes the calculation for this correction factor: eq 3.6: β(t)=|O2Hb(t)| / (|O2Hb(t)|+ |HHb(t)| Where β is the blood volume correction factor, t represents time, O2Hb is the oxygenated hemoglobin/myoglobin signal, and HHb is the deoxygenated hemoglobin/myoglobin signal. β is a nondirectional factor that represents the proportionality of the blood volume change (values range from 0 to 1). Mitochondrial Oxidative Respiratory Capacity Motobe and colleagues (Motobe et al., 2004) first described a method for measuring mitochondrial capacity using transient arterial occlusions following a bout of skeletal muscle contraction to deplete the ATP so the subsequent recovery curve can be measured. Electrical stimulation and voluntary exercise can be used as the bout of contraction, both methods have been shown to provide comparable results (Ryan et al., 2013), and the recovery curve not influenced by exercise intensity (Ryan et al., 2013). The signal should be corrected using ischemic calibration if a