Here, we describe a simple, non-invasive approach using near-infrared spectroscopy to assess reactive hyperemia, neurovascular coupling and skeletal muscle oxidative capacity in a single clinic or laboratory visit.
Exercise represents a major hemodynamic stress that demands a highly coordinated neurovascular response in order to match oxygen delivery to metabolic demand. Reactive hyperemia (in response to a brief period of tissue ischemia) is an independent predictor of cardiovascular events and provides important insight into vascular health and vasodilatory capacity. Skeletal muscle oxidative capacity is equally important in health and disease, as it determines the energy supply for myocellular processes. Here, we describe a simple, non-invasive approach using near-infrared spectroscopy to assess each of these major clinical endpoints (reactive hyperemia, neurovascular coupling, and muscle oxidative capacity) during a single clinic or laboratory visit. Unlike Doppler ultrasound, magnetic resonance images/spectroscopy, or invasive catheter-based flow measurements or muscle biopsies, our approach is less operator-dependent, low-cost, and completely non-invasive. Representative data from our lab taken together with summary data from previously published literature illustrate the utility of each of these end-points. Once this technique is mastered, application to clinical populations will provide important mechanistic insight into exercise intolerance and cardiovascular dysfunction.
The hyperemic response to a brief period of tissue ischemia has emerged as a key non-invasive measure of (micro)vascular function. During occlusion of a conduit artery, downstream arterioles dilate in an effort to offset the ischemic insult. Upon release of the occlusion, the decreased vascular resistance results in hyperemia, the magnitude of which is dictated by one's ability to dilate the downstream microvasculature. While reactive hyperemia is a strong independent predictor of cardiovascular events1,2 and therefore a clinically significant endpoint, its functional significance to exercise tolerance and quality of life is less clear.
Indeed, dynamic exercise represents a major cardiovascular stress that demands a highly coordinated neurovascular response in order to match oxygen delivery to metabolic demand. For example, skeletal muscle blood flow can increase nearly 100-fold during isolated muscle contractions3, which would overwhelm the pumping capacity of the heart if such a hemodynamic response were extrapolated to whole-body exercise. Accordingly, to avoid severe hypotension, sympathetic (i.e., vasoconstrictor) nervous activity increases to redistribute cardiac output away from inactive and visceral tissues and towards active skeletal muscle4. Sympathetic outflow is also directed to the exercising skeletal muscle5; however, local metabolic signaling attenuates the vasoconstrictor response in order to ensure adequate tissue oxygen delivery6,7,8,9,10,11. Collectively, this process is termed functional sympatholysis12, and is imperative to the normal regulation of skeletal muscle blood flow during exercise. Since skeletal muscle blood flow is a key determinant of aerobic capacity — an independent predictor of quality of life and cardiovascular disease morbidity and mortality13— understanding the control of skeletal muscle blood flow and tissue oxygen delivery during exercise is of great clinical significance.
Oxygen delivery is only half of the Fick equation, however, with oxygen utilization satisfying the other half of the equation. Among the major determinates of oxygen utilization, mitochondrial oxidative phosphorylation plays an essential role in supplying adequate energy for cellular processes both at rest and during exercise. Indeed, impairments in muscle oxidative capacity can limit functional capacity and quality of life14,15,16. Various measures are commonly used to provide an index of muscle oxidative capacity, including invasive muscle biopsies and expensive and time-consuming magnetic resonance spectroscopy (MRS) techniques.
Here, we propose a novel, non-invasive approach, using near-infrared spectroscopy (NIRS), to assess each of these three major clinical endpoints (reactive hyperemia, sympatholysis, and muscle oxidative capacity) in a single clinic or laboratory visit. The major advantages of this approach are three-fold: First, this technique is easily portable, relatively low cost, and easy to perform. Current Doppler ultrasound approaches for measuring reactive hyperemia are highly operator-dependent — requiring extensive skill and training — and require sophisticated, high-cost, data acquisition hardware and post-processing software. Moreover, this could conceivably be introduced into the clinic and/or large clinical trials for bedside monitoring or testing therapeutic efficacy. Second, by virtue of the methodology, this technique focuses specifically on the skeletal muscle microvasculature, increasing the overall specificity of the technique. Alternative approaches using Doppler ultrasound focus entirely on upstream conduit vessels and infer changes downstream, which can dampen the signal. Third, this technique is completely non-invasive. Skeletal muscle oxidative capacity is traditionally assessed with invasive and painful muscle biopsies, and functional sympatholysis may be assessed with intra-arterial injection of sympathomimetics and sympatholytics. This approach avoids these requirements all together.
This protocol follows the guidelines of the institutional review board at the University of Texas at Arlington and conforms to the standards set by the latest version of the Declaration of Helsinki. Accordingly, written informed consent was (and should be) obtained prior to commencement of research procedures.
1. Instrumentation
NOTE: The following instrumentation description is based on the near-infrared (NIR) spectrometer and data acquisition system used in our lab (see Table of Materials). Thus, the instructions include steps that are necessary for the optimal function of these devices. These steps include the calibration of the NIR probe using the accompanying software and calibration phantom, and the application of a dark cloth to exclude ambient light. In the event that different data collection hardware and/or software are used, investigators should consult their own specific user manuals for calibration and ambient light considerations. Figure 1 illustrates the experimental set-up and instrumentation described immediately below.
2. Skeletal Muscle Oxidative Capacity
NOTE: A representative data tracing illustrating the experimental procedure for measuring skeletal muscle oxidative capacity is depicted in Figure 2. This experimental approach has previously been validated against in vivo phosphorus MRS18 and in situ muscle respirometry19, and is gaining widespread acceptance20.
3. Reactive Hyperemia
NOTE: A representative data tracing illustrating the experimental procedure for measuring reactive hyperemia is depicted in Figure 3.
4. Functional Sympatholysis
NOTE: A representative data tracing illustrating the experimental procedure for measuring functional sympatholysis is depicted in Figure 4.
Skeletal muscle oxidative capacity
Figure 2 illustrates a representative participant response during a NIRS-derived skeletal muscle oxidative capacity assessment. Panel A shows the tissue saturation profile during a 5 min arterial cuff occlusion protocol, handgrip exercise, and intermittent arterial occlusion during recovery from exercise. Panel B illustrates the expected tissue desaturation/re-saturation profile during the intermittent arterial occlusions during the recovery period. The rate of desaturation is directly proportional to the rate of muscle oxygen consumption, and is plotted in Panel C for each of the intermittent cuff occlusion periods. The calculated muscle oxygen consumption recovery data is then fit to a monoexponential curve and the recovery time constant derived. Using the same approach, a growing number of studies have evaluated skeletal muscle oxidative capacity for both health and disease, across a variety of muscle groups (Table 1).
Reactive Hyperemia
Figure 3 illustrates the NIRS-derived reactive hyperemia profile during a representative vascular occlusion test. This same approach has been used across a wide range of study populations and muscle groups with good success (Table 2). The data indicate that NIRS-derived reactive hyperemia not only provides valuable insight into vascular reactivity, but that the test is easily adaptable and clinically meaningful.
Functional Sympatholysis
Table 3 summarizes the existing literature using the exact same neurovascular coupling approach described herein to measure functional sympatholysis, showing both mechanistic and clinically relevant outcomes. In healthy control subjects, when LBNP is superimposed on mild handgrip, the reflex decrease in muscle oxygenation is attenuated by ~50% (Figure 4). Failure to attenuate sympathetic (vasoconstrictor) nerve activity during exercise, as with cardiovascular or neurological disease (Table 3), disrupts the balance between oxygen delivery and utilization, and causes functional muscle ischemia.
Figure 1. Experimental set-up and instrumentation. (A) Representative experimental set-up, with a typical subject lying supine on a bed with their legs inside the LBNP chamber and fully instrumented. (B) Dominant arm instrumented with a non-invasive beat-to-beat blood pressure device for beat-to-beat arterial blood pressure measurement, and a brachial artery blood pressure cuff for calibration and verification of the beat-to-beat system. (C) Instrumentation of the non-dominant arm. The hand is comfortably gripping a handgrip dynamometer (connected to data acquisition system), and the forearm muscle is instrumented with the near-infrared spectroscopy probe. (D) Once instrumented, the NIRS optodes are covered with a black vinyl cloth (to eliminate interference from ambient light). In addition, a rapid cuff inflation system is placed over the brachial artery. Please click here to view a larger version of this figure.
Figure 2. Skeletal muscle oxidative capacity protocol. (A) Raw data tracing from a representative subject measured via NIRS, showing tissue saturation (StO2) over time. After establishing a stable baseline, the brachial artery of the non-dominant arm is occluded for five min in order to establish the subject's desaturation reserve (difference between baseline StO2 and the nadir). After recovery from the occlusion, the subject is instructed to perform a 50% isometric handgrip, followed by 18 rapid cuff inflation series to assess muscle oxygen consumption recovery kinetics. (B) Data analysis is then performed offline by calculating the average slope of each cuff occlusion series following exercise; illustrated here using hypothetical cuff occlusion series data. (C) In order to calculate the recovery time constant of muscle oxygenation, the slope of each of the 18 rapid cuff occlusions (i.e., post-exercise muscle oxygen consumption, mV̇O2) from A is plotted against time and fit to a monoexponential curve. Please click here to view a larger version of this figure.
Figure 3. Reactive hyperemia experimental protocol. With the subject lying supine, record at least 1 min of baseline data, followed by 5 min of total arterial cuff occlusion, and at least 3 min of recovery following cuff release. Note the obvious overlap between the skeletal muscle oxidative capacity protocol (Figure 2) and this protocol. 'Baseline' defines the period of time prior to arterial cuff occlusion. 'Slope 1' defines the desaturation rate during cuff occlusion, and is regarded as a measure of resting skeletal muscle metabolic rate. The lowest StO2 value obtained during ischemia is defined as 'StO2 minimum', and is regarded as a measure of the ischemic stimulus to vasodilate. The tissue saturation reperfusion rate is denoted as 'Slope 2', and is an index of reactive hyperemia; as are StO2 maximum, and the reactive hyperemia 'area under the curve' (AUC). To gain insight into the hyperemic reserve, the StO2 maximum is expressed as a percent change from baseline. Please click here to view a larger version of this figure.
Figure 4. Functional sympatholysis experimental protocol. Left panel: Raw data tracing from a representative subject. With the subject lying supine in the LBNP chamber, allow 3 min of steady-state baseline data collection. Turn on LBNP to -20 mmHg for 2 min. Oxyhemoglobin/myoglobin should decrease in response to the reflex sympathetic vasoconstriction (blue circle, shaded area). Allow 2 min for recovery. Ask the subject to perform rhythmic handgrip exercise at 20% MVC (measured prior to data collection). After 3 min of rhythmic exercise, repeat -20 mmHg LBNP for 2 min while the subject continues to exercise, followed by 2 min of exercise without LBNP. The reduction in oxyhemoglobin/myoglobin should be significantly attenuated (red circle, shaded area). If not already performed, inflate a blood pressure cuff over the brachial artery of the exercising arm for 5 min to establish the subject's range of desaturation. Note that the shaded areas in the figure are only meant to highlight the changes in oxyhemoglobin/myoglobin; see protocol for details on how to analyze the outcome variables used to calculate sympatholysis. Right Panel: LBNP-induced change in oxyhemoglobin/myoglobin at rest and during handgrip exercise calculated from the data on the left. Please click here to view a larger version of this figure.
Reference/Data Set | Study Population | Sample size (n) |
Age of participants (years ± SD) |
Tau (τ) (s) |
Muscle group | NIRS Variable reported | Device |
Brizendine et al. (2013) | Endurance Athletes | 8 | 25 ± 3 | 19 | Vastus lateralis | Hbdiff/total blood volume | Continuous wave (Oxymon MK III) |
Ryan et al. (2014) | Young, healthy | 21 | 26 ± 2 | 55 | Vastus lateralis | HHb | Continuous wave (Oxymon MK III) |
Southern et al. (2015) | Elderly | 23 | 61 ± 5 | 63 | Wrist flexor | Hbdiff | Continuous wave (Oxymon MK III) |
Elderly + Heart Failure | 16 | 65 ± 7 | 77 | Wrist flexor | Continuous wave (Oxymon MK III) |
||
Adami et al. (2017) | Smokers with normal spirometry | 23 | 63 ± 7 | 80 | Medial forearm | Tissue saturation index (TSI) | Continuous wave (Portamon) |
COPD Gold 2-4 | 16 | 64 ± 9 | 100 | Medial forearm | Continuous wave (Portamon) |
||
Erickson et al. (2013) | Spinal cord injury | 9 | 43 ± 11 | 143 | Vastus lateralis | HbO2 | Continuous wave (Oxymon MK III) |
Table 1: Summary of previously published reports across the health continuum using near-infrared spectroscopy to measure skeletal muscle oxidative capacity.
Reference | Study Population | Muscle Group | Reported Outcomes | Outcome Value |
Lacroix, J Biomed Opt, 2012 | Healthy Males | Forearm | Peak Oxyhemoglobin | 28.05 ± 3.15 μM |
Peak Total Hemoglobin | 10.56 ± 1.80 μM | |||
Increase Rate to Peak HbO2 | 0.75 ± 0.22 μM/s | |||
Increase Rate to Peak Total Hb | 0.52 ± 0.16 μM/s | |||
Kragelj, Ann Biomed Eng, 2001 | Peripheral Vascular Disease | Forearm | Oxygen Consumption | 0.68 ± 0.04 mL/min |
Time to Peak | 153 ± 16 s | |||
Maximal Absolute Change in HbO2 | 2.93 ± 0.22 μM/100 mL | |||
Suffoletto, Resuscitation, 2012 | Post-Cardiac Arrest ICU Admittants | Thenar Eminence | Desaturation Rate | -5.6 ± 2 %/min |
Resaturation Rate | 0.9 ± 0.6 %/sec | |||
Dimopoulos, Respir Care, 2013 | Pulmonary Artery Hypertension | Thenar Eminence | Baseline Saturation with 21% O2 | 65.8 ± 14.9 % |
O2 Consumption Rate with 21% O2 | 35.3 ± 9.1 %/min | |||
Reperfusion Rate with 21% O2 | 535 ± 179 %/min | |||
Doerschug, Am J Physiol Heart Circ Physiol, 2007 | Organ Failure & Sepsis | Forearm | Baseline Saturation | 84% |
Reoxygenation Rate | 3.6 %/s | |||
Mayeur, Crit Care Med, 2011 | Septic Shock | Thenar Eminence | Baseline Saturation | 80 ± 1.0 % |
Desaturation Slope | -9.8 ± 3.7 %/min | |||
Recovery Slope | 2.3 ± 1.4 %/sec | |||
McLay, Exp Physiol, 2016 | Healthy Males | Tibialis Anterior | Baseline Saturation | 71.3 ± 2.9 % |
Minimum Saturation | 44.8 ± 8.6 % | |||
Desaturation Slope | -0.1 ± 0.03 %/s | |||
Recovery Slope | 1.63 ± 0.5 %/s | |||
Peak Saturation | 82.6 ± 2.3 % | |||
McLay, Physiol Rep, 2016 | Healthy Males | Tibialis Anterior | Baseline Saturation | 71.1 ± 2.4 % |
Minimum Saturation | 46.2 ± 7.5 % | |||
Peak Saturation | 82.1 ± 1.4 % | |||
Recovery Slope | 1.32 ± 0.38 %/s |
Table 2: Summary of previously published reports across the health continuum using near-infrared spectroscopy to measure reactive hyperemia.
Reference | Study Population | % Attenuation |
Nelson MD, J. Physiol, 2015 | Healthy | -57 |
Becker Muscular Dystrophy | -13 | |
Vongpatanasin, J. Physiol, 2011 | Healthy | -93 |
Hypertension | -14 | |
Fadel, J. Physiol, 2004 | Pre-Menopause | -84 |
Post-Menopause | -19 | |
Sander, PNAS, 2000 | Healthy | -74 |
Duchenne Muscular Dystrophy | .+7 | |
Nelson MD, Neurology, 2014 | Healthy | -54 |
Duchenne Muscular Dystrophy | -7 | |
Price, Hypertension, 2013 | Hypertension Pre-Treatment | -52 |
Hypertension Post-Nebivolol Treatment | -97 | |
Hansen, J. Clin. Invest., 1996 | Healthy Exercise at 20% MVC | -92 |
Healthy Exercise at 30% MVC | -125 |
Table 3: Summary of previously published reports across the health continuum using near-infrared spectroscopy, in combination with lower body negative pressure and handgrip exercise, to assess functional sympatholysis.
The methods described herein enable non-invasive, clinical evaluation of reactive hyperemia, neurovascular coupling, and skeletal muscle oxidative capacity in a single clinic or laboratory visit.
Critical Considerations
Although NIRS is relatively robust and easy to use, collection of these data require careful placement of the optodes directly over the muscle belly, secured tightly in place to avoid movement artifact, and covered with a black vinyl sheet in a dimly lit room to avoid interference of the near infrared from external light. In addition, obtaining good quality data relies heavily on clear communication between the tester and the subject, and the testing team. We, and others, have found that when performed with appropriate care and attention, NIRS is highly reproducible within a single study visit, and across multiple visits10,11,24,29. Moreover, the physiological outcome variables reported herein (i.e., skeletal muscle oxidative capacity, reactive hyperemia, and neurovascular coupling) are sensitive to experimental/clinical intervention, both within and between study visits30,31,10,11.
There is currently limited consensus on the appropriate reporting of the NIRS outcome variables. For example, when measuring skeletal muscle oxidative capacity, investigators have fit the recovery kinetics of HbO232, HHb19, Hbdiff30 and tissue O2 saturation (present study and others33). Likewise, a similar spread in the outcome variables have also been reported for NIRS-based reactive hyperemia.34,35,36,37Some of this discrepancy may relate to the type of NIRS device used. For example, frequency-domain devices (as used here) provide absolute quantification of HbO2 and HHb, and thus may not be affected by acute changes in total Hb content (negating the need to correct the data). In contrast however, continuous-wave devices are greatly affected by acute changes in total hemoglobin, requiring data correction25.
Modifications and troubleshooting
One important and currently unavoidable limitation of NIRS is its limited penetration depth (~2 cm). Therefore, limb adiposity can significantly reduce — and even completely eliminate — the NIRS signal and should be considered when screening potential subjects. To control for this, investigators are encouraged to measure forearm skinfold thickness, and exclude participants with significant peripheral adiposity.
Any factor that can modulate vascular responsiveness, neurovascular coupling, and/or skeletal muscle oxidative capacity (i.e., medication, genetic mutations, etc.) will indeed affect the primary end-point measurements described herein. Investigators are therefore encouraged to take these factors into account when adapting this protocol and planning future experimentation.
For functional sympatholysis determination, investigators may wish to include a second resting LBNP challenge to ensure the signal is still present and that the differences observed during exercise-LBNP were not simply due to a loss of signal or measurement error. It is recommended to allow 3 – 5 min to allow the oxyhemoglobin signal to full recovery to baseline values before repeating the resting LBNP challenge.
Future applications or directions after mastering this technique
NIR spectroscopy uses laser light to assess the concentration of oxygenated and deoxygenated hemoglobin in tissue. During measurement of reactive hyperemia and functional sympatholysis, relative changes in these parameters are believed to represent changes in microvascular flow. Diffuse correlation spectroscopy (DCS) is an emerging near-infrared imaging approach which, in addition to evaluating the concentration of oxy- and deoxyhemoglobin, can also quantify microvascular perfusion38. Given the obvious similarities between these two imaging approaches, incorporation of DCS into the proposed techniques would be virtually seamless and may provide additional insight into the quantification of microvascular function and perfusion.
Once this technique is mastered, application to clinical populations, such as those with heart failure, will provide important mechanistic insight into exercise intolerance and cardiovascular dysfunction.
The authors have nothing to disclose.
This work was supported by a University of Texas at Arlington Interdisciplinary Research Program grant.
Dual-channel OxiplexTS Near-infrared spectroscopy machine | Iss Medical | 101 | |
NIRS muscle sensor | Iss Medical | 201.2 | |
E20 Rapid cuff inflation system | Hokanson | E20 | |
AG101 Air Source | Hokanson | AG101 | |
Smedley Handgrip dynometer (recording) | Stolting | 56380 | |
Powerlab 16/35, 16 Channel Recorder | ADInstruments | PL3516 | |
Human NIBP Set | ADInstruments | ML282-SM | |
Bio Amp | ADInstruments | FE132 | |
Quad Bridge Amp | ADInstruments | FE224 | |
Connex Spot Monitor | Welch Allyn | 71WX-B | |
Origin(Pro) graphing software | OrignPro | Pro | |
Lower body negative pressure chamber | Physiology Research Instruments | standard unit |