Near-infrared diffuse correlation spectroscopy tracks changes in oxygen delivery and utilization during exercise with and without isolated arterial compression

Abstract

Near-infrared diffuse correlation spectroscopy (NIR-DCS) is an emerging technology for simultaneous measurement of skeletal muscle microvascular oxygen delivery and utilization during exercise. The extent to which NIR-DCS can track acute changes in oxygen delivery and utilization has not yet been fully established. To address this knowledge gap, 14 healthy men performed rhythmic handgrip exercise at 30% maximal voluntary contraction, with and without isolated brachial artery compression, designed to acutely reduce convective oxygen delivery to the exercising muscle. Radial artery blood flow (Duplex Ultrasound) and NIR-DCS derived variables [blood flow index (BFI), tissue oxygen saturation ($\text{S}{\text{t}}_{{\text{O}}_2}$), and metabolic rate of oxygen ($\text{M}{\text{R}}_{{\text{O}}_2}$)] were simultaneously measured. During exercise, both radial artery blood flow (+51.6 ± 20.3 mL/min) and DCS-derived BFI (+155.0 ± 82.2%) increased significantly (P < 0.001), whereas $\text{S}{\text{t}}_{{\text{O}}_2}$ decreased −7.9 ± 6.2% (P = 0.002) from rest. Brachial artery compression during exercise caused a significant reduction in both radial artery blood flow (−32.0 ± 19.5 mL/min, P = 0.001) and DCS-derived BFI (−57.3 ± 51.1%, P = 0.01) and a further reduction of $\text{S}{\text{t}}_{{\text{O}}_2}$ (−5.6 ± 3.8%, P = 0.001) compared with exercise without compression. $\text{M}{\text{R}}_{{\text{O}}_2}$ was not significantly reduced during arterial compression (P = 0.83) due to compensatory reductions in $\text{S}{\text{t}}_{{\text{O}}_2}$, driven by increases in deoxyhemoglobin/myoglobin (+7.1 ± 6.1 μM, P = 0.01; an index of oxygen extraction). Together, these proof-of-concept data help to further validate NIR-DCS as an effective tool to assess the determinants of skeletal muscle oxygen consumption at the level of the microvasculature during exercise.

INTRODUCTION

Diffuse correlation spectroscopy (DCS) is an emerging optical imaging approach to measure skeletal muscle perfusion during exercise (3, 22, 39, 46). DCS is completely noninvasive and has high temporal resolution for measuring relative changes in blood flow at the level of the microvasculature (3, 7). In addition to its validation in skeletal muscle, DCS has also been validated for assessing tissue perfusion in a variety of other organs and tissues, against several different standards, including laser Doppler (41), Xenon-enhanced computed tomography (28), fluorescent microsphere flow measurements (49), and arterial spin labeled-MRI (48).

Pairing DCS with near-infrared spectroscopy (NIRS), an established noninvasive technique for characterizing the transport and utilization of oxygen through the microcirculation (4, 5, 11, 15, 19, 23, 39, 42, 46), allows for simultaneous measurement of oxygen delivery and utilization (determinants of skeletal muscle oxygen consumption) at the level of the microvasculature. Indeed, recent studies performed in our laboratory (3, 39, 46) demonstrate good concurrent validity between near-infrared diffuse correlation spectroscopy (NIR-DCS) and conventional measures of oxygen delivery (Doppler Ultrasound) and utilization (venous oxygen saturation) during both fixed workload (46) and incremental (39) handgrip exercise. However, it is still unclear whether NIR-DCS can track acute perturbations in oxygen delivery and utilization at the level of the microvasculature.

In a prior study from our laboratory (46), we attempted to alter microvascular oxygen delivery by manipulating forearm arterial perfusion pressure (by raising the arm above or lowering the arm below the level of the heart) (6, 40, 45). Whereas bulk conduit brachial artery blood flow (measured by Doppler ultrasound) followed changes in perfusion pressure during arm manipulation, microvascular perfusion was preserved during steady-state handgrip exercise, which we interpreted as evidence of myogenic autoregulation (46). In a separate, more recent NIR-DCS study, we challenged the system using a graded handgrip exercise test, hoping to observe a divergence between oxygen delivery and oxygen extraction at higher workloads (39). However, in this small muscle mass model, oxygen delivery and extraction increased linearly with each workload. Although these observations were insightful, our fundamental objective (i.e., challenging the balance between microvascular oxygen delivery and extraction) has yet to be achieved.

To address this knowledge gap, we performed the present ‘proof-of-concept’ study, designed to acutely reduce convective oxygen delivery to exercising skeletal muscle (i.e., forearm), through isolated arterial compression of the brachial artery. This model has previously been used at rest and during rhythmic handgrip exercise to reduce shear rate (8, 37, 38, 44) and downstream oxygen delivery . The major advantage of this approach over the more conventional cuff occlusion technique is that arterial oxygen delivery is reduced without disrupting venous return, which can alter local skeletal muscle hemodynamics as well as systemic hemodynamics via the metaboreflex (1, 10, 17, 34). We hypothesized that, during compression, NIR-DCS would effectively track reductions in skeletal muscle microvascular perfusion (measured by DCS) and compensatory reductions in tissue saturation (i.e., increased oxygen extraction) (measured by NIRS) to maintain skeletal muscle oxygen consumption throughout exercise.

MATERIALS AND METHODS

Ethical Approval and Subjects

The study was approved by the Institutional Review Board for research involving Human Subjects at the University of Texas at Arlington. All subjects provided informed written consent before participation in the study.

Healthy recreationally active men (18–35 yr old) were recruited from the University of Texas at Arlington campus and invited to participate in the study. Exclusion criteria included current medication or tobacco use, history of cardiovascular, liver, pulmonary, or metabolic disease, orthopedic limitations, blood pressure > 140/90 mmHg, or body mass index <18.5 or >35 kg/m².

Study Design

Experimental testing visit.

All subjects came to the laboratory for one experimental testing visit. All experiments were performed in a quiet, climate- (~22°C temperature, 40% relative humidity) and ambient light-controlled room. Subjects were instructed to abstain from alcohol and vigorous exercise for >24 h, caffeine for >12 h, and food for >4 h before the testing visit. Upon arrival at the laboratory, subjects were positioned supine on a bed and asked to rest quietly for a minimum of 20 min while being instrumented for continuous measurement of heart rate, blood pressure, radial blood flow, and NIR-DCS-derived variables. A Smedly handgrip dynamometer was positioned next to the subject’s exercising arm in such a way that the arm could be fully extended, supported, and comfortable.

Once fully instrumented, subjects performed a 2-min resting baseline followed by 7 min of rhythmic handgrip exercise at 30% of maximal voluntary contraction using a 50% contraction duty cycle (2-s contraction, 2-s relaxation) at a rate of 15 contractions per min, with simultaneous measurements of Doppler-derived radial blood flow and NIR-DCS-derived variables [blood flow index (BFI), tissue oxygen saturation ($\text{S}{\text{t}}_{{\text{O}}_2}$), and metabolic rate of oxygen ($\text{M}{\text{R}}_{{\text{O}}_2}$)]. After 3 min of exercise, one of the experimenters (R. Rosenberry) compressed the brachial artery of the subject’s exercising arm with finger pressure applied to the brachial pulse for 2 min to reduce arterial oxygen delivery to the active muscle (further details about arterial compression described in Brachial artery compression). After 2 min of exercise with arterial compression, the experimenter (R. Rosenberry) released compression and the final 2 min of exercise were performed. A schematic of this experimental protocol and brachial artery compression model is presented in Fig. 1. A large PC monitor was used to display real-time exercise handgrip force outputs to give the subject visual feedback on force production. A recorded voice prompt was used to guide contraction/relaxation cycles.

Fig. 1.

Experimental protocol (top right), experimental compression model to reduce arterial blood flow to the exercising limb (bottom left), and effect of experimental compression model on radial artery blood velocity (cm/s) during 30% maximal voluntary contraction (MVC) rhythmic handgrip exercise in a representative subject (bottom right).

Brachial artery compression.

To reduce blood flow/oxygen delivery to the exercising forearm, the brachial artery was palpated proximal to the antecubital fossa in the left arm and manually compressed (Fig. 1). Real-time blood velocity values displayed on our online data acquisition system acted as feedback for the experimenter who was applying the compression. The goal was to apply enough compression pressure to reduce radial artery blood velocity by 40–50% and maintain that pressure for 2 min. The use of this arterial compression model to reduce blood velocity and shear rate has been utilized by others (8, 17, 37, 38, 44). Although others have used a pneumatic piston or linear actuator device to apply constant pressure to the brachial artery during the compression at rest (37, 38), this approach is less feasible during exercise because of subject movement (38). Therefore, an arterial compression achieved via manual experimenter finger pressure applied to the brachial pulse was preferred for this study.

Instrumentation and Measurements

Anthropometrics.

Height and weight were measured with a dual-function stadiometer and weighing scale (Professional 500KL, Health-O-Meter, McCook, IL).

Heart rate and blood pressure responses.

Heart rate was measured via three-lead electrocardiography using standard CM₅ placement of ECG electrodes (MLA 0313; ADInstruments, Colorado Springs, CO). Beat-to-beat arterial blood pressure was measured with a small finger cuff placed around the middle finger of the nonexercising hand at heart level using photoplethysmography (Finometer PRO, Finapres Medical Systems, Arnhem, The Netherlands), which was calibrated against an automated brachial artery blood pressure cuff (Connex Spot Monitor, Model 71WX-B, Welch Allyn, Skaneateles Falls, NY). Analog outputs for blood pressure and heart rate were connected to a high-performance physiological data acquisition system (Powerlab 16/35, ADInstruments, Colorado Springs, CO) for real-time, continuous data recording.

Radial artery blood flow.

Radial artery blood velocity and diameter were measured with a duplex ultrasound system (Vivid-I, GE Healthcare, Little Chalfont, United Kingdom) at the midpoint of the left exercising forearm (~10 cm proximal from the carpal joint) using a 12-MHz linear array probe with 60 degrees of insonation. Ultrasound gate was optimized to ensure complete insonation of the entire artery cross-section with constant intensity. A continuous Doppler audio signal was converted into real-time mean blood velocity waveforms using a validated Doppler audio converter (26) and recorded using a PowerLab data acquisition system (ADInstruments, Colorado Springs, CO). Radial artery diameter was measured manually in triplicate by a single observer with B-mode ultrasound imaging, with measurements made during the resting baseline and during the final 15 s of each experimental condition (exercise, exercise with compression, exercise following compression release) within the 7 min of uninterrupted rhythmic handgrip exercise. Analog outputs from the Doppler ultrasound module, along with the amplified signal from the handgrip dynamometer, were connected to the data acquisition system detailed above for time-aligned simultaneous data recording. The radial artery was selected for downstream imaging because of concerns over mechanically induced brachial artery vasodilation or constriction that may accompany the arterial compression in close proximity to the vessel being imaged (14, 30).

Near-infrared diffuse correlation spectroscopy.

The methodology and validation of our in-house, dual-wavelength DCS system has previously been published (3, 39, 46). The system consists of two continuous-wave, long-coherence-length laser diodes (785 nm and 852 nm; Crystalaser, Reno, NV) that are alternatively switched (MEMS 2×2 Blocking Switch, Dicon Fiberoptics, Richmond, CA) and a single-photon-counting avalanche photodiode (APD) as the photon detector (SPCM-AQRH-14-FC, Pacer USA, Palm Beach Gardens, FL). The output of the APD is connected to a computer with a 32-bit, 8-channel data acquisition card (PCI-6602, National Instruments, Austin, TX). A LabVIEW (National Instruments, Austin, TX) program was developed for photon counting. A software autocorrelator calculates the autocorrelation function and absolute intensity (sum of photon counts) of diffused light (16). A three-dimensional-printed probe was used to hold a multimode fiber (125 μm in core diameter) from the optical switch and a single-mode fiber (5 μm in core diameter) to the APD detector. The probe was affixed to the forearm of the exercising arm, directly over the belly of the flexor digitorum profundus, using Velcro strips. The source-to-detector distance was 2.5 cm. The data sampling rate was 0.25 Hz.

By being operated at two wavelengths (785 nm and 852 nm), this DCS system measures not only relative changes in microvascular BFI (%), but also oxygenated hemoglobin/myoglobin (oxy [Hb+Mb]; %) and deoxygenated hemoglobin/myoglobin (deoxy [Hb+Mb]; %) concentrations based on conventional NIRS principles. Additionally, by obtaining optical properties (tissue absorption and scattering) before the experiment using a frequency-domain near-infrared tissue oximeter (OxiplexTS, ISS, Champaign, IL), we are able to estimate absolute changes in oxy [Hb+Mb] (μM) and deoxy [Hb+Mb] (μM) and calculate changes in $\text{S}{\text{t}}_{{\text{O}}_2}$ (%). Finally, by combining the relative changes in oxy [Hb+Mb] and deoxy [Hb+Mb] provided by NIRS and the relative changes in BFI from DCS, the relative change in skeletal muscle $\text{M}{\text{R}}_{{\text{O}}_2}$ was calculated (9). For detailed calculations corresponding to each of these DCS- and NIRS-derived variables, please refer to (3, 39, 46).

Forearm deep venous blood sampling.

In a subset of individuals (n = 3), an 18-gauge intravenous catheter was inserted retrograde to venous blood flow into a deep forearm vein, as previously described (39, 46). Briefly, blood samples were taken during the resting baseline, the final 30 s of 3 min of rhythmic handgrip exercise, and the final 30 s of 2 min of rhythmic handgrip exercise with brachial artery compression to measure changes in venous oxygen saturation ($\text{S}{\text{V}}_{{\text{O}}_2}$). A 3-mL discard was drawn before the 2-mL blood sample. A 2-mL saline flush followed each sample to prevent the catheter from clotting. An I-STAT analyzer (I-STAT 1, Abbott Point of Care, Princeton, NJ) was used to analyze the whole blood sample for forearm $\text{S}{\text{V}}_{{\text{O}}_2}$ using CG8+ test cartridges (Abbott Point of Care, Princeton, NJ). All blood samples were analyzed in duplicate and averaged.

Data analysis and calculated variables.

Heart rate, blood pressure, Doppler ultrasound, and NIR-DCS-derived variables were measured during rest, exercise, and postexercise recovery. To minimize muscle-fiber motion artifact during exercise, NIR-DCS measurements were only made and analyzed during the relaxation phase of the handgrip duty cycle, as previously described (3, 21, 39, 46). Accordingly, radial artery blood flow measurements were also made and analyzed during the relaxation phase of the handgrip duty cycle to match DCS. With the exception of data presented continuously for representative individuals, mean resting baseline values represent a 1-min average of the final minute of a 2-min baseline and exercise values represent a 1-min average of the final minute for each experimental condition (exercise, exercise with compression, exercise following compression release) within the 7 min of uninterrupted rhythmic handgrip exercise.

Heart rate, blood pressure, and Doppler data were analyzed offline by a single observer (W. J. Tucker) using LabChart Pro Software (ADInstruments, Colorado Springs, CO). For real-time arterial blood pressure waveforms, a peak-detection algorithm was used to identify systolic and diastolic components of each arterial wave. Thereafter, mean arterial pressure (MAP, mmHg) was calculated on a beat-to-beat basis. MAP was calculated as the mean pressure across a continuous average of arterial waveforms (1-min average of final min of resting baseline and 1-min average of final min of each experimental exercise condition). Radial artery blood velocity (MBV, cm/s) was calculated as the beat-to-beat average from the real-time Doppler ultrasound waveforms. Radial artery blood flow (RBF) was estimated as [MBV × π (radial artery diameter/2)²] × 60. Forearm vascular conductance was calculated as RBF/MAP × 100 mmHg.

NIRS and DCS.

All raw NIRS and DCS data were analyzed in MATLAB (Version R2016A, MathWorks, Natick, MA) and exported to Microsoft Excel (Microsoft Corporation, Redmond, WA) for subsequent statistical analyses.

Statistical Analysis

Data were analyzed using SPSS Software (SPSS 24.0; IBM Corporation, Armonk, NY). All data are reported as means ± SD. P values of < 0.05 were considered statistically significant. All data were initially assessed for normality using the Shapiro-Wilk test. Once normality was verified, a one-way repeated measures analysis of variance (ANOVA) was used to test for differences between rest, exercise, and exercise during and after release of arterial compression for all physiological and NIR-DCS variables. For all repeated measures ANOVA testing, if the sphericity assumption was violated (Greenhouse-Geisser ε < 0.75), the Greenhouse-Geisser adjustment was used to interpret main effects. In the event of a significant main effect, a Bonferroni correction post hoc analysis was performed to determine where significant differences were.

RESULTS

Seventeen subjects volunteered to participate in the study. Two of these subjects were excluded from the final analysis because of technical difficulties associated with successfully implementing the 2-min arterial compression during handgrip exercise. A failure to adequately compress the artery was defined as no reduction in real-time radial artery blood velocity using our online data acquisition system and later confirmed in both subjects as no reduction in radial artery blood flow or DCS-derived BFI during postprocessing offline analysis (data not shown). In a third subject, technical difficulties prevented data collection during the final 2 min of exercise (following release of the arterial compression). Therefore, 14 subjects were included in the final analyses (subject characteristics in Table 1).

Table 1.

Subject characteristics

Variable	Mean ± SD
Age, yr	25 ± 3
Height, cm	177.4 ± 5.8
Weight, kg	81.0 ± 15.5
BMI, kg/m2	25.6 ± 3.7
Systolic blood pressure, mmHg	121 ± 10
Diastolic blood pressure, mmHg	74 ± 9
Heart rate, beats/min	67 ± 8
MVC, kg	47 ± 8
Absolute workload at 30% MVC, kg	14 ± 2

Data presented as means ± SD; n = 14 men. BMI, body mass index; MVC, maximal voluntary contraction.

Radial Artery Blood Flow and Microvascular Tissue Perfusion

Doppler-derived radial artery blood flow increased significantly with handgrip exercise (Fig. 2A, Table 2), with DCS-derived BFI also following this pattern of response (P < 0.001) (Fig. 2B). Increases in radial artery blood flow were driven by a ~3.5-fold increase in blood velocity during handgrip exercise (Table 2). Compression of the brachial artery during handgrip exercise caused a significant drop in radial artery blood flow (Fig. 2A, Table 2), with DCS-derived BFI also following this pattern of response (P = 0.01) (Fig. 2B, Table 2). The reduction in radial artery blood flow was driven by a 44 ± 15% decrease in blood velocity associated with the upstream arterial compression (Table 2). Release of the compression caused a rapid increase in both radial artery blood flow (Fig. 2A) and DCS-derived BFI (P = 0.01) (Fig. 2B). In fact, the release of the arterial compression during handgrip exercise augmented radial artery blood flow (P = 0.001) and DCS-derived BFI (P = 0.06) to higher levels than those observed during exercise before the compression (Fig. 2, A and B; Table 2). Individual subject responses for radial artery blood flow, DCS-derived BFI, and NIRS-derived variables are illustrated in Fig. 3.

Fig. 2.

Effect of brachial artery compression and release during 30% maximal voluntary contraction (MVC) rhythmic handgrip exercise on radial artery blood flow (BF; mL/min) (measured by Doppler Ultrasound) (A), microvascular tissue perfusion [measured by diffuse correlation spectroscopy (DCS)-derived blood flow index (BFI; %)] (B), near-infrared spectroscopy-derived tissue saturation (Sat.; %) (C), and metabolic rate of oxygen ($\text{M}{\text{R}}_{{\text{O}}_2}$) (D) responses in a representative subject (line graphs on the left) and for the entire group (bar graphs on right that depict mean ± SD). *P < 0.05 compared with rest; ***P < 0.001 compared with rest; †P < 0.05 compared with both exercise (EX) periods without compression; ‡P < 0.05 compared with first EX period before compression; §P < 0.05 compared with both the first EX period before compression and EX plus compression. Data are means ± SD; n = 14 men.

Table 2.

Hemodynamic and NIR-DCS responses to 30% maximal voluntary contraction rhythmic handgrip exercise with and without arterial compression

	Rest	Exercise	Exercise + Compress	Exercise	Time Effect P Value
Hemodynamics
MBV, cm/s	6.9 ± 2.8	23.1 ± 4.5***	13.1 ± 4.5*†	25.8 ± 5.7***	<0.001
Radial artery diameter, cm	0.24 ± 0.03	0.25 ± 0.03***	0.26 ± 0.03***	0.27 ± 0.03***‡	<0.001
RBF, mL/min	20.1 ± 11.1	71.7 ± 24.4***	39.7 ± 13.6*†	86.7 ± 31.7***§	<0.001
MAP, mmHg	87 ± 8	94 ± 10*	100 ± 10*‡	99 ± 9*‡	<0.001
FVC, mL·min−1·100 mmHg−1	24 ± 14	76 ± 22***	40 ± 13*†	87 ± 28***‡	<0.001
Heart rate, beats/min	62 ± 9	67 ± 7	69 ± 8*	69 ± 7*‡	0.02
NIR-DCS
BFI, %	106.1 ± 7.7	261.1 ± 81.0***	203.8 ± 86.9*†	307.9 ± 113.2***	<0.001
Oxy [Hb+Mb], μM/L	69.4 ± 7.6	64.6 ± 9.5*	59.9 ± 10.2*‡	63.7 ± 9.1*	<0.001
Deoxy [Hb+Mb], μM/L	29.4 ± 6.0	39.4 ± 11.6*	46.5 ± 14.7***‡	44.9 ± 11.9***‡	<0.001
$\text{S}{\text{t}}_{{\text{O}}_2}$, %	70.3 ± 5.3	62.4 ± 9.8*	56.8 ± 10.9***‡	58.9 ± 9.2***‡	<0.001
$\text{M}{\text{R}}_{{\text{O}}_2}$, %	105.4 ± 9.6	332.4 ± 131.7***	296.5 ± 147.4*	433.9 ± 209.1***§	<0.001

Data presented as means ± SD; n = 14 men. BFI, blood flow index; deoxy [Hb+Mb], deoxygenated hemoglobin/myoglobin; FVC, forearm vascular conductance; MAP, mean arterial pressure; MBV, radial artery blood velocity; $\text{M}{\text{R}}_{{\text{O}}_2}$, metabolic rate of oxygen; NIR-DCS, near-infrared diffuse correlation spectroscopy; oxy [Hb+Mb], oxygenated hemoglobin/myoglobin; RBF, radial artery blood flow; $\text{S}{\text{t}}_{{\text{O}}_2}$, tissue oxygen saturation.

^*P < 0.05 compared with rest;

^***P < 0.001 compared with rest;

^†P < 0.05 compared with both exercise periods without compression;

^‡P < 0.05 compared with first exercise period before compression;

^§P < 0.05 compared with both the first exercise period before compression and exercise plus compression.

Fig. 3.

Individual responses to brachial artery compression and release during 30% maximal voluntary contraction (MVC) rhythmic handgrip exercise on radial artery blood flow (BF; measured by Doppler Ultrasound) (A), microvascular tissue perfusion [measured by diffuse correlation spectroscopy (DCS)-derived blood flow index (BFI)] (B), near-infrared spectroscopy-derived tissue saturation (Sat.; %) (C), and metabolic rate of oxygen ($\text{M}{\text{R}}_{{\text{O}}_2}$) (D). Data presented as percentage change from initial exercise values (100%) before compression and release; n = 14 men.

Central Hemodynamic Responses

As expected, MAP increased significantly (+7 ± 7 mmHg, P = 0.02) in response to handgrip exercise (Table 2). Compression of the brachial artery caused further augmentation of arterial pressure (+6 ± 3 mmHg, P = 0.001) during handgrip exercise; however, there was no change in MAP after the compression was released (Table 2). Forearm vascular conductance followed changes in radial artery blood flow with each condition, with a significant increase in conductance in response to handgrip exercise (P < 0.001), that was reduced with compression (P < 0.001) and further increased with subsequent compression release (P = 0.001) (Table 2).

Forearm Tissue Oxygenation

As expected, $\text{S}{\text{t}}_{{\text{O}}_2}$ decreased −7.9 ± 6.2% (P = 0.002, Fig. 2C and Table 2) during handgrip exercise. Compression of the brachial artery caused a further reduction of $\text{S}{\text{t}}_{{\text{O}}_2}$ (−5.6 ± 3.8%, P = 0.001), with subsequent compression release nonsignificantly increasing $\text{S}{\text{t}}_{{\text{O}}_2}$ (+2.1 ± 3.6%, P = 0.31) (Fig. 2C and Table 2).

The changes in $\text{S}{\text{t}}_{{\text{O}}_2}$ during handgrip exercise were driven by increases in deoxy [Hb+Mb] and reductions in oxy [Hb+Mb] (Table 2). Similarly, further reductions in $\text{S}{\text{t}}_{{\text{O}}_2}$ observed during the compression were caused by associated increases in deoxy [Hb+Mb] and reductions in oxy [Hb+Mb] (Table 2).

The pattern of change observed with $\text{S}{\text{t}}_{{\text{O}}_2}$ was confirmed in all three of the participants instrumented with a retrograde deep venous catheter in the exercising arm, with $\text{S}{\text{V}}_{{\text{O}}_2}$ decreasing from 66 ± 5% at rest to 36 ± 3% during handgrip exercise, followed by a further decline to 26 ± 6% during handgrip with brachial artery compression.

Forearm Skeletal Muscle Relative Oxygen Consumption

$\text{M}{\text{R}}_{{\text{O}}_2}$ increased significantly in response to handgrip exercise (Fig. 2D and Table 2). However, compression of the brachial artery during exercise did not cause a significant reduction in $\text{M}{\text{R}}_{{\text{O}}_2}$ (P = 0.83) (Fig. 2D and Table 2), in part due to compensatory reductions in $\text{S}{\text{t}}_{{\text{O}}_2}$ in opposition to reduced microvascular flow. Release of the compression during exercise caused an increase in $\text{M}{\text{R}}_{{\text{O}}_2}$ that was significantly higher than exercise with compression (P = 0.01) and exercise before compression (P = 0.04).

DISCUSSION

This study sought to extend prior investigations using NIR-DCS to evaluate microvascular determinants of skeletal muscle oxygen consumption during exercise by experimentally restricting oxygen delivery using an isolated arterial compression model. As hypothesized, NIR-DCS effectively tracked acute reductions in skeletal muscle oxygen delivery (i.e., microvascular perfusion measured by DCS), along with compensatory reductions in tissue oxygen saturation (measured by NIRS), during handgrip exercise. As a result, skeletal muscle oxygen consumption was preserved because of the compensatory reduction in $\text{S}{\text{t}}_{{\text{O}}_2}$ in opposition to the reduced microvascular flow. Together, these proof-of-concept data help to establish NIR-DCS as an effective tool to assess the determinants of skeletal muscle oxygen consumption at the level of the microvasculature during exercise.

Prior studies conducted in our laboratory (39, 46) have shown that NIR-DCS is able to measure the determinants of skeletal muscle oxygen consumption at the level of the microvasculature during exercise. Indeed, NIR-DCS demonstrated good concurrent validity and directionally similar responses to conventional measures of oxygen delivery (Duplex Ultrasound) and utilization (venous oxygen saturation acquired through retrograde catheter) during both fixed-workload (46) and incremental (39) handgrip exercise. Despite these advancements, these validation studies did not fully address the sensitivity of NIR-DCS to detect changes in oxygen delivery and extraction when convective oxygen delivery is reduced. The current study addresses this knowledge gap by using isolated arterial compression to highlight the temporal responsiveness of NIR-DCS during an acute physiologic perturbation. The major strengths of this isolated arterial compression model, in contrast to the more conventional cuff occlusion approach, are twofold. First, it avoids venous pooling, which can independently alter arterial hemodynamics (12, 27). Second, it avoids the build-up of metabolic byproducts, which are known to stimulate the metaboreflex and alter arterial hemodynamics (1, 2, 10, 34, 47). The data show a marked reduction in microvascular perfusion of the exercising muscle with arterial compression, together with a concomitant increase in muscle oxygen extraction, as measured by deoxyhemoglobin/myoglobin (and confirmed in a subset of participants using a deep venous catheter to measure $\text{S}{\text{V}}_{{\text{O}}_2}$ in the exercising arm). In accordance with the expected physiologic response, these responses resolved after release of the arterial compression, further validating NIR-DCS as a sensitive tool for evaluating the determinants of muscle oxygen consumption at the level of the microvasculature.

By design, brachial artery compression caused a robust reduction in downstream radial artery blood flow during exercise, which was quite uniform across all individuals (Fig. 3A). In contrast, although microvascular perfusion (measured by DCS) was also significantly reduced in response to arterial compression, the magnitude and direction of change was not nearly as uniform across individuals (Fig. 3B), as was observed in the radial artery. For example, 3 of 14 individuals were able to fully maintain microvascular perfusion despite marked reductions in radial artery blood flow during the arterial compression. Moreover, the absolute magnitude of change is visibly lower in the exercising microvasculature compared with the conduit (radial) artery. Although the exact mechanism responsible for this disparity between macro- and microvascular blood flow is beyond the scope of this investigation, it is interesting to speculate that blood flow redistribution (31, 32, 43), by way of sympatholysis and/or other vasoregulatory control mechanisms (e.g., myogenic autoregulation) (13, 24, 25, 35), may be contributing. Indeed, the radial artery is supplying muscles adjacent to those primarily performing handgrip exercise, which presumably do not secrete the same vasoregulating substances as the active muscles or have the same metabolic demand. However, caution is warranted when making this comparison because the flexor digitorum profundus is fed not by the radial artery but rather the anterior interosseous artery, which was not imaged in this investigation due to direct competition for available imaging space between the NIR-DCS and Doppler ultrasound.

That we observed skeletal muscle oxygen extraction increase in response to an acute reduction in microvascular oxygen delivery provides proof-of-concept that NIR-DCS can effectively assess the determinants of skeletal muscle oxygen consumption and helps to establish the overall utility of this technology. We have previously reported a strong linear relationship between DCS-derived blood flow index and NIRS-derived deoxyhemoglobin/myoglobin [a surrogate measure of microvascular oxygen extraction in skeletal muscle (15, 18–20, 29)] during incremental handgrip exercise. Here we dissociate these two parameters, demonstrating the interdependence of each measure, as they relate to muscle oxygen consumption. Indeed, where skeletal muscle oxygen delivery was allowed to increase in proportion to handgrip intensity in our previous study (39), in this current investigation absolute workload was held constant while oxygen delivery was reduced. Under these conditions, where oxygen delivery and utilization are unbalanced, we clearly observe a compensatory increase in oxygen extraction when oxygen delivery is reduced, evidenced by a further reduction in $\text{S}{\text{t}}_{{\text{O}}_2}$ when arterial compression was superimposed onto handgrip. That these data were confirmed by directionally similar responses in $\text{S}{\text{V}}_{{\text{O}}_2}$ in subjects instrumented with a deep venous catheter supports this interpretation and extends our prior observations. Taken together, NIR-DCS is now poised to extend our understanding of exercise intolerance in clinical populations. Indeed, the demonstrated sensitivity of NIR-DCS in the present study holds great promise for precision-based medicine and targeted therapies designed to treat specific pathophysiologic mechanisms (e.g., impaired microvascular function, oxygen utilization, or both).

This study is not without limitation. The effectiveness of manual brachial artery compressions to reduce blood flow to the exercising forearm were confirmed by measuring changes in downstream radial artery blood flow with Duplex Ultrasound. As mentioned above, although imaging the anterior interosseous artery would have provided better insight because it directly perfuses the flexor digitorum profundus, we were unable to image the anterior interosseous artery due to space limitations associated with placement of the NIR-DCS device in a similar area of the forearm. Accordingly, caution is warranted when interpreting comparisons made between the radial artery Doppler data and DCS-derived blood flow index. Additionally, we cannot be sure that the finger pressure applied to the brachial artery during each compression was identical between subjects. However, the reductions in radial artery blood flow caused by brachial artery compressions were quite uniform across individuals (Fig. 3A) and in line with our a priori defined target.

To optimize signal quality, the NIR-DCS probe was carefully placed over the belly of flexor digitorum profundus by an experienced operator before each experiment. This is, however, a small muscle with multiple neighboring muscle groups. As such, we cannot rule out the possibility of competing signals from other surrounding forearm muscles during handgrip exercise. NIRS measures changes in tissue oxygenation in small arterioles, capillaries, and venules within skeletal muscle. Therefore, the relative contributions of each of these vessels to the optical signal during handgrip exercise at 30% maximal voluntary contraction is unclear. However, it has been shown that the majority (~84%) of the skeletal muscle microvascular volume comprises capillaries, with the remainder split between the arterioles and venules (4, 36). As such, it is typically inferred that changes in optical signal strength observed with NIRS during exercise likely represent changes in microcirculatory oxygen transport (33). Finally, this study only included healthy young men because of technical concerns related to imaging radial artery blood flow during handgrip exercise in women. Future studies are needed to assess NIR-DCS responses in women, older adults, and diseased populations who have impaired muscle blood flow (e.g., heart failure, peripheral arterial disease).

Perspectives and Significance

These proof-of-concept data help to further validate NIR-DCS as an effective tool to assess the determinants of skeletal muscle oxygen consumption at the level of the microvasculature during exercise.

GRANTS

This research was completed in the Potratz Family Clinical Research Laboratory, funded by a National Institutes of Health Grant No. R15 HL140989-01, and the Nagy Family Endowment. W. J. Tucker was financially supported by the American Heart Association (AHA) Postdoctoral Fellowship Grant (AHA Award No. 18POST33990210). M. J. Haykowsky is financially supported by the Moritz Chair in Geriatrics at the University of Texas at Arlington.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

W.J.T., R.R., R.F.B., M.J.H., and M.D.N. conceived and designed research; W.J.T., R.R., D.T., B.S., and M.D.N. performed experiments; W.J.T. analyzed data; W.J.T., R.R., D.T., R.F.B., M.J.H., F.T., and M.D.N. interpreted results of experiments; W.J.T. prepared figures; W.J.T. and M.D.N. drafted manuscript; W.J.T., R.R., D.T., B.S., R.F.B., M.J.H., F.T., and M.D.N. edited and revised manuscript; W.J.T., R.R., D.T., B.S., R.F.B., M.J.H., F.T., and M.D.N. approved final version of manuscript.