Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

5.1 Introduction

Muscle oxidative metabolism has often in the past been examined by traditional analytical biochemistry based on obtaining biopsy samples. Invasive techniques such as intravascular catheterization have provided information on muscle oxygen status at rest and during exercise [1]. Myoglobin O2 saturation and reduced nicotinamide adenine dinucleotide (NADH) redox state can be detected using freeze-clamped tissue [2]. The disadvantages of the invasive approach include poor time resolution due to repeated measurements and, of course, the invasive nature of the sampling method, so there is a need for noninvasive approaches to measure muscle oxygen status during exercise. NIRS has been utilized to evaluate skeletal muscle O2 dynamics and energetics during exercise. Application of this technology has focused on the validity and calibration of measurements, and biochemical, physiological, and pathological research of muscle oxidative metabolism [3–6]. The primary reason why NIRS methodology is useful for evaluating muscle is the visibility of heme molecules and their oxygen-dependent characteristics.

Visible light has been used to monitor changes in tissue oxygenation since the 1930s [7]. Mitochondrial NADH signal measured using optical methods has shown a rapid change caused by electrical muscle stimulation (within a fraction of a second), indicating coupling of muscle contraction and mitochondrial activity [8, 9]. Jöbsis [10] then discovered that NIR light easily penetrates the skull, which set the stage for recent application of NIRS to scientific research in brain and muscle as well as in various clinical settings. A muscle NIRS system was then developed [11–13] that served as one of the first models to provide an opportunity for noninvasive and portable clinical muscle research.

5.2 Biochemistry and Physiology of Muscle Oxidative Metabolism

Skeletal muscle employs three major biochemical processes to synthesize ATP: oxidative phosphorylation (32–36 ATPs), anaerobic glycolysis (2 ATPs), and the PCr pathway (1 ATP). During low- to moderate-intensity exercise, like that involved in normal daily activities, skeletal muscle intensively relies on oxidative metabolism. Skeletal muscle O2 consumption (VO2) can be elevated 50-fold with up to tenfold abrupt increases in O2 delivery (DO2). The net oxidative energy pathway in muscles can be described by the following equation [14]:

$$ 3\mathrm{ ADP} + 3\mathrm{ Pi} + \mathrm{ NA}\mathrm{ DH} + {{\mathrm{ H}}^{+}} + {\scriptstyle 1} / {\scriptstyle 2}{{\mathrm{ O}}_2} = 3\mathrm{ ATP} + \mathrm{ NA}{{\mathrm{ D}}^{+}} + {{\mathrm{ H}}_2}\mathrm{ O}, $$
(5.1)

where ADP is adenosine diphosphate, Pi is inorganic phosphate, ATP is adenosine triphosphate, and NAD+ is the nicotinamide adenine dinucleotide.

The kinetic control model of respiration that describes metabolic rate as a function of regulatory substrate concentrations using the Michaelis-Menten equation is written as [15]

$$ V/{V_{\mathrm{ m}}} = 1/\left( {1+{k_1}/\mathrm{ ADP} + {k_2}/\mathrm{ Pi}+{k_3}/{{\mathrm{ O}}_2}+{k_4}/\mathrm{ NADH}} \right), $$
(5.2)

where V is observed velocity, V m is maximum velocity, k 1−4 represent affinity constants for various substrates, Pi is inorganic phosphate, PCr is phosphocreatine, and NAD is the nicotinamide adenine dinucleotide. The in-vivo mitochondrial concentrations of ADP, Pi, O2, and NADH are around 20, 1,000, 1, and 100 μM, respectively, and the in-vitro K m (half-maximum velocity) values for ADP, Pi, O2 (cytochrome aa3 for O2), and NADH are 20, 300, 0.1, and ~10 μM, respectively. The in-vitro data indicate that the primary candidate for metabolic control is ADP [15]. However, it should be noted that no final agreement has been reached as to what is the metabolite controller of mitochondrial respiration.

For the phosphorylation cycle to function, ADP and Pi must be transported into the mitochondrial matrix, and then ATP is transported from there back to the cytoplasm. This reaction is catalyzed by the adenine nucleotide translocase of the inner mitochondrial membrane. It has thus been proposed that the rate of mitochondrial respiration can be determined by the rate of adenine nucleotide translocation and that, therefore, the [ATP]/[ADP] ratio regulates the respiratory rate under physiological conditions [16]. Holian et al. [17] generalized metabolic regulation to include ATP as a controller, as well as ADP and inorganic phosphate. The near-equilibrium hypothesis states that there is a correlation among the cellular [ATP]/[ADP][Pi] ratio, the mitochondrial [NAD]/[NADH] ratio, and respiration rate. Respiration rate would thus be dependent on [ATP]/[ADP][Pi] under conditions of Pi (~1 mM) and ATP (~6 mM). In general, ATP is maintained constant during submaximal exercise by creatine kinase equilibrium via the creatine shuttle in muscle, and ATP is not operative in respiration regulation.

The thermodynamic regulation model [18], an analogue of the electrical model for respiration control, in which PCr can be directly related to respiration rate, should thus be applicable to mitochondrial respiration (see Fig. 5.1). The current (I cy) represents the cytosolic adenosine triphosphatase (ATPase) rate, and voltage V 0 represents the free energy potential available in the mitochondria. Capacitance C is due to the creatine kinase reaction, and PCr is analogous to the stored charge on a capacitor. Resistor R m is a function of the number and properties of the mitochondria in the cell, and current I rm is the rate of oxidative phosphorylation [18], where I rm = I c + I cy and I c = CdV 0/dt. The free energy of ATP hydrolysis (ΔG ATP) becomes a function of ln{[PCr]/(Total-Cr – [PCr])2} (see [18] for details). To a good approximation, ln{[PCr]/(Total-Cr – [PCr])2} is linear for PCr between 20% and 70% of total-Cr. The cellular mechanism for mitochondrial respiration and oxygen delivery is illustrated in Fig. 5.2. We have confirmed that both the kinetic and thermodynamic control models of mitochondrial respiration are operative in working human muscles [19].

Fig. 5.1
figure 00051

Electrical analog model with a first-order linear system for respiratory control in muscle at submaximal oxidative rates. The current Icy represents the cytosolic adenosinetriphosphatase (ATPase) rate, the voltage V0 represents the free energy potential available in the mitochondria. Capacitance C is due to the creatine kinase reaction, and PCr is analogous to stored charge on a capacitor. Resistor Rm is a function of the number and properties of the mitochondria in the cell, and current Irm is the rate of oxidative phosphorylation [18]

Full size image
Fig. 5.2
figure 00052

Cellular mechanism for mitochondrial respiration and oxygen delivery to muscle. Mitochondrial respiration is controlled by the concentrations of ADP, Pi, and Ca2+, the NAD/NADH ratio, and so on. Oxygen delivery (blood flow) is controlled by sympathetic vasoconstriction, varying metabolites, NO, muscle pump, etc. ATP is synthesized via the three main steps – namely, oxidative phosphorylation (32 ATPs), anaerobic glycolysis (2 ATPs), and PCr pathway (1 ATP). Mit mitochondrion, NA noradrenaline, NO nitric oxide: HbO 2 oxyhemoglobin, ATP adenosine triphosphate, ADP adenosine diphosphate, Pi inorganic phosphate, PCr Phosphocreatine, NAD nicotinamide adenine dinucleotide, NADH reduced nicotinamide adenine dinucleotide, VO 2 oxygen consumption: DO 2 , oxygen delivery

Full size image

5.3 Principles of Muscle NIRS

Light within the NIR region (700–3,000 nm) shows less scattering and thus better penetration into biological tissue than visible light. However, light absorption by water limits tissue penetration at longer wavelengths (above ~900 nm), leaving the 700–900 nm window for biological monitoring. The major absorbing compounds in this wavelength region are vascular hemoglobin (Hb), intracellular Mb, and mitochondrial cytochrome c oxidase [10]. Therefore, NIRS measurements rely on O2-dependent absorption changes that occur in the heme-containing compounds.

The most common commercially available NIRS devices use single-distance continuous-wave light (NIRSDCWS). To calculate changes in oxy-Hb/Mb, deoxy-Hb/Mb, or total-Hb/Mb, the equation of a two- or multiple-wavelength (770–850 nm) method can be applied according to the following Beer-Lambert Law:

$$ \Delta \mathrm{ OD}=-\mathrm{ lo}{{\mathrm{ g}}_{\mathrm{ e}}}(I/{I_0})=\varepsilon\mathrm{ PL}\Delta\left[ C \right], $$
(5.3)
$$ \Delta\left[ C \right]=\Delta\mathrm{ OD}/\varepsilon\mathrm{ PL}\text{,} $$
(5.4)

where ε is the extinction coefficient (OD/cm/mM) (= constant), PL is pathlength, [C] is concentration of absorber (mM), I is detected light intensity, I 0 is incident light intensity, and OD is optical density.

A major limitation of NIRSDCWS is that it provides only relative values of tissue oxygenation. The main reason for this lack of quantification by NIRSDCWS is the unknown path of NIR light through biological tissue. The pathlength of NIR light can be measured using other optical approaches, e.g., time-resolved spectroscopy (NIRTRS) [20–23] and phase-modulation spectroscopy (NIRPMS) [23–25]. These approaches provide experimentally observed values of oxygenated and deoxygenated Hb/Mb and Hb/Mb O2 saturation (SO2) in skeletal muscle.

With NIRTRS, light is input to tissues in picosecond pulses, and the temporal point spread function (TPSF) is detected as a function of time with picosecond resolution [20, 26]. When a time-correlated single-photon counting system is employed, the distribution of arrival times of a large number of photons follows the TPSF. In principle, the optical properties of tissue can be determined from the TPSF analogous to electronic engineers determining the equivalent circuit of a black box from its response to a voltage or current impulse [26]. Optical properties such as absorption and scattering coefficients can be determined after having an accurate model for light transport in tissues, and represent averaged values for these parameters that best match the monitored data [26].

In NIRPMS instruments, the light source is intensity modulated at radio frequencies (RFs), and measurement is made not only of detected light intensity, but also of its phase shift and modulation depth. Detection of small changes in phase shift and modulation depth is difficult, so almost all these instruments employ some downconversion scheme to bring the RF signal down to an intermediate frequency or audio frequency. In general, two different setups have been used. First, detector gain is modulated at a reference frequency offset by a few kHz from the light source modulation frequency, resulting in direct demodulation within the detector. Second, the detector signal is fed into a double-balanced mixer together with the reference RF, which requires a simpler scheme and much lower RF powers (~10 mW) than the former [26]. The phase shift for typical tissues at frequencies below 200 MHz is linearly related to average optical pathlength. Knowing phase shift enables direct calculation of changes in oxy- and deoxy-hemoglobin concentrations.

Spatially resolved NIRSRCWS (NIRSRCWS) uses multiple light sources coupled to one detector. In NIRSRCWS, the slope of light attenuation versus distance is measured at a distance from the light input, from which the tissue oxygenation index (TOI) can be calculated using photon diffusion theory. Therefore, NIRSRCWS provides relative changes in Hb/Mb and absolute values of SO2 [27].

With NIRSDCWS measurements, there is an assumption that pathlength does not show significant change during exercise, recovery, and other intervention periods; otherwise, the values obtained are either underestimated or overestimated, as shown in Eqs. 5.3 and 5.4. However, the pathlength of light might vary due to variations in tissue composition, blood volume, and muscle geometry. During and after arterial occlusion, changes in pathlength for forearm muscle ranged from −8.3% to −2.1% at 780 nm and from −2.2% to 0.74% at 830 nm [22]. Changes in pathlength were less than 10% during arterial occlusion with maximum voluntary contraction (MVC) [21]. The change in pathlength at 780 and 830 nm during forearm exercise from moderate to high intensity and hyperemic recovery was small (<8%) relative to the resting level [28]. The differential pathlength factor (DPF) in thigh muscle decreased slightly, but significantly, from baseline (DPF at 690 nm = 5.22, at 830 nm = 4.49, on average) to peak cycle exercise (DPF at 690 nm = 4.88; at 830 nm = 4.27 on average) (−6.5% at 690 nm and −4.9% at 830 nm) [29]. For an accurate evaluation of muscle oxygenation during arterial occlusion, exercise, and recovery, changes in pathlength should be extensively examined in a wide range of exercise mode/intensity.

5.4 Quantification of In-Vivo NIRS Measurements

5.4.1 Mb/Hb

A recent study reported that the spectral peak appearing around 760 nm in deoxygenated muscle tissue included both deoxy-Hb and deoxy-Mb signals, and that the peak position shifted as a linear function of the relative contributions of Hb and Mb to the optical spectra. The investigators reported that Hb accounts for 87% of the overall signal in mouse muscle but for only 20% in human skeletal muscle [30]. 1H-magnetic resonance spectroscopy (1H-MRS) has also been employed to detect the deoxy-Mb signal by measuring the N-delta proton of proximal histidine [31, 32]. Simulation results based on combined measurements of 1H-MRS and NIRS concluded that the overall NIR signal is greater than ~50% Hb [33]. There is also a report indicating differential kinetics between MbO2 and NIRS measurements. Another study [34] found that, at rest, human tibialis anterior intramuscular O2 stores (measured by appearance of 1H-NMR deoxy-Mb signal during cuff occlusion) began to decrease after 1 min and that maximal Mb desaturation was achieved at about 6.5 min. Richardson et al. found that deoxy-Mb levels did not appear to increase after 2 min of cuff ischemia [35], while NIRS-measured oxygen signals declined almost immediately and reached near-maximal levels at 4 min [19]. In contrast, It is reported that deoxy-Mb signal correlated with NIR deoxygenation in a study using 1H-MRS and NIRS [36]. These differing conclusions highlight the need for additional study to clarify not only the contribution of Mb to the NIR signal, but also the kinetics and amount of Mb desaturation during exercise under different conditions.

5.4.2 Effect of Multiple Layers

The light path from a light source to the detector follows a banana-shaped curve in which penetration depth into tissue is approximately equal to half the distance between the light source and detector [12]. If light source–detector separation was set at 3 cm, penetration depth was equal to 1–2 cm and measured volume ~4 cm [12, 37]. Hampson et al. studied how the signal from skin might interfere with muscle NIRS measurement, employing NIRSDCWS, changes in skeletal muscle O2 content kinetics and skin deoxygenation during 10-min ischemia in human forearm [13]. They found that muscle O2 content began to fall at onset of arterial occlusion. Increased pressure caused skin oxygenation to rapidly fall within 2–3 min, while muscle oxygenation declined within 6 min. Muscle O2 content was completely depleted (functional anoxia) about 6 min after onset of arterial occlusion. This indicated that the time course of skin deoxygenation changes during ischemia correlate poorly with changes in skeletal muscle. Heating of the thigh at 37°C and 42°C caused a marked increase in cutaneous vascular conductance (CVC) at rest and with exercise, and an increase in SO2 by several percentage points at rest, but not during exercise. One might thus suppose that skin vasodilatation has a marginal influence on NIRS experiments where skin blood flow can change markedly [38]. A fixed source–detector spacing of 4 or 5 cm ensures more accurate quantitation of the oxygenation changes occurring at the muscular level and minimizes the influence of skin vasculature [39].

Subcutaneous adipose tissue thickness has a substantial influence on signal intensity [40–42]: signal intensity for 5-mm fat thickness mm was reduced by ~0.2 (80% of signal for 0 mm) with a source–detector separation of 30–40 mm, and was further reduced by 0.3–0.6 with a separation of 15–20 mm [41, 43]. Other work documented the influence of adipose tissue thickness ranging from 0 to 15 mm with a source–detector separation of 15–40 mm. The curve was based on the results of Monte Carlo simulation as well as in-vivo experiments [41, 43]:

$$ S=\exp \left( {-{{{\left( {h/{A_1}} \right)}}^2}} \right) - {A_2}G(\alpha, \beta ), $$
(5.5)

where S is normalized measurement sensitivity, h is adipose tissue thickness, G(α,β) is a gamma distribution, and constants A 1, A 2, α, and β at a source–detector separation of 15 mm are 6.9, 1.15, 7.86, and 0.80, respectively. Considering that the value of S is mainly determined by h, then corrected values are obtained by dividing the measured values by S.

The influence of adipose tissue thickness in human muscle on NIR spectra was also studied by Monte Carlo simulation in a two-layer model using phantom experiments [44]. The results of this work suggested that subject-to-subject variation in adipose tissue optical coefficients and thickness can be ignored if the thickness is less than 5 mm when the source–detector separation is 40 mm. Subcutaneous adipose tissue thickness should thus be measured for muscle NIRS measurement. In addition, it is preferable to use a 4- to 5-cm source–detector separation for measurements at the muscle level without attenuating signal intensity.

5.5 In-Vitro and In-Vivo Calibration of NIRS Measurements

Several researchers have demonstrated and justified their equations by in-vitro and in-vivo experiments. Chance et al. [12] adopted an “in vitro yeast and blood model” for evaluation of an NIR apparatus. Their model system included intralipid as a scatterer (0.5–1.0%) and blood at concentrations of 10–200 μM hemoglobin. O2 bubbling caused reoxygenation, while cessation of bubbling caused deoxygenation of hemoglobin due to yeast respiration. Addition of blood correlated with increased in-vivo blood volume. Subtracted signals at 760 and 850 nm showed a linear relationship with hemoglobin deoxygenation [12]. Wilson et al. [45] demonstrated a linear relationship between near-infrared spectroscopic measurements and venous hemoglobin saturation in an animal model. It should be noted that the contribution of Hb/Mb should be different for human and mouse muscle [30]. Mancini et al. found muscle oxygenation and venous O2 saturation (SvO2) of human forearm muscles to be closely related during exercise [46]. They also demonstrated that muscle oxygenation decreased with intravascular norepinephrine administration and increased with administration of a vasodilator (nitroprusside). A good association was found between regional quadriceps oxygenation at three different measurement sites and SvO2 during one-legged dynamic knee extension exercise under normoxic conditions [47]. A good relationship was also found between vastus lateralis oxygenation and femoral arteriovenous O2 difference (a-vO2D) during one-legged dynamic knee extension exercise under normoxic as well as hypoxic and hyperoxic conditions [4]. However, several studies have failed to validate NIRS measurement under normoxic conditions [48, 49]. A possible explanation for this discrepancy is that the NIRS signal contains information on arteriolar, capillary, venular, and intracellular Mb, while the O2 gradient from arteriole to venule is great during normoxic conditions, such that variations in blood volume from arteriole to venule could alter the NIRS signal without changing venous oxygen signals [40]. Lower oxygen levels during hypoxic conditions would reduce this effect. It is thus broadly accepted that the NIRS oxygenation/deoxygenation signal correlates strongly with changes in SvO2 and/or a-vO2D under varying oxygenation status in human muscle.

The use of a physiological calibration in combination with MRS measurements has allowed NIRS to measure muscle oxygen consumption (mVO2) [19]. The rate of decline of muscle oxygenation during ischemia can be compared with that of muscle PCr in mM/s or a conversion to mVO2 in mM/s. The transient arterial occlusion method has been used to measure forearm muscle metabolism during exercise [19, 42, 50]. Steady-state exercise at varying intensities was used to provide a range of mVO2, and resulting mVO2 values were compared to simultaneous MRS measurements of phosphorus metabolites. A significant correlation was found between mVO2 and PCr, and ADP concentrations. The linear relationships between exercise intensity and NIRS- and MRS-measured indicators support both thermodynamic [18] and kinetic [15] regulation models of mitochondrial respiration in skeletal muscle. The rate of deoxygenation using transient arterial occlusion has been compared to rate of PCr recovery, a biochemical process of ATP resynthesis via oxidative phosphorylation after exercise (Fig. 5.3). The rate of deoxygenation was quantitatively correlated with the rate of PCr recovery, proving the validity of transient arterial occlusion [50].

Fig. 5.3
figure 00053

Rate of hemoglobin/myoglobin deoxygenation post-exercise transit ischemia and phosphocreatine ( PCr ) resynthesis rate. Muscle oxygen consumption (VO2) using the transient arterial occlusion method was compared to rate of PCr recovery, a biochemical process of oxidative ATP resynthesis, at 30 s after exercise. The PCr kinetics is expressed as simulated data (Adapted with permission from J Appl Physiol 90(1):338–344. Copyright © 2001, American Physiological Society)

Full size image

Previous studies demonstrated that oxygenation of forearm flexor muscles closely reflects exercise intensity and metabolic rate determined by MRS during exercise and recovery [28, 51], and that muscle oxygenation level (% of arterial occlusion) shows a linear relationship with mVO2, though in a limited range (3.2 < mVO2 < 13.3-fold of resting), during exercise and recovery [52]. Therefore, where the transient arterial occlusion method is not applicable, muscle oxygenation level can serve to approximate mVO2.

5.6 Application of NIRS to Physiological Science

Early studies employed onset and recovery kinetics of oxygen saturation to evaluate oxygen use and oxygen delivery [12]. Muscle reoxygenation recovery time reflects the balance of oxygen delivery and oxygen demand in localized muscles. Recovery time measurements are based on extensive study of PCr recovery time, or on the PCr/Pi ratio. Chance et al. [12] compared male elite rowers to female elite rowers and made suggestions for improved performance. They reported a prolonged recovery time, which suggested increased energy deficit when exercise intensity increased. They also compared recovery time in submaximal and maximal work with plasma lactate, and demonstrated a significant correlation between blood lactate and muscle reoxygenation recovery time after exercise. Several studies have reported that muscle reoxygenaiton recovery time after submaximal to maximal exercise is one of the indicators for evaluating muscle oxidative capacity [12, 53]. The deoxy-Hb/Mb pattern during ramp cycling exercise has been monitored to distinguish trained cyclists from physically active subjects [54]. One set of authors proposed a method for noninvasively approximating muscle capillary blood flow kinetics using the kinetics of the primary component of pulmonary O2 uptake and deoxy-Hb/Mb in humans during exercise [55]. Other investigators compared rate of deoxygenation at onset of intermittent aerobic plantar flexion (7 on and 3 s off, repeatedly), where anaerobic glycolysis is negligible, to muscle oxidative enzyme activity [56], and demonstrated a good correlation between rate of deoxygenation and citrate synthase activity. Thus, we can assume that the rate of onset of deoxygenation reflects muscle oxidative capacity. One might consider that rate of mitochondrial respiration can be determined by rate of adenine nucleotide translocation and, therefore, that the [ATP]/[ADP] ratio regulates respiratory rate under physiological conditions [16]. However, ATP is maintained constant during submaximal aerobic exercise by creatine kinase equilibrium, so that we need not consider ADP adenine nucleotide translocation during aerobic exercise.

The deoxygenation patterns of the vastus lateralis (VL) and the lateral head of the gastrocnemius (GL) have been examined during graded treadmill exercise [57]. This study found a negative relationship between pulmonary VO2 and muscle oxygenation level, that the pattern of deoxygenation between VL and GL is somewhat different, and that muscle oxygenation level is associated with pulmonary VO2. Evaluation of muscle oxygenation in conjunction with systemic oxygen uptake would offer insight into the physiology of healthy and athletic individuals and provide a better exercise prescription for functional improvement. The influence of increased as well as decreased activity on muscle function has also been measured employing NIRS. Most studies have evaluated acute changes in muscle oxygenation during aerobic exercise, but several have also examined high-intensity exercise [58, 59]. In addition, NIRS has been employed to evaluate the training effects of exercise on muscle oxygenation and oxidative metabolism for various types of athletes (e.g., endurance [60, 61] and sprinters [62, 63]) in various sports. Costes et al. [64] attempted to determine whether exercise training–induced adaptations in muscle can be determined by NIRS and found that training does not change the pattern of muscle oxygenation, though a significant relationship was found between blood lactate and muscle oxygenation at the end of exercise. Motobe et al. [65] used NIRS in immobilized forearm muscle to measure changes in skeletal muscle oxidative function and to evaluate the preventive effect of the endurance training protocol on deterioration of skeletal muscle. They found that muscle oxidative function was determined by the time constant for recovery of mVO2 applying repeated transient arterial occlusions after exercise (see Fig. 5.4). NIRS measurement indicated delayed mVO2 recovery after exercise during immobilization.

Fig. 5.4
figure 00054

(a) Schematic representation of muscle oxygen consumption (VO 2mus mus) and typical changes in muscle oxygenated Hb/Mb. Schematic representation of VO2mus and typical changes in muscle oxygenated Hb/Mb at rest and during exercise and recovery. VO2mus was calculated from the rate of decline of oxygenated Hb/Mb during arterial occlusion (A.O.) at rest (Slope rest), and recovery (Slope recovery). (b) Typical kinetics of VO 2 mus recovery after exercise. VO2mus during recovery was compared with that at rest, normalized by the maximum value, and plotted as a function of time in seconds. Time constant of VO2mus for this subject was 56 s (Adapted with permission from Dyn Med 3 Galadriel 3:2. Copyright 2004, BioMed Central)

Full size image

In summary, NIRS can provide useful information on noninvasive monitoring of deconditioning and reconditioning of skeletal muscle oxidative function. However, most studies on the influence of training have been performed using a cross-sectional study design. There is still a need for more longitudinal studies on exercise training that employ NIRS measurements.

Several multichannel NIRS systems have been developed to detect regional differences in muscle oxygenation [66–68]. By simultaneously collecting data from multiple 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 study of regional differences in how skeletal muscle responds to exercise. Another rationale for using multichannel NIRS systems is that measuring at multiple locations provides better agreement between NIRS signal and oxygen saturation in the entire limb [47]. By simultaneously collecting data from multiple muscle regions, these devices avoid the variability caused by position-dependent differences in muscle oxygenation that plague all single-location measurements. Multichannel NIRS also holds an advantage over NMR and PET devices in terms of better time resolution.

5.7 Application of NIRS to Clinical Practice

The strong dependence of skeletal muscle on oxidative metabolism during exercise means that an improvement in the body’s oxidative system leads to higher performance during athletic events. On the other hand, impairment of VO2 and/or DO2 will limit performance and thus lead to functional deterioration. NIRS is suitable for measuring attenuation of VO2 and DO2 in patients with various diseases and possesses great advantages in terms of portability and real-time monitoring over conventional technologies.

5.7.1 Congestive Heart Failure (CHF)

NIRS has also been employed to evaluate skeletal muscle oxygenation in patients with heart disease. Patients with congestive heart failure (CHF) [45, 69–71] showed greater deoxygenation compared with controls partly due to heart pump failure and the consequent skeletal muscle hypoperfusion. The authors of [71] found a correlation between changes in total hemoglobin (tHb) and leg vessel conductance in patients with and without cardiac dysfunction during submaximal dynamic exercise. This result indicated that tHb reflects muscle vasodilation in these patients and that tHb can be used to assess vascular conductance in this patient group. Another study [72] demonstrated that changes in deoxygenation during submaximal exercise are steeper but peak deoxygenation lower in heart transplant recipients (HTRs) than in control subjects. The authors suggested that NIRS allows for detection of impairment of both DO2 and O2 extraction in HTR skeletal muscle. NIRS has also been used to examine the effect of pharmacological treatment on blunted microvascular oxygen delivery to muscle [73] and the effect of rehabilitation on muscle oxygenation [74] in patients with CHF.

5.7.2 Patients with Chronic Obstructive Pulmonary Disease (COPD)

Patients with COPD frequently develop skeletal muscle and vascular abnormalities as complications of their disease, similar to those with heart disease [75, 76]. Muscle reoxygenation recovery after exercise has been shown to correlate with pulmonary VO2 recovery kinetics in patients with COPD [77]. NIRS has also been employed to obtain the time constant of reoxygenation recovery during three work exercise tests, one below and two above the lactic acidosis threshold [78], where patients trained for 6 weeks on a cycle ergometer at high work rates. These investigators found significant correlations between changes in oxidative enzyme activity and changes in recovery time constant and endurance. Leg training is known to accelerate speed of reoxygenation of the vastus lateralis muscle after exercise, and this improvement has been correlated with changes in oxidative enzymes [79].

5.7.3 Neuromuscular Disorders

NIRS has also been employed to study patients with neuromuscular disorders. An increase in muscle oxygenation (paradoxical oxygenation) at onset of treadmill exercise was been reported in patients with cytochrome c oxidase deficiency [80], metabolic myopathy [81], and Friedreich’s ataxia [82]. This paradoxical oxygenation, a diagnostic indicator, is due to a combination of impaired mVO2 and a normal physiological increase of DO2 (vasodilatation) stimulated by muscle pump and/or myogenic activity. Grassi et al. [83] tested patients with mitochondrial myopathy (MM) or myophosphorylase deficiency (McArdle’s disease, McA) for changes in capacity for O2 extraction, maximal aerobic power, and exercise tolerance during cycle exercise using NIRS. They found that peak deoxygenation during exercise, an index of O2 extraction, was lower in MM and McA than in controls.

5.7.4 Peripheral Arterial Disease (PAD)

Several studies have used rate of recovery of reoxygenation to evaluate DO2 in the calf muscles of patients with peripheral artery disease (PAD) [84]. In a rather extensive study, the authors of [85] reported good agreement between faster PCr recovery kinetics and better oxygenation kinetics measured by NIRS [85]. A slower rate of calf reoxygenation after exercise is a consistent finding in PAD patients (see [86–88]).

Patients with varying severity of PAD have been classified by using patterns of calf oxygenation kinetics during treadmill exercise and recovery [87]. Impaired muscle O2 usage at rest [89] and at onset of exercise [90] have been observed as well. The authors of [84] found a good correlation between Doppler pressure waveform measurements and ankle arm systolic pressures (AAI) and the NIRS recovery time constant [84]. Mohler et al. [86] examined the interaction between PAD and presence or absence of diabetes mellitus (DM) employing changes in muscle capillary blood expansion and reoxygenation recovery, and found that capillary blood expansion is reduced in patients with DM regardless of existence of PAD. This parameter might thus be a good indicator of vascular impairment in patients with DM. Peripheral venous occlusive disease (PVOD) has also been tested using NIRS [91–93], and investigators were able to distinguish between successfully treated patients and those with deep vein thrombosis after 12 months. So we can conclude that NIRS is able to identify and quantify severity of DO2 in patients with PAD and PVOD.

5.7.5 Spinal Cord Injury (SCI)

NIRS has also been employed to evaluate changes in oxygenation that occur in the leg muscles of patients with spinal cord injury (SCI). Bhambhani et al. [94] found lower muscle deoxygenation during maximal exercise and faster changes in muscle deoxygenation with respect to pVO2 during functional electrical stimulation cycle exercise in SCI patients compared to healthy subjects. NIRS has been used to evaluate potential SCI therapies. The authors of [95] placed motor-complete SCI subjects and neurologically normal controls on a gait-training apparatus that enabled subjects to stand and move their legs passively, and found that oxygenation gradually increased and deoxygenation decreased in the experimental group, quite different from the response in normal controls. This result demonstrated that passive leg movement can not only induce muscular activity but can also alter muscle oxygenation in a paralyzed lower leg, and that induced muscular activity seems to correlate with increased muscle perfusion. In another NIRS study [96], SCI patients underwent electrical stimulation training (45 min daily, 3 days a week for 10 weeks) with different muscle oxygenation loads in paralyzed lower limbs. The vastus lateralis muscle of statically trained legs was found to show significantly increased oxidative capacity compared to both baseline and dynamically trained legs. This study demonstrated that NIRS is able to detect attenuated muscle deoxygenation after static training.

5.7.6 End-Stage Renal Disease (ESRD)

Muscle oxygenation and metabolism were examined using NIRS in children with end-stage renal disease (ESRD) before and after kidney transplantation and compared to controls during submaximal handgrip exercise [97]. These investigators looked at rate of deoxygenation, an indicator of mVO2, during transient arterial occlusion post-exercise and recovery time to reoxygenation after exercise. Both parameters were significantly improved after renal transplantation but not significantly different from controls (Fig. 5.5). Another group [98] employed NIRS to evaluate the potential for vascular and metabolic dysfunction in patients with renal failure, the effect of handgrip exercise training on forearm vasodilator responses, and forearm vasodilator responses to 3-min arterial occlusion in patients receiving hemodialysis, but found no improvement in vasodilator response after exercise training, and vasodilator response estimated by maximum oxygenation after release of arterial occlusion was significantly smaller in renal failure patients compared to controls. These studies show that NIRS can be used to detect muscle hypoperfusion in patients with renal failure and to find the functional alterations of muscle oxidative metabolism that occur after renal transplantation.

Fig. 5.5
figure 00055

Changes in recovery time (Rec-time) for Hb/Mb reoxygenation after renal transplantation (RT) in 10 renal transplant recipients. Individual data shown in black circles and solid lines; mean and SD shown in black squares and dotted line. Significant difference p < 0.05. Recovery time after transplantation was shortened in nine patients and delayed in one patient, who showed the shortest TR before RT (Adapted with permission from Am J Kidney Dis 48:473–480. Copyright © 2006, Elsevier)

Full size image

The results of clinical NIRS application suggest that it is a promising tool for noninvasively monitoring metabolic impairment in a follow-up setting, and in assessment of therapies and interventions.

5.8 Summary

There are several issues involved with in-vivo muscle NIRS measurement: (a) the origin of the NIR signal (from arterioles, capillary, and venules, as well as from Hb and Mb); (b) the NIR penetration depth or measurement area in tissue with varying source–detector arrangements in a multilayer model, including the effect of non-muscular tissue; (c) changes in optical properties within a wide range of tissue oxygenation status; (d) the wide variety of subjects; and (e) the choice of exercise modality. Nevertheless, there is little question that NIRS is suitable and quite useful for measuring changes in mVO2 and DO2 in healthy subjects as well as in patients with various disease states (e.g., chronic heart failure, chronic obstructive pulmonary disease, various muscle diseases, peripheral vascular disease, spinal cord injury, renal failure). In addition, NIRS has great advantages over conventional technologies in terms of portability and real-time monitoring.