INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

SKELETAL MUSCLE OXYGEN CONSUMPTION is a key parameter of metabolism under various conditions and has been a popular research subject for many years. In contrast to pulmonary O₂ uptake, which is measured noninvasively, muscle O₂ consumption can be measured invasively by measuring muscle blood flow and the arteriovenous O₂ difference. Since phosphorus magnetic resonance spectroscopy (³¹P-MRS) was first applied for the measurement of human skeletal muscle metabolism by Cresshull et al. in 1981 (9), it has become "a gold standard" of noninvasive measurement. ³¹P-MRS monitors the kinetics of intramuscular ATP, creatine phosphate (PCr), and inorganic phosphate (P_i) as well as pH. Many studies have examined the role of phosphate metabolites in muscle metabolism, especially in oxidative phosphorylation during exercise (2, 7, 24, 31) and recovery (1, 4, 6, 27, 31). Postexercise PCr resynthesis rate (Q) measured by ³¹P-MRS has been recognized as one of the most reliable parameters for quantifying the rate of oxidative ATP production (7, 19, 21, 25).

Near-infrared continuous wave spectroscopy (NIRcws) was first applied to the study of exercising skeletal muscle in humans by Chance et al. in 1985 (7). Over the past several years, many more groups have applied this technique (3, 10, 11, 15, 18, 23, 29). The parameters commonly measured by NIRcws are oxyhemoglobin/myoglobin (Hb/MbO₂), deoxyhemoglobin/myoglobin (Hb/MbR), and total hemoglobin/myoglobin (THb/Mb). NIRcws measures the relative changes of these parameters as the balance between O₂ supply and O₂ consumption. Therefore, it is necessary to know the amount of the O₂ supply to the muscle to estimate muscle O₂ consumption. The rate of Hb/MbO₂ decline under the condition of interrupting the O₂ supply to the muscle reflects O₂ consumption. Hamaoka et al. (14) examined muscle oxidative metabolism postexercise relative to the resting value (muscle O₂ consumption ratio) by occluding arterial blood flow with a pneumatic tourniquet during these two periods. Hamaoka et al. (12) also succeeded in quantitatively calculating muscle oxidative metabolic rate postexercise by multiplying the muscle O₂ consumption ratio by the muscle resting metabolic rate (RMRmus) measured using ³¹P-MRS (O_2NIR). It was assumed that O_2NIR reflected the absolute values of muscle O₂ consumption, and O_2NIR correlates with both intramuscular ADP and PCr concentrations, thought to be important regulators of mitochondrial oxidative metabolism. However, no studies have examined the validity of the quantitative values of O_2NIR. Therefore, the objective of this study was to examine the validity of NIRcws for quantitatively estimating O_2NIR in exercise. To validate the measurements made by NIRcws, its absolute values (the product of the muscle O₂ consumption ratio and RMRmus) were compared with the standard absolute values obtained by ³¹P-MRS.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects. Twelve male subjects (25 ± 5 yr) were tested in this study. All subjects were physically active, but none had participated in regular training programs requiring forearm exercise at the time of testing. The subjects were fully informed of the risks and gave their consent before the experiments.

Experimental protocol. Each subject sat on a chair with their forearm positioned horizontally and perpendicularly to the trunk (Fig. 1). The elbow was naturally extended, and the handgrip on the ergometer handle was adjusted so the subjects could grip the handle comfortably. The measurement site for both NIRcws and ³¹P-MRS was the upper part of the forearm over the finger flexor muscles. A new setup was used that allowed same-site measurement of the flexor muscles by both NIRcws and ³¹P-MRS under the same conditions (13). Each subject underwent three separate sessions, once for RMRmus measurement by ³¹P-MRS and two additional sessions for oxidative metabolism measurements postexercise. Each session was separated by at least 1 day. During the first session, RMRmus of the finger flexor muscles was measured using ³¹P-MRS (Fig. 2), which is explained in more detail below in Measurement of muscle metabolic rate. Two separate bouts of a rhythmic handgrip exercise were used to measure muscle oxidative metabolic rate postexercise, once for measuring O_2NIR by NIRcws and once for a standard measure by ³¹P-MRS (Fig. 2). Before each bout of exercise, the subject underwent arterial occlusion for 1 min followed by 2 min of rest. Each subject performed handgrip exercises at a frequency of one contraction every 2 s (0.5 Hz) for 3 min at one of the following intensities: 12% (n = 3), 18% (n = 3), or 24% (n = 6) of maximum voluntary isometric contraction. To determine the reproducibility of the exercise bouts between the different sessions while NIRcws was being measured, the kinetics of PCr, P_i, and intramuscular pH were simultaneously measured by ³¹P-MRS.

View larger version (31K): [in this window] [in a new window]   Fig. 1.   Experimental setup. Subject is seated with the exercising arm extended gripping the ergometer positioned inside a 31P-magnetic resonance spectroscopy (MRS) magnet. Both the 31P-MRS surface coil and the near-infrared continuous wave spectroscopy (NIRcws) light probes are positioned beneath the forearm. MVC, maximum voluntary isometric contraction.

View larger version (23K): [in this window] [in a new window]   Fig. 2.   Experimental protocols. During the first session, resting muscle metabolic rate (RMRmus) was determined using both 31P-MRS and NIRcws during 15 min of arterial occlusion (AO). Two separate exercise sessions were used to measure NIRcws and 31P-MRS. For both sessions, 3 min of exercise were preceded by 1 min of arterial occlusion and 2 min of rest. The subject performed handgrip exercise at 12, 18, or 24% of MVC at a rate of 0.5 Hz for 3 min. Measurements indicated by the arrows were taken at 30 s postexercise. Subjects underwent arterial occlusion during NIRcws measurement. Hatched bars, arterial occlusion induced by placing the cuff on the upper arm; solid bars, handgrip exercise (Ex); arrows, measurement of muscle oxidative metabolic rate. O2NIR(30), the quantitative value of muscle oxidative metabolic rate (O2NIR) at 30 s postexercise; Q(30), creatine phosphate (PCr) resynthesis rate at 30 s postexercise.

NIRcws and ³¹P-MRS measurements. NIR signals were obtained by an O₂ monitor (OM-100A, Shimadzu) with wavelengths of 780, 805, and 830 nm. Details of these instruments were previously reported by Tamura et al. in 1989 (30). Changes in Hb/MbO₂, Hb/MbR, and THb/Mb were calculated by the least-squares method using data from the changes in the absorbance of these lights with different wavelengths (18). To maintain an identical measurement site for both NIRcws and ³¹P-MRS, a 3-cm separation between the light source and the detector was used for NIRcws (Fig. 1). ³¹P-MRS signals were obtained by a NMR spectrometer (Otsuka Electronics) with a 2.0-T superconducting 26-cm-bore magnet. A double-tuned (¹H and ³¹P), 3-cm-diameter radio-frequency surface coil tuned to 34.58 Hz with 60-µs pulse width was used for phosphorus-signal acquisition. Pulse repetition time was 2 s. Five pulses were averaged to obtain a free induction decay, so each spectrum was obtained every 10 s. Three spectra were averaged during the preexercise period and during exercise. Each spectrum was used during the first 3 min postexercise, and then three spectra were averaged from 3 to 10 min postexercise. All ³¹P-MRS spectra were fitted to a Lorentzian line shape using the least-squares method. The relative area and frequency of the individual peaks were determined (Otsuka Electronics software) to calculate the areas of PCr, P_i, and -ATP peaks. The PCr and P_i intensities were normalized using the sum of PCr and P_i to avoid influence from possible changes in the sensitivity of ³¹P-MRS signals. Saturation correction was done by using saturation factors of PCr, P_i, and -ATP peaks, which were calculated by comparing the data from the 2-s and fully relaxed spectra. The saturation factors of PCr, P_i, and -ATP peaks in this study were 1.330, 1.081, and 1.184, respectively. Absolute PCr concentrations were calculated using PCr-to-ATP ratio and an ATP concentration of 8.2 mM, which was based on biopsy data (16).

Measurement of muscle metabolic rate. RMRmus was measured for all subjects during 15 min of arterial occlusion. With the use of the signal from ³¹P-MRS, RMRmus was defined as the rate of PCr decline under this ischemic condition (Fig. 2). With the assumption that the muscle metabolic rate does not change throughout the arterial occlusion during rest, the rate of the decline of Hb/MbO₂ level should also reflect the same metabolic rate as RMRmus (12). This premise was used during the exercise session when the NIRcws measurements were being taken. By induction of arterial occlusion both before and 30 s after exercise, the slope of the changes in the Hb/MbO₂ reflects the rate of decline of Hb/MbO₂, which corresponds to the relative change in the oxidative metabolic rate of the muscle during rest (S_R) and 30 s postexercise [S_PE(30)]. The ratio of S_PE(30) to S_R multiplied by RMRmus quantifies this measurement, providing absolute values for muscle oxidative metabolic rate 30 s postexercise O_2NIR(30). This is represented by the following equation (12)
With the use of ³¹P-MRS to reflect the rate of oxidative phosphorylation (28), the PCr resynthesis rate was also determined 30 s postexercise [Q₍₃₀₎] and was used as the standard measurement of the muscle oxidative metabolic rate. These measurements were under the same conditions as the NIRcws but were made without arterial occlusion postexercise. PCr monoexponentially recovers during the recovery period (22), and the PCr concentration can be described by the following equation
where PCr is the PCr concentration at time t during recovery, PCr₀ is the PCr concentration at the end of exercise, PCr is the change in the amount of PCr concentration during recovery, and k is the rate constant of the monoexponential curve. To obtain Q₍₃₀₎, the slope of a monoexponential curve at 30 s following exercise was calculated using the following equation
Sample calculations of both O_2NIR(30) obtained by NIRcws and Q₍₃₀₎ obtained by ³¹P-MRS are shown in Fig. 5. To compare the absolute values of O_2NIR(30) and Q₍₃₀₎, both values were expressed as millimolar ATP per second. The absolute values of muscle O₂ consumption (in mM O₂/s) were also calculated from the P-to-O₂ ratio, which is 6 for in vivo skeletal muscle (4, 12). The O_2NIR(30) measured by NIRcws should be equal to the standard value of Q₍₃₀₎ because the subjects performed identical exercises protocols for both measurements.

Data analysis. Data are presented as means ± SD. A linear regression analysis was used to examine the relationship between O_2NIR(30) and Q₍₃₀₎ as measured by NIRcws and ³¹P-MRS, respectively. P < 0.05 was defined as statistically significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Determination of RMRmus is shown in Fig. 3. Immediately after arterial occlusion was initiated, the Hb/MbO₂ level began to decrease linearly. The Hb/MbO₂ level ceased to decline after 5-6 min of arterial occlusion and remained unchanged throughout the rest of the measurement (Fig. 3). As soon as the Hb/MbO₂ level reached its lowest value, the PCr concentration began to decrease linearly throughout the remaining period of arterial occlusion. RMRmus was determined from the rate of decline of PCr, and the average value was 0.0076 ± 0.0008 mM ATP/s. Figure 4 shows the typical kinetics of Hb/MbO₂ and THb/Mb levels and PCr concentration measured by NIRcws and ³¹P-MRS, respectively, during rest, exercise, and recovery periods. For each Hb/MbO₂ and THb/Mb data point, the data are averaged over 1.5 and 30 s for PCr concentration. The Hb/MbO₂ level decreased rapidly under both arterial occlusion at rest (S_R) and at the onset of exercise. During exercise, once the Hb/MbO₂ level reached its lowest value, it remained at a constant level. The Hb/MbO₂ level rapidly increased immediately after exercise ceased but decreased even more steeply when arterial occlusion was applied postexercise [S_PE(30)] compared with S_R. During arterial occlusions at rest and postexercise, there were no significant changes in the THb/Mb levels as shown in Fig. 4, top trace. PCr concentrations also did not change at rest and during resting arterial occlusion (Fig. 4). During exercise, the PCr concentration decreased continuously and reached its lowest point at the end of exercise, after which it increased exponentially (Fig. 4). The average PCr concentration measured by ³¹P-MRS was 28.5 ± 4.3 mM at rest and decreased to 45.9 ± 17.4% (13.3 ± 5.8 mM) of the resting value at the end of exercise. Intramuscular pH was 7.02 ± 0.01 at rest and decreased to 6.70 ± 0.19 at the end of exercise. Intramuscular pH was even lower at 30 s after exercise (6.61 ± 0.30). Sample measurements of the rate of Hb/MbO₂ decline pre- and postexercise for estimating O_2NIR(30) and PCr recovery kinetics postexercise for estimating Q₍₃₀₎ from one subject are shown in Fig. 5. The RMRmus of this subject was 0.007 mM ATP/s. The calculated O_2NIR(30) and Q₍₃₀₎ were 0.060 and 0.081 mM ATP/s, respectively. Q₍₃₀₎ obtained by ³¹P-MRS ranged between 0.041 and 0.209 mM ATP/s. O_2NIR(30) obtained by NIRcws ranged between 0.018 and 0.187 mM ATP/s. These values were also equivalent to 3.0 and 31.2 µM O₂/s, respectively, which were calculated by using a P-to-O₂ ratio of 6. The relationship between muscle oxidative metabolic rate as measured by both NIRcws and MRS is shown in Fig. 6. There was a high correlation between O_2NIR(30) and Q₍₃₀₎ with r = 0.965 (P < 0.001).

View larger version (25K): [in this window] [in a new window]   Fig. 3.   Typical kinetics of oxyhemoglobin/myoglobin (Hb/MbO2) level and PCr concentration during resting metabolic rate measurements. Sample recordings are shown from 1 subject for RMRmus determination. Tracings of Hb/MbO2 and PCr measured by NIRcws and 31P-MRS, respectively, are shown during arterial occlusion. This ischemic condition caused an immediate decline in Hb/MbO2, which plateaued ~5 min after the start of occlusion. At this point, PCr began to decline linearly, which continued until the end of the 15 min of arterial occlusion. SR, slope at rest; OD, optical density.

View larger version (22K): [in this window] [in a new window]   Fig. 4.   Typical kinetics of Hb/MbO2 and total hemoglobin/myoglobin (THb/Mb) level and PCr concentration. Hatched bars, arterial occlusion (Occl); solid bars, handgrip exercise (Ex).

View larger version (26K): [in this window] [in a new window]   Fig. 5.   Calculations of oxidative metabolic rate. Top: O2NIR(30) calculation by NIRcws. The solid line and the dotted line are the linear regression lines of Hb/MbO2 levels during arterial occlusion at rest and at 30 s postexercise, respectively. Bottom: Q(30) calculation by 31P-MRS. PCr concentration increased in a monoexponential manner during recovery. The dotted line is the slope 30 s after exercise as calculated by differentiating the monoexponential curve. SPE(30), relative change in the oxidative metabolic rate 30 s postexercise; k, Rate constant of monoexponential curve; , change.

View larger version (14K): [in this window] [in a new window]   Fig. 6.   Relationship between muscle O2NIR(30) and Q(30) by NIRcws and 31P-MRS (n = 12). O2NIR(30) values are shown both in mM ATP/s (left) and in µM O2/s based on the calculation using a P-to-O2 ratio of 6 (right). The dotted line is a regression line of all data points. O2NIR(30) correlated significantly with Q(30) (P < 0.001).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

RMRmus measurements. Measuring RMRmus accurately was one of the most important parts of this study for estimating the absolute values of O_2NIR(30). Blei et al. (4) and Hamaoka et al. (12) reported that the resting PCr breakdown rate after the complete depletion of muscle O₂ stores by arterial occlusion at rest reflected RMRmus. In this study, as in previous studies, there were no changes in intramuscular pH and ATP concentration along with no O₂ utilization during RMRmus measurements. Because the contribution of the glycolysis to the PCr synthesis was negligible at rest (8), it is reasonable to conclude that the PCr breakdown rate after O₂ depletion reflects RMRmus (Fig. 3). The average value of RMRmus in this study was 0.0076 mM ATP/s, which is similar to both 0.008 and 0.0073 mM ATP/s previously measured by ³¹P-MRS (4, 12). This result is also comparable to the value of 6.3 µmol · min¹ · 100 ml¹ measured invasively (17).

Validity of NIRcws measurement. Several studies have shown that the Q values measured by ³¹P-MRS reflect the rate of oxidative ATP production (6, 16, 17, 21). The high correlation between Q₍₃₀₎ measured by ³¹P-MRS and O_2NIR(30) measured by NIRcws under the same conditions demonstrates the validity of NIRcws as a quantitative measurement of muscle O₂ consumption. Several studies have used NIRcws to estimate oxidative metabolism (10, 12, 18), and one study correlated NIRcws with a standard invasive measurement (18). To the best of our knowledge, this is the first study to show the validity of the quantitative values of O_2NIR measured by NIRcws. The muscle oxidative ATP production rates obtained in this study [Q₍₃₀₎] ranged between 0.041 and 0.209 mM ATP/s, which correspond to between 5.6- and 28.2-fold higher than the RMRmus. The initial Q for PCr ranged between 0.067 and 0.341 mM ATP/s. The intramuscular pH at the end of exercise ranged between 6.37 and 6.96. The wide range of these parameters would suggest the effectiveness of NIRcws measurements for evaluating muscle oxidative metabolic rates at various metabolic levels. To look into the wide range of the metabolic rates, we used different muscle loads. The possible reason of why different metabolic rates were obtained at the given intensity was the varying physiological and metabolic properties in the individuals. Blei et al. (4) reported that the average Q following fully excited muscle contraction was 0.28 mM ATP/s. Hartling et al. (17) reported that forearm O₂ uptake during maximal forearm dynamic exercise was 201 ± 56 µmol · min¹ · 100 ml¹, which is equivalent to 0.2 mM ATP/s (17). In this study, Q values from the three subjects were equivalent or higher than the values measured by Blei et al. (4) and Hartling et al. (17). This result indicates that the finger flexor muscles, which served as the measurement site in this study, were maximally activated in some cases and confirms the feasibility of the measurement of maximum muscle O₂ consumption by NIRcws.

Advantages and limitations in measurements. NIRcws has several advantages as a method of evaluating muscle oxidative metabolic rates compared with ³¹P-MRS measurements, which are currently held to be the gold standard. First, the portability of NIRcws allows it to be used anywhere it is needed. This characteristic allow for more opportunity to examine both clinical and nonclinical (experimental) aspects of energy metabolism. Second, NIRcws has a higher sensitivity to changes in muscle metabolism because NIRcws is a more direct measurement of the changes in O₂ content compared with the indirect measurement made by ³¹P-MRS. In other words, NIRcws can be applied for measuring muscle oxidative metabolism even in the case where there is a lack of a significant decrease in PCr such as very low-intensity exercise. Third, NIRcws has a higher time resolution than ³¹P-MRS in the measurements of muscle oxidative metabolic rates. The time resolution of NIRcws used in our study was 1.5 s; however, improvements in technology have increased the time resolution to 0.1 s. On the other hand, the time resolution for one data point by ³¹P-MRS in our study was 10 s, which was the same or even higher than the other studies (4, 25), although there were some studies in which ³¹P-MRS was able to be measured every 1 s (5). Furthermore, to measure muscle oxidative ATP production rate using ³¹P-MRS, the postexercise Q has to be determined. When the rate constant of PCr recovery is used to calculate Q, PCr kinetics postexercise has to be monitored for at least 5 min (4, 21, 25-27). Some studies achieved a much better time resolution in which the initial Q values were measured directly from the changes in PCr concentrations between successive time points (5, 20); however, it was pointed out that obvious random error occurred in the case of Q calculated in 10 s following exercise (5). Furthermore, it was suggested that a solution to this error would be to use a window of PCr resynthesis wider than 10 s during recovery because it would provide greater signal-to-noise ratio (5). In other words, the ideal condition is essential to calculate the oxidative metabolic rate within 10 s following exercise using ³¹P-MRS. In our setup, better reproducible results of the rate of PCr resynthesis were obtained when the rate constant of PCr recovery was used rather than when the changes in PCr concentrations between successive time points were used. On the other hand, as already shown in this paper, NIRcws takes 6 s and would take even shorter when NIRcws with a higher time resolution are used.