INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES APPENDIX

RECENTLY, AN INCREASE in concentrations of HCO₃ and K⁺ and a decline in H⁺ concentration (increase in pH) without an increase in PCO₂ was found in femoral vein blood during early exercise (27) (Fig. 1) and could be attributed to the hydrolysis of phosphocreatine (PCr), a reaction that takes up H⁺ (20). The increase in HCO₃ and K⁺ were stoichiometrically linked (27). This early exercise-induced alkalosis (E-I Alk) must be a component of the transient increase in CO₂ stores. The fixing of metabolic CO₂ as HCO₃ (22.3 ml/mmol of HCO₃), as a result of this early E-I Alk, contributes to a reduction in CO₂ production (CO₂) relative to O₂ uptake (O₂) and to a decrease in respiratory exchange ratio (RER) during the period of high rate of HCO₃ formation (Fig. 1). The decrease in RER takes place between 15 and 90 s after the start of exercise, with the lowest value at ~30 s. Although this phenomenon is consistent at all work rates (WRs), it is more marked at high WRs. However, with heavy exercise, buffering of the developing lactic acidosis partially masks the increase in early CO₂ stores (23, 27).

View larger version (25K): [in this window] [in a new window]   Fig. 1.   Respiratory exchange ratio (RER) and femoral vein (FV) HCO3, PCO2, and PO2 in response to 40% (left) and 85% (right) of peak O2 uptake (O2) during the first 90 s of leg cycling exercise in an upright position. Data shown are averaged for 5 subjects. Data reported in this figure come from studies previously reported from this laboratory (25, 27). The femoral vein HCO3 and PCO2 data were reported in Ref 27. PO2 data come from the same studies and were reported in Ref 25. RER data are from the same studies but were not previously reported. Bars show SE at selected times [at rest and at 30 s (dotted lines) and at 60 s of exercise]. Significant differences: * P < 0.05 from rest; + P < 0.05 between values at 30 and at 60 s of exercise.

To generate high-energy phosphate, muscle metabolizes primarily glycogen (7, 9, 13, 21). Thus muscle substrate respiratory quotient (RQ_m) is relatively high. In fact, muscle biopsy as well as gas-exchange measurements have shown that the RQ_m is equal to ~0.95 (4, 7, 9). Therefore, the decrease in RER at 15-90 s after the onset of exercise (Fig. 1) must be due to an increase in CO₂ stores. For moderate [40% of peak O₂ or ~80% of lactic acidosis threshold (LAT)] constant-work-rate (CWR) exercise (Fig. 1, left), RER reaches a peak value of <1.0 before 4 min and thereafter remains relatively constant or decreases slowly with time (17, 28). If the WR is above the subject's LAT (Fig. 1, right), the initial decrease in RER is followed by an increase to a value >1.0 because of release of CO₂ from HCO₃ as HCO₃ buffers the newly formed lactic acid. It then decreases toward a lower value as the rate of lactate increase slows (28).

Four mechanisms can cause tissue CO₂ stores to increase: 1) blood CO₂ that results from increase in PCO₂, 2) CO₂ bound to hemoglobin (Hb) as carbamate due to oxyhemoglobin (HbO₂) desaturation (Haldane effect) (8), 3) increase in CO₂ stored in tissues as PCO₂ increases, and 4) fixation of metabolic CO₂ as HCO₃ due to the alkalinization resulting from the conversion of PCr to creatine and inorganic phosphate (2, 20, 27, 31). The purpose of this study was to determine the dynamics and volume of the increase in CO₂ stored during moderate exercise and to apportion the quantities attributable to each mechanism. Because exercise above the LAT will result in additional CO₂ as HCO₃ buffers lactic acid, and hyperventilation resulting from ventilatory compensation for metabolic acidosis (32), we restricted our studies to WRs that were below the LAT.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES APPENDIX

Subjects

Preliminary Ramp Exercise Testing

The slope (S) of the increase of O₂ as a function of increase in WR was derived for each individual by least squares regression line of O₂ vs. WR. The mean value and SD were 10.1 ± 0.7 ml · min¹ · W¹ and were not significantly different from the value of 10.2 ± 1.0 ml · min¹ · W¹ previously reported for normal subjects (15). The WRs at O₂ of 60 and 80% LAT were calculated as follows where b is the O₂ for unloaded cycling.

CWR Test

View larger version (14K): [in this window] [in a new window]   Fig. 2.   Schematic representation of testing protocol. Vertical bars indicate time of loaded exercise (load). Each subject performed 2 protocols on different days, in random order, one at 60% of the lactic acidosis threshold (LAT) and the other at 80% LAT. Subjects maintained a cycling rate of 60 rpm on the unloaded ergometer (marked as 1, 2, 3, 4) before the workload was imposed. Workload was repeated 4 times for 4-min workload periods, with 5-min recovery periods of unloaded cycling (unl) between each workload. Unloaded O2 averaged 372 ± 9 ml/min for the 60% LAT protocol and 379 ± 6 ml/min for the 80% LAT protocol.

Measurements and Calculations

Pulmonary gas exchange. The O₂, CO₂, and other respiratory variables were measured breath by breath (Medical Graphics). These data were converted by interpolation to points at intervals of 1 s. The points at the beginning of each WR increase were time aligned and then averaged, second by second, for the four WR repetitions. Because all work repetitions were not exactly equal to the second in time (they may differ by several seconds, despite computer control, because of variable breath lengths), the averaging was restricted to the length of the shortest work repetition. Similarly, all transitions to unloaded cycling were aligned at the end of the exercise load and were averaged over the length of the shortest unloaded repetition. The transition points were correctly aligned so that the transition behavior was averaged accurately. The resulting averaged exercise and recovery repetitions were joined to give a true picture of O₂ and CO₂ dynamics over the entire cycle of loaded cycling (exercise) followed by unloaded cycling (recovery).

RQ_m. RQ_m was calculated as the ratio of the total CO₂ produced during the 9 min of the exercise-recovery period (CO_{2 work} + CO_2rec) minus the baseline volume of CO₂ (CO_{2 baseline}) to the total O₂ consumed during the 9 min of the exercise-recovery period (O_{2 work} + O_{2 rec}) minus the baseline volume of O₂ consumed (O_{2 baseline}) for each subject, i.e.

CO₂ stores from pulmonary gas-exchange measurements. The increase in aerobically generated CO₂ per unit of time (CO_{2 work}) in response to exercise is described by the increase in O₂ per unit of time above baseline (O_{2 work}) times RQ_m. The O_{2 work} is equal to the difference between pulmonary O₂ during unloaded and loaded exercise (O_{2 work}) plus the decrease in venous O₂ stores (O_{2 venous}) and muscle myoglobin (Mb) O₂ stores (O_{2 Mb}). O_{2 Mb} is relatively small, being ~24 ml for 80% LAT WR (see DISCUSSION). Adding this to the O₂ measured by external respiration would result in an additional 24 ml (less for the 60% LAT WR) to the E-I Alk CO₂ stores. Because it is a small and fixed value, we have not added this volume to the volume of the increase in CO₂ stored, as determined from external respiration measurements. In turn

(1)

CO_{2 work} is also equal to the sum of the increase in pulmonary CO₂ output (CO_{2 work}) in response to loaded exercise plus CO₂ from aerobic metabolism stored in the tissues (CO_{2 store}) plus venous blood (CO_{2 venous}), i.e.

(2)

Substituting CO_{2 work} of Eq. 1 for CO_{2 work} of Eq. 2

(3)

Assuming RQ_m × O_{2 venous} = CO_{2 venous}, these terms would cancel each other; hence

(4)

Thus, by not including O_{2 venous} in the total O_{2 work}, we do not include CO_{2 venous} in the increase in CO₂ stores calculated from external respiration. This leaves three components of the increase in CO₂ stores that are determined from measurements of external respiration. These include 1) the increase in blood CO₂ content due to Hb deoxygenation (Haldane effect); 2) the increase in CO₂ stored in tissues because of increase in tissue PCO₂; and 3) the chemically bound CO₂ due to tissue alkalinization from hydrolysis of PCr.

(5)

Calculation of components of the increase in CO₂ stores. haldane effect. We utilized near-infrared spectroscopy (NIRS) (RunMan; NIM, Philadelphia, PA) to monitor kinetics of muscle HbO₂ and MbO₂ saturation. Changes in differences between the two wavelengths monitored [(760-850 nm)] are used to estimate changes in the relative desaturation of the (HbO₂ + MbO₂) in the tissue under the probe (18). The linearity of the desaturation output from the NIRS unit had been validated previously in our laboratory (6). The probe was positioned over the vastus lateralis muscle, 10-12 cm above the knee and parallel to the major axis of the thigh. All studies were performed after calibration with similar balance of the signal and difference-gain settings on the spectrometer. The muscle oxygenation signal was continuously monitored and recorded second by second throughout the protocol on a computerized system (DI200 PGH/PGC; Dataq Instruments, Akron, OH).

(6)

(7)

1

1

(8)

(9)

Femoral vein blood-gas changes. In constructing Fig. 1 to illustrate the relationship between the changes in RER and the gas-exchange events at the level of the muscle, we used previously reported data from studies by this laboratory (25, 27). The femoral venous HCO₃ and Pfv_CO2 were reported in Ref. 27. To compare the magnitude and timing of the metabolic alkalosis (see Fig. 7) with the E-I Alk CO₂ stores, we calculated standard HCO₃ from the data reported in Ref. 27. The PO₂ was reported in Ref. 25. The gas-exchange data for calculating RER are previously unreported but come from the same experiments.

O₂ time constant. The increase in O₂ from unloaded cycling to steady-state exercise was fit with a monoexponential model from the following equation by using least squares criteria (2, 30)

(10)

where Y₀ is the baseline (unloaded cycling) O₂, A₁ is the O₂ above baseline at steady-state (O_{2 ss}), and  is the time constant for O₂.

O₂ deficit. The O₂ deficit is simply the product of O_{2 ss} and  (30), i.e.

(11)

Data Analysis and Statistics

Comparison of multiple unloaded O₂ and CO₂ values was carried out by using a repeated-measures ANOVA. Comparison of variables between the two levels of exercise was done by using the paired t-test. Significance of differences among the three components in the CumCO₂ stores calculation was determined by using ANOVA. Relationships among variables were studied by using linear regression techniques and by calculating Pearson product-moment correlation coefficients. Significant change was determined to be P < 0.05. All values reported are means ± SD.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES APPENDIX

Gas Exchange During Exercise and Recovery

Dynamics of RER and Exercise-Induced Changes in CO₂ Stores

View larger version (23K): [in this window] [in a new window]   Fig. 3.   Example of kinetics of increase in CO2 stores and its accumulation (CumCO2 store) derived from the simultaneous changes in O2 and CO2 during a 4-min 80% LAT (60 W) constant work rate exercise and 5 min of recovery. Data are for subject 5. All data are calculated second by second (see text for details). dCO2 store/dt, rate of CO2 store increase in ml/min; CumCO2 store, accumulated CO2 stores as related to time; TD, time delay; T90%, time to 90% of response.

The dynamics of CO₂ store change in response to exercise are shown in Table 2 for all of the subjects. The average RQ_m values were 0.91 ± 0.04 and 0.97 ± 0.06 for the 60 and 80% LAT WRs, respectively. The increase in CO₂ stores became evident at the lung, with a time delay of 16 ± 6 and 12 ± 6 s from the start of exercise for the 60 and 80% LAT WRs, respectively. It then increased to a peak rate of 103 ± 27 and 184 ± 74 ml/min at 61 ± 8 and 55 ± 6 s, respectively, for the 60 and 80% WRs. The peak rate of increase in CO₂ stores was significantly larger (P < 0.01), and the time delay was significantly shorter (P < 0.05) for the 80% compared with the 60% LAT exercise (Table 2). However, the peak time for dCO₂ store/dt was not significantly different for the two WRs after subtracting the time delays for each exercise. The CumCO₂ store of 244 ± 93 ml for 80% LAT was significantly higher than the 141 ± 48 ml for the 60% LAT WR (P < 0.001; Table 2). The 90% time (T₉₀) for CumCO₂ stores for 80% LAT was the same as that for 60% LAT. During unloaded recovery, RER, CO₂ stores, and Cum CO₂ stores behaved with similar but reverse dynamics to the accumulating CO₂ stores.

Components of the Increase in CO₂ Stores

The Haldane effect. The Haldane effect should have similar kinetics to that of muscle deoxygenation, because this effect takes place in venous blood. Figure 4 shows the average dynamics of the leg deoxygenation immediately after the 80% LAT exercise load was imposed for the 12 subjects. It reached a constant value by 50 s (Fig. 4). The group half-time (T_50%) values for the leg deoxygenation were 21 ± 5 and 19 ± 4 s for the 60 and 80% LAT exercise, respectively (P > 0.05). The group T_90% for the leg deoxygenation was 37 ± 11 and 39 ± 12 s for the 60 and 80% LAT exercise, respectively (P > 0.05). Compared with the time course of increase in CO₂ stores (which had a T_90% of 133 ± 14 and 133 ± 18 s for the 60 and 80%-LAT exercise, respectively), the kinetics of the leg deoxygenation was much faster. The increase in CO₂ stores from the Haldane effect (Eq. 6) was 6 ± 1 and 13 ± 2 ml for 60 and 80% LAT WRs, respectively.

View larger version (17K): [in this window] [in a new window]   Fig. 4.   Leg deoxygenation measured with near-infrared spectroscopy (NIRS) and the simultaneous RER for 4 min of 80% LAT constant-work-intensity exercise. Data are average for 12 subjects in the study. Dotted line (left), start of loaded exercise from unloaded cycling; dotted line (right), time transition from loaded to unloaded cycling. Half-time (T50%) and 90% of response time (T90%) of the NIRS signal are indicated. Nadir of RER response was at 50 s and then gradually increased to a steady state by 3 min. A.U., arbitrary unit.

Physically dissolved CO₂. The increase in physically dissolved CO₂ (Eq. 7) was 50 ± 25 and 60 ± 21 ml for 60 and 80% LAT WRs, respectively.

E-I Alk CO₂. The difference between the total and the sum of physically dissolved and Haldane effect is 84 ± 27 ml (61% of total) and 171 ± 88 ml (68% of total) for 60 and 80% LAT WRs, respectively (Table 3). This is attributable to the E-I Alk CO₂ store (fixed as HCO₃). It is the largest fraction of the various components of the increase in total CO₂ stores measured by external respiration. The absolute, but not the percentage, values were significantly higher for the 80% LAT compared with the 60% LAT exercise.

The Accumulated CO₂ Stores and Fitness

View larger version (20K): [in this window] [in a new window]   Fig. 5.   Changes () in total CO2 stores and exercise-induced metabolic alkali (E-I Alk) CO2 stores (in ml) as a function of O2 deficit (in ml) for 60% (n = 9) and 80% LAT (n = 12) constant work rate exercise. In each panel, each symbol represents a different subject. Solid lines are the least squares best fit correlation between total CO2 stores or E-I Alk CO2 stores and O2 deficit.

Also, the time to the peak rate of increase in CO₂ stores was significantly correlated with the time constants for O₂ for the 60 and 80% LAT exercise (r = 0.74 and 0.79, P = 0.02 and < 0.01, respectively; Fig. 6). This correlation remained significant after subtracting the tissue-to-lung time delay (Table 2) from the time of peak rate increase for both levels of exercise (Fig. 6, bottom).

View larger version (19K): [in this window] [in a new window]   Fig. 6.   Time to peak change in CO2 stores (dCO2 store/dt) (top) and time to peak minus time delay (TD) (bottom) as a function of time constants () for O2 for 60% LAT (n = 9) and 80% LAT (n = 12) constant work rate exercise (in s). In each panel, each symbol represents a different subject. Regression lines for data are also shown.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES APPENDIX

Dynamics of RER and CO₂ Stores During CWR Exercise

O₂ and CO₂ abruptly increase at the start of exercise with the same RER as for unloaded cycling (resting RER if starting exercise from rest; e.g., Fig 1). RER then consistently decreases for all work intensities, starting at ~15 s, reaches a nadir between 30 and 60 s, and approaches a constant value by 180 s (Fig. 1, left; Figs. 3 and 4) (17, 26, 28, 30). The 15-s delay approximates the time required for blood to be transported from the muscle capillaries to the pulmonary capillaries. During the time of decreasing RER, some of the CO₂ generated from aerobic metabolism is stored rather than being delivered to the lungs in proportion to O₂.

It should be pointed out that the decrease and then increase in RER is not caused by a switch from carbohydrate to fatty acid metabolism in muscle followed by a switch back to carbohydrate metabolism. If that were the case, CO₂ would not be available in the stores for RER to increase in the early recovery period (no hyperventilation at these WRs below the LAT) and CO₂ would not increase above O₂ in recovery as shown in Figs. 3 and 4. In fact, the CO₂-balance studies during exercise and recovery shown here support a constant RQ_m during exercise. That the recharging of PCr stores at early recovery is the major cause for the postexercise rise in RER is supported by the systematic release of the stored CO₂ (Fig. 3) and reuptake of K⁺ by the muscle (see Fig. 5 in Ref. 27) during recovery, at the time that PCr is resynthesized (see Figs. 2 and 4 in Ref 31).

Components of the Early CO₂ Stores and Calculation Assumptions

The Haldane effect. O₂ desaturation allows Hb to bind additional CO₂ without increasing blood PCO₂. Thus, despite no increase in PCO₂ in the femoral vein blood during the first 30 s of exercise (Fig. 1), the decrease in PO₂ and, therefore, O₂ saturation can bind additional CO₂ as carbamate and account for some of stores. However, the kinetics of HbO₂ desaturation (T₅₀ ~19-21 s; T₉₀ ~37-39 s) measured by NIRS (Fig. 4) were clearly faster than the kinetics of the increase in CO₂ stores (Fig. 3, Table 2). Femoral vein O₂ desaturation for moderate-intensity exercise was shown to be complete within 30-50 s after its start (25). This finding is consistent with the kinetics of muscle deoxygenation reported here (Fig. 4). By using data reported previously (24), it can be calculated that the increase in volume of CO₂ stored as carbamate due to Hb deoxygenation (Haldane effect) at the same PCO₂ is relatively small, being only ~5% of the total increase in tissue CO₂ stores. The Haldane effect also causes changes in HCO₃ (16) when accompanied by a change in PCO₂. Change in the HCO₃ component in venous blood due to uptake or release of CO₂ when Hb is deoxygenated during exercise and oxygenated during recovery is offset by changes in venous O₂ content, as reflected in the reasonable assumption that CO_{2 venous} = RQ_m × O_{2 venous}.

Physically dissolved CO₂. The change in physically dissolved CO₂ depends on 1) the volume of the tissue exposed to the change in PCO₂, 2) the solubility of CO₂ in the tissue, and 3) the increase in tissue PCO₂ (10). By using equations previously reported (12), we estimated the leg muscles weighed 18 ± 3 kg. We assumed that the volume of both legs does not change during 4 min of moderate exercise. The muscle tissue fluid volume of both legs would be ~11 liters, assuming that the fluid component is 60% of the leg muscle weight. The solubility of CO₂ is changed with the change in body temperature. During light and moderate exercise, the body temperature does not rise >2°C, for which the solubility of CO₂ would fall from 0.0301 to 0.0294 mmol · l¹ · mmHg¹ (19, 22). The increase in Pfv_CO2 approximated from the increase in PET_CO2 above unloaded cycling for each subject at each WR (see APPENDIX) averaged 7 ± 3 and 8 ± 2 Torr for the 60 and 80% LAT WRs, respectively. Thus, given these approximations, the increase in the physically dissolved CO₂ stores averaged 50 and 60 ml for the 60 and 80% LAT, respectively (Table 3).

The E-I Alk production. As soon as exercise starts, PCr hydrolyzes and exerts an alkalinizing effect as shown by studies that employed ³¹P-nuclear magnetic resonance spectroscopy (1, 2, 31). This alkalinizing effect causes an increase in femoral vein pH and HCO₃ without an increase in PCO₂ during the first 30 s of leg cycling exercise (27). As shown in Fig. 7, standard HCO₃ increases in the femoral vein but not in the arterial blood during this early period of exercise. Wasserman et al. (27) pointed out that the major mechanism for the exercise hyperkalemia is the reduction in nondiffusible intracellular anions in the myocytes when PCr hydrolyzes, because creatine is essentially neutral, whereas PCr is an acid (pK = 3.87) at the pH of cell. Thus new anions are needed to balance the free K⁺ made available when highly dissociated PCr is converted to neutral creatine. The new anion available to the myocyte is HCO₃, because the H⁺ of carbonic acid would be consumed in the hydrolysis of PCr (20). In this way, the splitting of PCr slows the rate of muscle CO₂ output from aerobic metabolism into the exhaled gas (increases the CO₂ stores). Thus venous-arterial standard HCO₃ difference increases transiently during the time of a high rate of PCr hydrolysis (Fig. 7). For a WR similar to 80% LAT exercise, the average increase in femoral vein standard HCO₃ during the first 2 min of exercise is ~0.8 mmol/l (Fig. 7) (25, 27). Because the leg blood flow would be on the order of 8 l/min in steady-state (from an assumed leg arteriovenous O₂ content difference of 100 ml/l and leg O₂ of 800 ml/min, calculated from total body O₂), increasing from 3 l/min at unloaded cycling (14), a WR of 80% LAT would result in an E-I Alk component of the increase in CO₂ stores of ~178 ml (2 min × 0.8 mmol/l × 22.3 ml/mmol × 5 l/min). Thus the average value that we obtained for the E-I Alk (171 ml for the 80% LAT WR shown in Table 3) is of the expected order of magnitude.

View larger version (21K): [in this window] [in a new window]   Fig. 7.   Arterial (Art; circles) and femoral vein (Fem V; squares) standard HCO3 (top; solid symbols) and actual HCO3 (bottom; open symbols) in response to 80% LAT upright leg cycling exercise. Data are average for 5 subjects from materials of Ref. 27. Bars, SE. Significant differences are provided at rest and at 30, 60, 90, and 180 s of exercise: * P < 0.05 compared with rest; + P < 0.05 compared with 30 s; # P < 0.05 compared with 60 s.

Potential miscellaneous factors. For the reasons described in METHODS, the increase in venous blood CO₂, which is a true increase in CO₂ stores, does not enter into our estimate of the increase in total CO₂ stores calculated from external respiration because it is offset by the consumption of O₂ from the venous O₂ stores that we did not measure. Similarly, changes in lung CO₂ stores from changing functional residual capacity have not been included, because this should be very small (<10 ml) and offset by reciprocal changes in lung O₂ stores.

Possible estimation errors. The O₂ consumption and CO₂ production of skeletal muscles that support respiration and posture are included in the measured O₂ and CO₂ during exercise. These muscles contribute <6% of the increase in whole body O₂ and CO₂ during moderate CWR leg muscle exercise (14). The magnitude of error was estimated for the Haldane effect and the physically dissolved CO₂ stores by applying previously reported measurements. For sensitivity analysis involving one SE of the mean, the Haldane effect is estimated to be 3 ± 1 to 6 ± 2% for the 60% LAT WR exercise and 5 ± 2 to 7 ± 3% for 80% LAT WR exercise. The physically dissolved CO₂ store is 30 ± 6 to 39 ± 7% for 60% LAT and 23 ± 8 to 30 ± 10% for the 80% LAT WR exercise. The EI-Alk CO₂ is 55 ± 7 to 66 ± 6% for the 60% LAT WR exercise and 64 ± 13 to 72 ± 10% for the 80% LAT WR exercise.

CO₂ Stores and Muscle Bioenergetics

O₂ Deficit and CO₂ Stores

Conclusions