Journal of Applied Physiology
Vol. 83, No. 2, pp. 631-643, August 1997
EXERCISE AND MUSCLE

Mechanism of the exercise hyperkalemia: an alternate hypothesis

Karlman Wasserman, William W. Stringer, Richard Casaburi, and Yong-Yu Zhang

Division of Respiratory and Critical Care Physiology and Medicine, Harbor-UCLA Medical Center, Torrance, California 90509

ABSTRACT

INTRODUCTION

METHODS

RESULTS

DISCUSSION

FOOTNOTES

REFERENCES

ABSTRACT

Wasserman, Karlman, William W. Stringer, Richard Casaburi, and Yong-Yu Zhang. Mechanism of the exercise hyperkalemia: an alternate hypothesis. J. Appl. Physiol. 83(2): 631-643, 1997.A progressive hyperkalemia is observed as exercise intensity increases. The current most popular hypothesis for the hyperkalemia is that the Na⁺-K⁺ pump cannot keep pace with the K⁺ efflux from muscle during the depolarization-repolarization process of the sarcolemmal membrane during muscle contraction. In this report, we present data that suggest an alternate hypothesis to those previously described. Because phosphocreatine (PCr) is a highly dissociated acid and creatine is neutral at cell pH, the concentration of nondiffusible anions decreases, and an alkaline reaction takes place when PCr hydrolyzes. This creates a state of cation (K⁺) excess and H⁺ depletion in the cell. To examine the balance of K⁺ and H⁺ for exercising muscle during the early period of exercise when PCr changes most rapidly, catheters were inserted into the brachial artery and femoral vein (FV) in five healthy subjects who performed two 6-min cycle ergometer exercise tests at 40 and 85% of peak oxygen uptake. FV blood was sampled every 5 s during the first 2 min, then every 30 s for the remaining 4 min of exercise and the first 3 min of recovery, and then less frequently for the next 12 min. Arterial sampling was every 30 s during exercise and simultaneous with FV sampling during recovery. Arterial K⁺ concentration ([K⁺]) increase lagged FV [K⁺] increase. The hyperkalemia observed during early exercise results from K⁺ release from skeletal muscle. FV [K⁺] increased by 5 s of the start of exercise and followed the rate of H⁺ loss from the FV blood for the first 30 s of exercise. FV lactate and Na⁺ kinetics differed from K⁺ kinetics during exercise and recovery. As predicted from the PCr hydrolysis reaction, the exercising limb took up H⁺ and released K⁺ at the start of exercise (first 30 s) at both exercise intensities, resulting in a FV metabolic alkalosis. K⁺ release was essentially complete by 3 min, the time at which oxygen uptake (and, presumably, PCr) reached its asymptote. These findings lead us to hypothesize that the early K⁺ release by the cell takes place with H⁺ exchange and that the major mechanism for the exercise hyperkalemia is the reduction in nondiffusible intracellular anions in the myocyte as PCr hydrolyzes.

phosphocreatine; acid-base balance; hydrogen ion transport; lactate; potassium

INTRODUCTION

IT IS WELL KNOWN that plasma K⁺ concentration ([K⁺]) increases during exercise and that the major source for the plasma K⁺ increase is the exercising muscle (12, 17, 22, 23, 26, 31, 37). The increase in arterial plasma [K⁺] during exercise has been postulated to possibly have a functional role in the control of ventilation (5, 24, 27, 41) and the circulation (11, 32, 39). It may also play a detrimental role by eliciting cardiac arrhythmias (13, 18).

The changes in arterial and femoral vein (FV) [K⁺] in response to leg cycling exercise have been described by several groups. Linton et al. (23), using a continuously recording K⁺-sensitive electrode in four normal subjects, found arterial plasma [K⁺] to increase after a delay of ~15 s of the start of exercise. Immediately after 1 min of very heavy exercise, Medbo and Sejersted (26) found that plasma [K⁺] increases were similar in femoral artery and FV. They found no relationship between plasma [K⁺] and either pH or glycogen breakdown. Vollested et al. (37), using K⁺-sensitive electrodes in the FV and femoral artery, determined the FV  arterial (FV-a) K⁺ differences during submaximal and supramaximal exercise and recovery. They found that FV [K⁺] increased immediately at the start of exercise. Recovery was also quite rapid. Arterial [K⁺] changes lagged the FV changes, and the major release from muscle was during early exercise.

Hallen et al. (16) postulated that the mechanism of K⁺ release from muscle during exercise was due to lag in the activation of the Na⁺-K⁺ membrane pump in response to changes in sarcolemmal membrane potential during exercise (22). Other mechanisms include the effect of epinephrine on the Na⁺-K⁺ pump (15), metabolic acidosis (26), increased glycolysis (26), and muscle damage.

We studied the simultaneous rapid FV and arterial K⁺ concentration changes along with Na⁺ and acid-base changes during exercise and recovery. We conclude that the timing of the K⁺ release from exercising muscle and its reuptake in recovery, and the simultaneous acid-base changes, support the hypothesis that the primary mechanism for the exercise hyperkalemia is linked to a reduction in nondiffusible intracellular anions, which accompanies phosphocreatine (PCr) hydrolysis in response to exercise.

METHODS

Five healthy nonsmoking male subjects performed upright leg cycling exercise on an electromagnetically braked, computer-controlled, calibrated cycle ergometer (Godart, DeBilt, The Netherlands) at ~40 and 85% of peak oxygen uptake (O_{2 peak}) determined on the basis of a previously performed incremental exercise test. Each work rate was performed for 6 min. Table 1 describes the subjects and the work rates performed, the O_{2 peak}, and the lactic acidosis threshold measured by gas exchange (4). The protocol was approved by the Institutional Review Board. Each subject, after being briefed on the nature of the study and on what would be required of him, gave written consent to serve as subject.

Table  1.   Subject characteristics, aerobic parameters, and work rates at 40 and 85% of O2 peak Subject No. Age, yr Height, cm Weight, kg  O2 peak, l/min LAT, l/min WRmax, W WR at 40% O2 peak WR at 85% O2 peak 1 23 183 71 4.09 2.61 378 120 300 2 28 168 66 3.45 2.60 320 120 265 3 22 173 79 4.48 2.60 325 125 270 4 28 168 61 3.50 1.55 300 100 250 5 29 168 66 3.60 2.25 278 130 230 Mean ± SD 26 ± 3  172 ± 7  68.6 ± 6.8  3.82 ± 0.45  2.32 ± 0.46  320 ± 37  119 ± 11  263 ± 26 O2 peak, peak oxygen uptake (O2) at peak work rate (WR) during a progressively increasing maximal-effort WR test (WRmax); LAT, lactic acidosis threshold = the O2 at which CO2 production increased because of HCO3 buffering of lactic acid measured by V-slope method during a progressively increasing WR test (see Ref. 4).

Before exercise, a 10-cm 8-Fr catheter (Cordis, Miami, FL) was inserted percutaneously into the right FV (tip of the catheter ~4 cm above the inguinal ligament) and secured in place with a single suture. The catheter was kept patent with a 15 ml/h continuous infusion of heparinized normal saline. Another catheter was inserted percutaneously into the left brachial artery with the use of the Seldinger technique. FV blood was sampled every 5 s during the first 2 min of exercise by using a computer-controlled anaerobic collector (6). The latter consisted of a roller pump and 24 3-ml syringes on a manifold, which filled sequentially by a switching device under computer control. After 2 min of exercise, the syringes were filled by hand every 30 s during exercise and the first 3 min of recovery and then at 5, 10, and 15 min of recovery. Arterial blood was manually sampled every 30 s during the 6 min of exercise and the first 3 min of recovery and then at 5, 10, and 15 min of recovery. Blood samples from both sites were immediately iced and analyzed for blood lactate with a Yellow Springs Instrument Laboratory lactate/glucose analyzer (model 2300, Yellow Springs, OH). pH, PCO₂, and plasma [K⁺] and Na⁺ concentration ([Na⁺]) were measured with electrodes (model BGE, Instrumentation Laboratories, Lexington, MA) within 45 min of collection. Actual HCO₃ was calculated from the Henderson-Hasselbalch equation. Gas exchange was measured breath by breath and interpolated second by second as previously described (3). Arterial and femoral venous oxygen contents were calculated from the product of hemoglobin concentration, oxyhemoglobin percent saturation (Instrumentation Laboratories CO-oximeter), and 1.34 ml/g hemoglobin (oxygen capacity for hemoglobin when fully saturated).

K⁺, Na⁺, and lactate balances were calculated from the simultaneous FV and arterial concentration differences. The rate of K⁺ release from muscle during exercise was calculated from the following equation where _leg is the blood flow through both exercising extremities and 0.55 is the fraction of whole blood, which is plasma into which K⁺ is assumed to distribute (0.55 of whole blood) rather than whole blood water (average arterial hemoglobin concentration = 15.3 g/dl for the 5 subjects).

_leg was measured from arterial-FV oxygen content difference [C(a-v_FV)_O2] and an estimate of oxygen consumption (O₂) by the lower extremities, as follows where resting O₂ of the lower extremities is estimated to be 0.2 of the resting total body O₂, and the exercise O₂ of the lower extremities is estimated to be 0.9 [approximate slope of the leg O₂ vs. pulmonary O₂ during exercise, from Poole et al. (30)] of the difference between the measured exercise O₂ plus 0.8 of the resting O₂. The rate of lactate release was similarly calculated.

Values are reported as means ± SE, except when otherwise noted. Significant differences were concluded when P was <0.05.

RESULTS

Dynamics of FV [HCO₃] and [K⁺] changes at the start of exercise.

O₂, K⁺, and lactate dynamics during constant work rate moderate- (40% of O_{2 peak}) and heavy- (85% of O_{2 peak}) intensity exercise.

DISCUSSION

The experimental data presented in this paper demonstrate that the major source of the increase in blood [K⁺] is from the exercising muscle during the first 3 min for both moderate- and heavy-intensity exercise (Fig. 3). Vollested et al. (37), using K⁺-sensitive electrodes in the FV and artery during leg cycling exercise, found similar kinetics in K⁺ release from the exercising extremity in humans. They observed the overshoot in FV K⁺ during the first 3 min of moderate exercise, as found in our study, and K⁺ release across the exercising extremity to be 90% complete by 3.5 min at all levels of exercise.

The primary determinant of K⁺ concentration in the cell is the concentration of nondiffusible anions (Donnan membrane equilibrium). Because of the electrochemical potential resulting from the high intracellular [K⁺], K⁺ diffuses through leak channels in the plasma membrane of the cell into the extracellular fluid. Na⁺ diffuses in the opposite direction because of its electrochemical potential. The Na⁺-K⁺ pump [Na⁺-K⁺-adenosinetriphosphatase (ATPase)] maintains these electrochemical potentials and K⁺ as the major intracellular cation and Na⁺ as the major extracellular cation. Both the change in nondiffusible anions in the muscle cell and the inability of Na⁺-K⁺-ATPase to regulate Na⁺ and K⁺ across the sarcolemmal membrane might cause the muscle to release K⁺ at the start of exercise.

Because the Na⁺-K⁺ pump is responsible for maintaining high intracellular [K⁺] and low intracellular [Na⁺], the concept that sarcolemmal potential change during muscular contraction and the failure of the Na⁺-K⁺ pump to keep pace with the rate of K⁺ loss is generally accepted as the primary mechanism for the exercise hyperkalemia (22). Clausen and his colleagues (7) have done a great deal of research to expand our knowledge of the Na⁺-K⁺ pump in skeletal muscle. In in vitro studies of skeletal muscles of subhuman species (7), only 2-6% of the pump capacity is required to maintain the cellular and extracellular concentrations of Na⁺ and K⁺. Thus there appears to be a large reserve in ATPase (7, 8). Hormones such as insulin, thyroid, and catecholamines have been shown to modify the pump activity (7).

Ouabain attaches to the K⁺ receptor site on ATPase (the pump). Thus it blocks K⁺ uptake by the pump and the antiport movement of Na⁺, allowing the latter to leak into the cell. Clausen et al. (8) demonstrated that inhibition of Na⁺-K⁺-ATPase by ouabain accelerates cellular efflux of K⁺ and influx of Na⁺ in rat soleus muscle in vitro. During electrical stimulation, the Na⁺-K⁺ pump is activated above that level which could be accounted for by the increase in intracellular [Na⁺] (10). Thus, if stimulation was not relatively high, the increase in intracellular [Na⁺] was not significant. Semb and Sejersted (35) suggest that intracellular [Na⁺] may change minimally because it is functionally restricted to intracellular space close to the sarcolemmal membrane and, therefore, has a small transient intracellular volume of distribution. A comparable restricted space for K⁺ efflux on the outer membrane surface has not been suggested, although such a phenomenon might be necessary to minimize diffusion distances for the pump to allow repeated depolarization and repolarization of the sarcolemmal membrane during isotonic exercise.

Because of the importance of the Na⁺-K⁺ pump in maintaining Na⁺ and K⁺ balance across the sarcolemmal membrane, exercise hyperkalemia is generally attributed to the failure of the pump to keep pace with K⁺ loss from the muscle during contraction. Indeed, an increase in femoral venoarterial K⁺ difference has been reported during exercise in heart failure patients following digitalization (34). Hallen et al. (16) attributed the net loss of K⁺ from muscle to a lag in the Na⁺-K⁺ membrane pump during isotonic exercise (the "Na⁺-K⁺ pump lag" theory). They postulated that the Na⁺-K⁺ pump activity is catecholamine dependent and attributed the increased FV and arterial [K⁺], observed to take place after -adrenergic blockade (15, 16, 23), to an accentuation of the lag in Na⁺-K⁺ pump. However, the increased hyperkalemia with -adrenergic blockade during steady-state exercise might be accounted for by the failure to redistribute blood flow adequately to the exercising muscles. This should result in increased PCr hydrolysis, as suggested by the finding of increased lactate during exercise for work above the anaerobic threshold (36), and slower O₂ kinetics (increased oxygen deficit) in response to exercise after -adrenergic blockade, compared with before blockade (28).

A mechanism that lags implies a dynamic process that would eventually "catch-up" if the activity were sustained. Because skeletal muscle Na⁺-K⁺-ATPase is in excess concentration over that normally required (7, 8), it might be expected that K⁺ would be pumped back into the muscle cell as exercise is sustained. However, the increased plasma [K⁺] is maintained constant throughout exercise after 3 min. Only after exercise stops does plasma [K⁺] return to the preexercise level (Fig. 5).

Hemoglobin concentration was measured in each blood sample. FV and arterial concentrations were virtually identical, and they changed together during exercise. For the 40% O_{2 peak} exercise, hemoglobin concentration increased on average from 15.3 ± 0.3 g/dl at rest, to 15.6 ± 0.3 g/dl at 30 s to 15.6 ± 0.4 g/dl at 6 min of exercise. For the 85% of O_{2 peak} exercise, hemoglobin concentration increased from 15.4 ± 0.5 g/dl at rest, to 15.6 ± 0.5 g/dl at 30 s, to 16.9 ± 0.4 g/dl at 6 min of exercise. At the higher exercise intensity, the apparent 10% shrinkage in plasma volume attributable to the 10% increase in hemoglobin concentration could contribute in only a minor way to the ~67% increase in [K⁺] observed at 6 min of exercise and could contribute nothing to the increased arteriovenous [K⁺] difference measured across the exercising extremity and calculated K⁺ release during the first 3 min of exercise.

Not consistent with the Na⁺-K⁺ pump lag is the observation that arterial and FV plasma [K⁺] and [Na⁺] do not change reciprocally or synchronously during exercise. Arterial and FV [K⁺] increased during moderate exercise without a change in [Na⁺]. During heavy-intensity exercise, arterial and FV [Na⁺] and [K⁺] both increased, although not with the same kinetics (Figs. 3 and 8) and never changed in opposite direction, as would be predicted by the Na⁺-K⁺ pump lag hypothesis.

The pump lag hypothesis can also be examined in terms of how much K⁺ release from the muscle could be accounted for by sarcolemmal membrane depolarization and whether it would be reflected in the blood plasma. To alter the membrane potential by 100 mV, K⁺ loss from the myocyte during muscle contraction, estimated from the Nernst equation for a typical cell of 10 µm, would be only ~1/100,000 of the total K⁺ in the cytosol (1). Assuming that K⁺ release by the contracting myocytes is distributed in the extracellular fluid volume (20% of the body weight, or 13.7 liters; see Table 1 for body weight of subjects), the total K⁺ release for the 85% O_{2 peak} study (see Fig. 3) would be 38.4 mmol (13.7 × 2.8 mmol). Assuming that the contracting muscles are 20% of the body weight and that 40% is cell water (5.5 liters) with a K⁺ concentration of 140 mmol/l, the total K⁺ in contracting muscle cells before K⁺ loss would be 768.3 mmol. Thus, during the period of K⁺ release, ~5.0% of the cellular K⁺ moves extracellularly, an amount much larger than that predicted from K⁺ loss from sarcolemmal membrane potential change. As pointed out by Semb and Sejersted (35), the ions that cause the change in membrane potential may be restricted to the surface layer by small diffusion distances and, therefore, occupy relatively small volumes of distribution rather than the total extracellular fluid space.

Kjeldsen et al. (19) found that plasma K⁺ was reduced after exercise training, without a change in Na⁺-K⁺-ATPase activity. However, McKenna et al. (25) and Green et al. (14) found a 16 and 13.6% increase in Na⁺-K⁺-ATPase activity, with a 27 and 4.5% reduction in the plasma [K⁺], respectively, after training. Despite a large reserve in muscle ATPase (8), with an ability to upregulate, every level of exercise, even the mildest, causes plasma [K⁺] to increase.

Metabolic acidosis is also sometimes suggested as a possible mechanism for the exercise hyperkalemia. However, as is evident from Figs. 3 and 4, K⁺ is lost from cells early in exercise without a metabolic acidosis (actually a metabolic alkalosis). Also, the changes in lactate and K⁺ concentrations, and the dynamics of their release from muscle, are quite different. Moreover, the patterns of their concentration change differ during recovery. In recovery, K⁺ is rapidly taken up by the exercised extremity, whereas lactate, in response to heavy-intensity exercise (e.g., 85% of O_{2 peak}), continues to be released from muscle for at least 15 min. As lactate is removed from the arterial blood, presumably primarily by the liver, the arterial blood with reduced lactate concentration (Fig. 7) gradually washes out lactate that accumulated in the muscle during exercise.

Increased glycolysis accounting for K⁺ release from liver and muscle would seem an unlikely explanation, since the release of K⁺ from muscle is largely complete by 3 min, the time that the rate of glycolysis is greatest, as evident from the O₂ response and rate of lactate release from the exercising muscle. The major K⁺ release from exercising muscle is before 3 min and, for the 40% O_{2 peak} work rate, the FV [K⁺] decreases after reaching a maximum value, despite O₂ and, therefore, the rate of glycolysis increasing. Furthermore, the K⁺ release does not parallel the lactate release (Fig. 10), the latter being a measure of the rate of glycolysis. Finally, the measurements made by Linton et al. (23), using K⁺-sensitive electrodes in the hepatic vein during exercise, did not demonstrate increased K⁺ release by the liver, suggesting that the increased glycolysis in liver did not account for the exercise hyperkalemia.

It is generally recognized that heavy ischemic work may damage muscles cells and make them more permeable to myoglobin and other cell contents. The experimental findings of this study do not support the concept that the exercise hyperkalemia is due to muscle injury. Muscle injury should cause a late release and should occur primarily during heavy intensity exercise. Also, the rapid K⁺ reuptake in recovery (Fig. 5) would not have been expected if muscle injury accounted for the hyperkalemia.

In this paper, we describe experimental data suggesting that the exercise hyperkalemia is linked to PCr hydrolysis during early exercise. At the beginning of exercise, the major source of high-energy phosphate for muscle contraction is the hydrolysis of PCr. As it splits into creatine and inorganic phosphate, there is a rapid reduction in nondiffusible anions in the contracting muscle cells because PCr is a highly dissociated acid at the cell pH (acidic dissociation constants pK_a1 and pK_a2 = 2.7 and 4.6, respectively) (9), and when it hydrolyzes, it is converted into a neutral molecule (40), creatine, and dibasic phosphate (Fig. 11). This reaction should alkalinize the cell during the period of rapid PCr hydrolysis until the acidifying effects of increased CO₂ production and, during heavy exercise, lactic acidosis overwhelm the alkalinizing effect of PCr hydrolysis. CO₂ and lactic acid, being relatively permeable acids, should not have significant holding power for intracellular cations. By early spectral analysis, it was possible to observe the early transient alkalinization in the muscle of humans by ³¹P-nuclear magnetic resonance spectroscopy (2). The early alkalinization in the muscle effluent blood reported in this study (Fig. 1) coincides in timing and pattern with the early alkalinization of the muscle cell attributable to PCr hydrolysis (29).

The concentration of PCr is ~14 mmol/kg wet wt, depending on fiber type (33). The hydrolysis of PCr is a reaction that is virtually complete during the period of increasing O₂, i.e., within the first 3 min of exercise, and the decrease in PCr is sustained during the entire course of exercise to a level depending on the work intensity (20). Because of the reduction of nondiffusible anions in the form of PCr, fewer cations are needed in the cell to balance the cell negative charges (Donnan effect). Creatine, being undissociated or a zwitter ion (21), takes up H⁺ as illustrated in Fig. 11. The early alkalinization of the contracting myocytes reflects the abrupt reduction of nondiffusible intracellular anions. Findings that the kinetics of the K⁺ loss from the cells is similar to the rate of PCr hydrolysis (or rate of change in O₂, i.e., dO₂/dt; Fig. 10); that K⁺ increase in blood is sustained throughout exercise, regardless of its duration; and that the FV blood is alkalinized during the early period of exercise when the rate of K⁺ efflux from the cells is greatest, are consistent with the PCr hydrolysis hypothesis for the exercise hyperkalemia.

Transport of K⁺ from the muscle to the extracellular fluid must be accompanied by a new anion or the uptake of a cation by the cell to maintain ionic balance. During the early period of both moderate and heavy exercise, the increase in FV [K⁺] is associated, stoichiometrically, with new HCO₃ ion in FV plasma. This is reflected in an increase in pH without an increase in FV PCO₂. It is likely that the H⁺ formed with HCO₃ is taken up by the alkalinized muscle cell (Fig. 11). The fixation of metabolic CO₂ as HCO₃ during this early period of exercise probably accounts for the decrease in gas-exchange ratio routinely observed at the airway from 15 to 45 s after the start of exercise (38).

The rate of change in PCr should be of a similar order of magnitude with similar kinetics to the rate of change in O₂. We found that peak K⁺ release was ~5 and 10 mmol/min for the 40 and 85% of O_{2 peak}, respectively. Recognizing the limitations of discontinuous measurements, it can be appreciated from Fig. 10 that the peak rate of K⁺ release from the exercising muscle and dO₂/dt overlap closely in time and are of the same order of magnitude for the work intensities studied.

Limitations of this study that might have affected our conclusions include our inability to measure PCr in muscle. However, its change should track the early O₂ kinetics as previously shown (2). Furthermore, we did not measure blood flow directly but made what we believe to be reasonable estimates based on prior experimental work on the fraction of exercise O₂ attibutable to exercising muscle. We also assumed that K⁺ change was not reflected in the red blood cell, based on reported low permeability of cations to human red blood cells. These assumptions might have resulted in a small quantitative difference in the calculated K⁺ release from muscle but not in the pattern of release. A further limitation was that we measured arterial concentrations at 30-s intervals, whereby we measured FV concentrations every 5 s. Consequently, when calculating balance across the leg, the peak arterial-FV differences could be underestimated.

The PCr hydrolysis hypothesis for the exercise hyperkalemia links muscle K⁺ release to the early change in muscle bioenergetics after the start of exercise. The release of K⁺ from muscle might be anticipated consequent to hydrolysis of PCr because nondiffusible anions in the myocyte are reduced when PCr is hydrolyzed. The K⁺ release from muscle and simultaneous H⁺ loss from the extracellular fluid are very early events, taking place at 5-10 s and preceeding the increase in local PCO₂ and lactate increase (Fig. 4). This early appearance signifies that the transit delay through the interstitial fluid is relatively small compared with the 3-min period during which the K⁺ release takes place. However, it could be anticipated that the measure of the rate of release into the blood (Fig. 10) is a damped underestimate of the rate of release from the muscles. However, the inability to see what is happening at the cell membrane does not negate the interpretations on balance across the exercising extremity allowed by the measurement of venoarterial differences.

Whereas an analysis was given above describing 1) the inadequacies of some previously rendered hypotheses and 2) shortcomings of this study, several important findings in this study implicate the intracellular nondiffusible anion changes (accompanying PCr hydrolysis) as the primary mechanism of the exercise hyperkalemia. The first of these findings is that FV alkalemia develops at the start of exercise (Fig. 1) during the period when K⁺ and HCO₃ are simultaneously increasing (Fig. 4) and at a time that the myocyte pH is also increasing (2). Alkalemia, secondary to H⁺ loss from the extracellular fluid, very closely parallels the K⁺ gain in the FV blood. The chance that differences in speed of H⁺ and K⁺ diffusion artifactually caused similar changes in K⁺ and HCO₃ at the start of exercise is highly unlikely. The measurements of venoarterial difference indicate that the exercising extremity is the source of the K⁺ increase. Because both the cell and extracellular fluid become alkaline simultaneously during the first 30 s of exercise, the mechanism for the quasi-stoichiometric increase in HCO₃ and K⁺ must be due to consumption of cellular and extracellular H⁺ by the cell. Figure 11 shows how intracellular cation would become in excess when PCr hydrolyzes and the mechanism by which the cell consumes H⁺ when PCr splits. The transmembrane H⁺-for-K⁺ exchange generates extracellular HCO₃, which serves to balance the positively charged K⁺ added to the extracellular fluid during the first 30 s of exercise (Fig. 4). Later, lactate might serve as the principal anion balancing the increase in extracellular K⁺ (Fig. 6).

The second major finding linking muscle K⁺ release to PCr hydrolysis is that measurements of venoarterial K⁺ difference indicate that the period of K⁺ loss from the exercising extremity is virtually complete by 3 min of exercise. This coincides with the time that PCr hydrolysis is complete (2) and the time when O₂ reaches an assymptote (3 min). The period of K⁺ release from myocytes should take place only during the period of PCr hydrolysis, since this is the period associated with decreasing cellular concentration of nondiffusible intracellular anions. The K⁺ release is virtually complete by 3 min, as would be predicted if the release were linked to the kinetics of O₂ and PCr hydrolysis.

In summary, principal findings leading us to the alternative hypothesis for the mechanism of the exercise hyperkalemia are that 1) during early exercise, K⁺ release is accompanied by a FV alkalemia and increase in HCO₃ with similar kinetics to the increase in K⁺ concentration and 2) the venoarterial K⁺ differences approach zero by 3 min of exercise, with the 3-min values sustained for the duration of exercise. We conclude that the major mechanism for the exercise hyperkalemia is the reduction in intracellular nondiffusible anions accompanying PCr hydrolysis.

FOOTNOTES

Address for reprint requests: K. Wasserman, Division of Respiratory and Critical Care Physiology and Medicine, Harbor-UCLA Medical Center, Torrance, CA 90509.

Received 31 July 1996; accepted in final form 26 March 1997.

REFERENCES

1.	Alberts, B., D. Bray, J. Lewis, M. Raff, K. Roberts, and J. D. Watson. Molecular Biology of the Cell (2nd ed.). New York: Garland Publishing, 1989, p. 304-317.
2.	Barstow, T. J., S. Buchthal, S. Zanconato, and D. M. Cooper. Muscle energetics and pulmonary oxygen uptake kinetics during moderate exercise. J. Appl. Physiol. 74: 1742-1749, 1994[Abstract/Free Full Text].
3.	Beaver, W. L., N. Lamarra, and K. Wasserman. Breath-by-breath measurement of true alveolar gas exchange. J. Appl. Physiol. 51: 1662-1675, 1981[Abstract/Free Full Text].
4.	Beaver, W. L., K. Wasserman, and B. J. Whipp. A new method for detecting the anaerobic threshold by gas exchange. J. Appl. Physiol. 60: 2020-2027, 1986[Abstract/Free Full Text].
5.	Busse, M. W., J. Scholtz, F. Saxler, N. Maassen, and D. Boning. Relationship between plasma potassium and ventilation during successive periods of exercise in men. Eur. J. Appl. Physiol. 64: 22-25, 1992.
6.	Casaburi, R., J. Daly, J. E. Hansen, and R. M. Effros. Abrupt changes in mixed venous blood gas composition after the onset of exercise. J. Appl. Physiol. 67: 1106-1112, 1989[Abstract/Free Full Text].
7.	Clausen, T. Regulation of active Na+-K+ transport in skeletal muscle. Physiol. Rev. 66: 542-580, 1986[Free Full Text].
8.	Clausen, T., M. E. Everts, and K. Kjeldsen. Quantification of the maximum capacity for active sodium-potassium transport in rat skeletal muscle. J. Physiol. (Lond.) 388: 163-181, 1987[Abstract/Free Full Text].
9.	Dawson, R. M. C., D. Elliott, W. H. Elliott, and K. M. Jones (Editors). Data for Biochemical Research. London: Oxford Univ. Press, 1969, p. 106-107.
10.	Everts, M. E., and T. Clausen. Excitation-induced activation of the Na+-K+ pump in rat skeletal muscle. Am. J. Physiol. 266 (Cell Physiol. 35): C925-C934, 1994[Abstract/Free Full Text].
11.	Fallentin, N., B. R. Jensen, S. Bystrom, and G. Sjogarrd. Role of potassium in the reflex regulation of blood pressure during static exercise in man. J. Physiol. (Lond.) 451: 643-651, 1992[Abstract/Free Full Text].
12.	Fenn, W. O. Factors affecting the loss of potassium from stimulated muscles. Am. J. Physiol. 124: 213-227, 1938.
13.	Gettes, L. S. Electrolyte abnormalities underlying lethal and ventricular arrhythmias. Circulation 85: 170-176, 1992.
14.	Green, H. J., E. R. Chin, M. Ball-Burnett, and D. Raney. Increases in human skeletal muscle Na+-K+-ATPase concentration with short-term training. Am. J. Physiol. 264 (Cell Physiol. 33): C1538-C1541, 1993[Abstract/Free Full Text].
15.	Gullestad, L., J. Hallen, and O. M. Sejersted. K+ balance of the quadriceps muscle during dynamic exercise with and without -adrenoceptor blockade. J. Appl. Physiol. 78: 513-523, 1995[Abstract/Free Full Text].
16.	Hallen, J., L. Gullestad, and O. M. Sejersted. K+ shifts of skeletal muscle during stepwise bicycle exercise with and without beta-adrenoceptor blockade. J. Physiol. (Lond.) 477: 149-159, 1994[Medline].
17.	Juel, C., J. Bangsbo, T. Graham, and B. Saltin. Lactate and potassium fluxes from human skeletal muscle during and after intense, dynamic, knee extensor exercise. Acta Physiol. Scand. 140: 147-159, 1990[Medline].
18.	Kjeldsen, K. Muscle Na,K-pump dysfunction may expose the heart to dangerous K+ levels during exercise. Can. J. Appl. Sports Sci. 16: 33-39, 1991.
19.	Kjeldsen, K., A. Norgaard, and C. Hau. Exercise-induced hyperkalemia can be reduced in human subjects by moderate training without change in skeletal muscle Na+-K+-ATPase concentration. Eur. J. Clin. Invest. 20: 642-647, 1990[Medline].
20.	Kushmerick, M. J. Energetics of muscle contraction. In: Handbook of Physiology. Skeletal Muscle. Bethesda, MD: Am. Physiol. Soc., 1983, sect. 10, chapt. 8, p. 189-236.
21.	Lide, D. R. (Editor). CRC Handbook for Chemistry and Physics. Boca Raton, FL: CRC Press, 1994, p. 8-47.
22.	Lindinger, J. I., R. S. McKelvie, and J. F. Heigenhauser. K+ and Lac distribution in humans during and after high-intensity exercise: role in muscle fatigue attenuation? J. Appl. Physiol. 78: 765-777, 1995[Abstract/Free Full Text].
23.	Linton, R. A. F., M. Lim, C. B. Wolff, P. Wilmshurst, and D. M. Band. Arterial plasma potassium measured continuously during exercise in man. Clin. Sci. (Lond.) 67: 427-431, 1984[Medline].
24.	McCoy, M., and M. Hargreaves. Potassium and ventilation during incremental exercise in trained and untrained men. J. Appl. Physiol. 73: 1287-1290, 1992[Abstract/Free Full Text].
25.	McKenna, M. J., T. A. Schmidt, M. Hargreaves, L. Cameron, S. L. Skinner, and K. Kjeldsen. Sprint training increases human skeletal muscle Na+-K+-ATPase concentration and improves K+ regulation. J. Appl. Physiol. 75: 173-180, 1993[Abstract/Free Full Text].
26.	Medbo, J. I., and O. M. Sejersted. Plasma potassium changes with high intensity exercise. J. Physiol. (Lond.) 421: 105-122, 1990[Abstract/Free Full Text].
27.	Paterson, D. J. Potassium and ventilation in exercise. J. Appl. Physiol. 72: 811-820, 1992[Abstract/Free Full Text].
28.	Petersen, E. S., B. J. Whipp, J. A. Davis, D. J. Huntsman, H. V. Brown, and K. Wasserman. Effects of -adrenergic blockade on ventilation and gas exchange during exercise in humans. J. Appl. Physiol. 54: 1306-1313, 1983[Abstract/Free Full Text].
29.	Piiper, J. Production of lactic acid in heavy exercise and acid-base balance. In: Lactate: Physiologic, Methodologic and Pathologic Approach, edited by P. R. Moret, J. Weber, J. Haissly, and H. Denolin. New York: Springer-Verlag, 1980, p. 35-45.
30.	Poole, D. C., G. A. Gaesser, M. C. Hogan, D. R. Knight, and P. D. Wagner. Pulmonary and leg O2 during submaximal exercise: implications for muscular efficiency. J. Appl. Physiol. 72: 805-810, 1992[Abstract/Free Full Text].
31.	Qayyum, M. S., C. A. Freemantle, C. J. Carey, B. C. Page, N. Soper, D. J. Paterson, and P. A. Robbins. Potassium loss from skeletal muscle during exercise in man: a radioisotope study. Exp. Biol. 78: 639-648, 1993.
32.	Sadamoto, T., Y. Mutoh, and M. Miyashita. Cardiovascular reflexes during sustained handgrip exercise: role of muscle fibre composition potassium and lactate. Eur. J. Appl. Physiol. 65: 324-330, 1992.
33.	Saltin, B., and P. D. Gollnick. Skeletal muscle adaptability: significance for metabolism and performance. In: Handbook of Physiology. Skeletal Muscle. Bethesda, MD: Am. Physiol. Soc., 1983, sect. 10, chapt. 19, p. 555.
34.	Schmidt, T. A., H. Bundgaard, H. L. Olesen, N. H. Secher, and K. Kjeldsen. Digoxin affects potassium homeostasis during exercise in patients with heart failure. Cardiovasc. Res. 29: 506-511, 1995[Medline].
35.	Semb, S. O., and O. M. Sejersted. Fuzzy space and control of Na+, K+- pump rate in heart and skeletal muscle. Acta Physiol. Scand. 156: 213-224, 1995.
36.	Twentyman, O. P., A. Disley, H. R. Gribbin, K. G. M. M. Alberti, and A. E. Tattersfield. Effect of -adrenergic blockade on respiratory and metabolic responses to exercise. J. Appl. Physiol. 51: 788-793, 1981[Abstract/Free Full Text].
37.	Vollested, N. K., J. Hallen, and O. M. Sejersted. Effect of exercise intensity on potassium balance in muscle and blood of man. J. Physiol. (Lond.) 475: 359-368, 1994[Abstract/Free Full Text].
38.	Wasserman, K. Coupling of external to internal respiration. Am. Rev. Respir. Dis. 129: S21-S24, 1984[Medline].
39.	Wilson, J. R., S. C. Kapoor, and C. G. Krishna. Contribution of potassium to exercise-induced vasodilation in humans. J. Appl. Physiol. 77: 2552-2557, 1994[Abstract/Free Full Text].
40.	Windholz, M. The Merck Index: An Encyclopedia of Chemicals, Drugs, and Biologicals. Rahway, NJ: Merck & Co., 1983, p. 368.
41.	Yoshida, T., T. Chida, M. Ichioka, K. Makiguchi, N. Tojo, and N. Udo. Ventilatory response and arterial potassium concentration during incremental exercise in patients with chronic airways obstruction. Clin. Physiol. 11: 73-82, 1991[Medline].