Supinski, G. S., D. Stofan, R. Ciufo, and A. DiMarco. N-acetylcysteine administration alters the response to inspiratory loading in oxygen-supplemented rats.J. Appl. Physiol. 82(4): 1119–1125, 1997.—Based on recent studies, it has been suggested that free radicals are elaborated in the respiratory muscles during strenuous contractions and contribute to the development of muscle fatigue. If this theory is correct, then it should be possible to attenuate the development of diaphragm fatigue and/or delay the onset of respiratory failure during loaded breathing by administering a free radical scavenger. The purpose of the present experiment was, therefore, to examine the effect ofN-acetylcysteine (NAC), a free radical scavenger and glutathione precursor, on the evolution of respiratory failure in decerebrate unanesthetized rats breathing against a large inspiratory resistive load. We compared the inspiratory volume and pressure generation over time in animals pretreated with either saline or NAC (150 mg/kg) and then loaded until respiratory arrest. After arrest, the diaphragm was excised, and samples were assayed for reduced (GSH) and oxidized glutathione. As a control, we also assessed respiratory function and glutathione concentrations in groups of nonloaded saline- and NAC-treated animals. We found that NAC-treated animals were able to tolerate loading better than the saline-treated group, maintaining higher inspiratory pressures and sustaining higher inspired volumes. Administration of NAC also increased the time that animals could tolerate loading before the development of respiratory arrest. In addition, although saline-treated loaded animals had significant reductions in diaphragmatic GSH levels compared with unloaded controls, the magnitude of this reduction was blunted by NAC administration (i.e., GSH averaged 965 ± 113, 568 ± 83, 907 ± 39, and 784 ± 61 nmol/g for unloaded-saline, loaded-saline, unloaded-NAC, and loaded-NAC groups, P< 0.05, with the value for the loaded-saline group lower than the values for the two unloaded groups; GSH for the loaded-NAC group was not different, however, from unloaded controls). These data demonstrate that administration of NAC, a free radical scavenger, slows the rate of development of respiratory failure during inspiratory resistive loading.
- free radicals
- skeletal muscle
recent studies suggest that oxygen-derived free radicals are produced in the respiratory muscles in response to strenuous contraction. In support of this concept, several biochemical markers of free radical generation have been found to increase in the respiratory muscles after electrically induced contractions with the use of in vitro and in situ experimental preparations (17, 18, 24) and after inspiratory loading of intact animals (2, 4, 8). For example, a number of experiments have detected alterations in diaphragmatic glutathione metabolism after loaded breathing [i.e., a reduction in reduced glutathione (GSH) and an increase in oxidized glutathione (GSSG)] that are consistent with the elaboration of free radicals (2, 4). In addition, several studies have indicated that it is possible to attenuate the rate of development of diaphragmatic fatigue during electrically induced contractions by administration of free radical scavengers, which, presumably, prevent free radical-mediated muscle dysfunction (19, 22).
Because respiratory muscle fatigue is thought to contribute to the development of respiratory failure during loaded breathing (1), and because scavengers reduce fatigue, it would seem possible to alter the course of development of respiratory failure during loading by administering a free radical scavenger. We recently tested this theory by administering N-acetylcysteine (NAC) to decerebrate rats just before applying a large inspiratory resistive load (23). We found that NAC administration largely prevented loading-induced depletion of diaphragmatic GSH stores. Despite this effect, NAC failed to delay the development of respiratory failure. In this past work, however, experiments were performed with animals breathing room air, with the result that profound hypoxemia developed during loading. This severe hypoxemia may have altered central nervous system and muscular physiology, accelerating the development of respiratory failure and obscuring any protective effect provided by NAC.
The purpose of the present study was to readdress these issues. We postulated that NAC administration would alter the course of development of respiratory failure during loaded breathing, providing that sufficient inspiratory oxygen supplementation was administered to blunt loading-induced hypoxemia. Studies were performed on decerebrate rats given either saline or NAC and then loaded with a large inspiratory resistance while receiving supplemental oxygen. As controls, we also studied unloaded saline- and NAC-treated animals. At the conclusion of load trials, animals were killed, and samples of diaphragm were obtained for assessment of in vitro isometric force generation and for determination of GSH and GSSG concentrations.
Animal preparation. Experiments were performed on 32 adult Charles River rats weighing between 400 and 600 g. Animals were housed in the Case Western Reserve Animal Facilities and fed rat chow ad libitum. On the day of study, animals were anesthetized with inhalational halothane anesthesia administered via a face mask; this was supplemented with a single injection of intramuscular ketamine (30 mg/kg). A neck incision was then made, and blunt dissection was used to expose the trachea. A small incision was made in the anterior wall of the trachea, a 7-mm-long polyethylene tube was inserted, and a silk suture was tightened around the trachea to secure this tube. Animals were then placed on mechanical ventilation, and halothane administration was continued via the ventilator. Femoral venous and arterial lines were placed via a cutdown, and the arterial line was attached to a blood pressure monitor. Halothane administration was then adjusted to reduce arterial pressure to ∼60 Torr.
Animals were subsequently placed in a prone position, and an incision was made in the skin over the cranium. A small burr hole was drilled in the skull, and a bone rongeur was used to extend this craniotomy and expose the cerebral cortex. A midcollicular section was performed by using a flat probe, and the cerebrum rostral to this section was aspirated. The exposed brain was then lightly packed with gelfoam and cotton to suppress bleeding. Halothane administration was stopped, and animals were returned to the supine position. As soon as vigorous spontaneous breathing efforts had returned, mechanical ventilation was discontinued. A small breathing circuit, consisting of a Hans Rudolf valve with a pneumotachograph on the expiratory limb, was then attached to the endotracheal tube. The tachygraph signal was amplified and integrated (Charles Ward Enterprises PI-830 integrator) to obtain tidal volume measurements. A side arm in the Hans Rudolf valve was connected to a Validyne pressure transducer (MP-45) and used to monitor airway pressure. An small-bore polyethylene catheter attached to a Validyne pressure transducer was then advanced into the esophagus and used to monitor esophageal pressure swings during breathing. Bipolar stainless steel electrodes were placed percutaneously into the diaphragm; electrical activity recorded by using these electrodes was amplified (WPI amplifiers) and integrated by using a Paynter filter (Charles Ward Enterprises). Airway pressure, esophageal pressure, systemic arterial pressure, and tidal volume were recorded by using a Gould RS3600 physiological chart recorder. Body temperature was monitored by using a rectal thermistor probe and maintained at 37°C with a heating lamp.
Animals were next assigned to one of four experimental groups:1) unloaded saline-treated animals (n = 8);2) unloaded NAC-treated animals (n = 7);3) loaded saline-treated animals (n = 9); and4) loaded NAC-treated animals (n = 8). Drugs were administered intravenously over a 5-min period (0.75 ml/kg of saline or 0.75 ml/kg of a 200 mg/ml solution of NAC). After drug administration, a 1-h period was provided to allow agents to distribute to interstitial spaces. Five minutes before this hour elapsed, supplemental oxygen (100%) was added to the inspiratory limb of the breathing circuit. This usage was based on previous studies demonstrating the development of profound hypoxemia almost immediately after the institution of inspiratory loading in this experimental model when animals breathed room air.
At the end of this equilibration period, a large inspiratory load, consisting of a long piece of polyethylene tubing (internal diameter 0.86 mm, resistance 20,000 cmH2O ⋅ l−1 ⋅ s) was attached to the inspiratory limb of the breathing circuit. Loading was subsequently continued until respiratory arrest occurred, with arrest defined as the cessation of breathing efforts for 10 s. Samples of arterial blood were drawn immediately before loading, at 5 min into load trials, and at the point of respiratory arrest. Blood-gas analysis was performed on these samples by using a Radiometer model ABL30 blood analyzer. Immediately after arrest, the abdomen was opened and the diaphragm was removed en bloc. A portion of this muscle was placed in a dissecting dish containing gassed Krebs-Henselheit solution (95% O2-5% CO2, pH 7.40, 135 mM NaCl, 5 mM KCl, 11.1 mM dextrose, 2.5 mM CaCl2, 1 mM MgSO4, 14.9 mM NaHCO3, 1 mM NaHPO4, 50 U/l insulin, and 16 mg/l d-tubocurare). The remaining muscle was immediately frozen in liquid nitrogen, stored at −70°C, and later assayed to determine GSH and GSSG concentrations.
Unloaded saline and NAC-treated control groups of animals were monitored for 60 min after completion of the equilibration period; at the end of this time, diaphragms from unloaded animals were removed en bloc and processed in the same way as were diaphragms from loaded animals. During this 60-min observation period, 100% oxygen was administered to the breathing circuit of unloaded animals.
In vitro assessment of diaphragm force generation. Diaphragm muscle strips were prepared from the portion of the excised diaphragm placed in the dissecting dish. Strips were then mounted vertically in a Radnoti water-jacketed organ bath containing gassed Krebs-Henselheit solution and maintained at a temperature of 27°C. One end of each strip was attached to a fixture in the bottom of the organ bath, whereas the other end was connected via a metal rod to an isometric Grass force transducer (model number FT10). Muscles were electrically stimulated by using a platinum mesh field electrode array, with current delivered to these electrodes from a constant-current amplifier (Applied Neural Control Laboratories) driven by a Grass S48 stimulator. Current intensity was increased to supramaximal levels, and muscle strip length was adjusted to the length at which force generation was maximal.
Diaphragm force-generating ability was subsequently assessed. To ensure that these measurements were done at a uniform time with respect to load trials, in vitro measurements were started 20 min after arrest in all experiments. Diaphragm strips were first stimulated with single electrical impulses to determine twitch contraction time (time-to-peak force development) and half-relaxation time (time for peak force to fall by 50%). After a 5-s rest, the diaphragm force-frequency relationship was determined by stimulating muscles with trains of 1-, 10-, 20-, 50-, and 100-Hz impulses. Each train was applied for 800 ms, and adjacent trains were separated by 5-s periods. An additional 30-s rest was provided after the force-frequency curve was completed. In vitro fatigability was then assessed by stimulating strips to contract 30 times/min for 5 min in response to 500-ms-long trains of 20-Hz impulses. Muscle strips were then removed from the organ bath, extraneous connective tissue was excised, and each strip was weighed.
Measurement of GSH and GSSG concentrations. Muscle GSH and GSSG concentrations were assessed by using the modified high-performance liquid chromatography (HPLC) method described by Reed et al. (16). In brief, this method is based on the reaction of iodoacetic acid with protein sulfhydryl groups to form S-carboxymethyl derivatives, followed by reaction with 1-fluoro-2,4-dinitrobenzene to form chromophores. To carry out his assay, we first pulverized ∼100 mg of muscle under a blanket of liquid nitrogen. This powder was mixed with ice-cold perchloric acid containing bathophenanthrolined sulfonic acid (to prevent further oxidation), homogenized, frozen in liquid nitrogen, thawed, and centrifuged. The supernatant was mixed with iodoacetic acid and an alkaline buffer and then incubated. 1-Fluoro-2,4-dinitrobenzene was then added, and this latter reaction allowed to proceed. Aliquots (100 μl) were then injected onto a Spherisorb-Amino column, individual sulfur-containing amino acids were separated by reverse-phase ion-exchange HPLC (Varian 5000 HPLC), and eluted amino acids were measured by using ultraviolet detection (365 nm). Known concentrations of GSH and GSSG were prepared as standards, and peak areas of GSH and GSSG in tissue samples were quantitated by comparison to areas obtained by using the standards.
Data analysis. Determination of inspiratory duration during loaded trials was accomplished by measuring the time from the beginning of inspiration to the most negative deflection of the esophageal pressure waveform. Breathing cycle duration was taken as the time between adjacent peaks of this waveform.
In vitro diaphragm strip force was expressed per unit muscle cross-sectional area, as calculated by using the following formula (5)
Statistical analysis.Comparison of single variables across experimental groups (e.g., twitch forces, GSH) was made by using a one-way analysis of variance, with post hoc testing used to detect statistical differences between individual groups. A repeated-measures multivariate analysis of variance was employed for those comparisons in which repeated measurements of a given variable were made over time or under different conditions (e.g., in vitro force-frequency curves). Because of significant animal mortality after the 50-min point in loaded breathing trials, it was not possible to make meaningful comparisons of several variables [i.e., pressure, volume, electromyographic (EMG) activity] after this point in time. For this reason, statistical comparison of the rate of fall of pressures and volumes between saline and NAC-treated animals at the end of loaded breathing trials was restricted to the time between 20 and 50 min of loading (i.e., from the point at which pressures began to fall to the 50-min time point). Animal survival (i.e., time to respiratory arrest) was compared between saline- and NAC-treated loaded groups of animals by using the Wilcoxan survival test (SPSS statistical sofware).
Data are presented as means ± SE. AP < 0.05 was taken as indicating statistical significance.
Breathing trials. In nonloaded saline- and NAC-treated animals, there were no significant changes over time in diaphragmatic EMG activity, tidal volume, respiratory timing, arterial pressure, or in arterial blood- gas tensions. For example, tidal volume averaged 2.06 ± 0.23 and 2.00 ± 0.29 ml, respectively, at the beginning and end of experimental trials in unloaded saline-treated animals and was 2.24 ± 0.32 and 2.46 ± 0.35 ml at the beginning and end of trials in unloaded NAC-treated animals. Arterial pressure and blood-gas data for unloaded animals are provided in Tables 1 and2.
For loaded trials, application of the inspiratory resistance was followed by an increase in EMG activity and a rise in inspiratory pressure generation. The pressure gradient across the inspiratory load increased to a peak level of 47 ± 7 cmH2O in saline-treated loaded animals and a level of 49 ± 6 cmH2O in NAC-treated loaded animals. Pressure generation rose and/or remained at high levels over the first 20 min of loading in all animals, with no appreciable difference between saline- and NAC-treated animals with respect to this initial response.
After 20 min of loading, inspiratory pressure generation gradually fell in both saline- and NAC-treated animals (see Fig.1). The rate of this decline was greater in the saline-treated group (P < 0.037 for comparison of pressure-time curves between 20 and 50 min of loading in saline- and NAC-treated animals). EMG activity was relatively stable over this time period (Fig. 2), and there was no statistically significant difference between EMG time curves in saline- and NAC-treated animals between 20 and 50 min of loading. Because of the high animal mortality after 50 min of loading, meaningful EMG time curves comparing loaded saline- and loaded NAC-treated groups could not be displayed as a function of time from the initiation of loaded breathing trials after the 50-min point. To permit an examination of EMG activity at the very end of loaded breathing trials, we also calculated average peak EMG activity as a function of time to the end of loaded trials; this information is presented in Table 3. As is evident, there was a rapid fall in peak EMG activity over the final 5 min of loading in both NAC and saline-treated loaded groups; the trajectory of this terminal decline in EMG activity was not different for the two groups.
Tidal volumes were also similar in saline and NAC-treated animals over the first 20 min of loaded breathing trials. From 20 min onward, there was a progressive reduction in tidal volume in both groups of animals (Fig. 3) but, as with pressure generation, volume decreased more slowly over time in the NAC-treated group (P < 0.05 for comparison of volume-time curves of saline- and NAC-treated animals between 20 and 50 min of loading).
Loaded breathing was also associated with progressive alterations in respiratory timing in both loaded groups, as shown in Table4. Breathing cycle duration became progressively longer during loading, and there was a small increase in inspiratory duration. Alterations in respiratory timing were similar during loaded trials for saline- and NAC-treated groups.
Arterial CO2 increased and arterial oxygen tensions decreased with loading, but oxygen tensions remained >55 Torr in all but one loaded animal (Table 2). Arterial blood pressure remained relatively constant during the early portion of loaded breathing trials but fell over the final few minutes of trials (Table 1). There was no difference between saline- and NAC-treated animals with respect to arterial pressure measurements.
On average, NAC-treated animals took longer to reach the point of respiratory arrest than saline-treated animals (Fig.4). Time to arrest averaged 62 ± 4 min in the NAC group and 49 ± 4 min in the saline-treated group (P < 0.05).
In vitro force generation. Diaphragms excised from both loaded groups had reduced force generation in vitro compared with diaphragms from unloaded saline- and NAC-treated animals (Fig. 5). For example, twitch forces for saline- and NAC-treated loaded animals were 4.4 ± 0.6 and 4.2 ± 1.3 N/cm2, respectively, whereas twitch force averaged 8.7 ± 0.4 and 8.5 ± 0.5 N/cm2, respectively, in unloaded saline- and NAC-treated animals (P < 0.0001 for comparison of loaded to unloaded twitch forces, with no significant difference between the two loaded groups or between the two unloaded groups).
In vitro twitch kinetics were not different among the four groups, however, with contraction time averaging 71 ± 3 ms and half-relaxation time averaging 61 ± 3 ms. Moreover, the in vitro fatigability of strips excised from the four groups of animals was not significantly different. Specifically, force fell to 0.22 ± 0.02, 0.20 ± 0.01, 0.25 ± 0.02, and 0.26 ± 0.05 N/cm2 of its initial value during in vitro fatigue trials in unloaded saline-treated, unloaded NAC-treated, loaded saline-treated, and loaded NAC-treated groups, respectively.
Glutathione concentrations. Loading elicited a reduction in diaphragmatic GSH levels in saline-treated animals (Fig. 6), as judged by comparison to both control groups (P < 0.05 for this comparison). Moreover, the magnitude of the reduction in GSH elicited by loading was attenuated in NAC-treated animals, with GSH levels in this group (784 ± 61 nmol/g) that were significantly higher than GSH levels for saline-treated loaded animals (568 ± 83 nmol/g, P < 0.05).
GSSG levels and, in turn, GSSG/GSH ratios were higher in the saline-treated loaded group compared with control animals and with the NAC-treated loaded group (Fig. 6). Specifically, GSSG/GSH was 6.1 ± 1.3, 2.5 ± 0.4, 21.6 ± 6.5, and 7.8 ± 2.7% in saline-unloaded, NAC-unloaded, saline-loaded, and NAC-loaded groups, respectively (P < 0.01 for comparison of the saline-loaded group with the other groups of animals).
In this study, we found that NAC-treated animals were able to tolerate loading better than saline-treated control animals, exhibiting higher pressure and volume generation near the end of loaded breathing trials. We also found that administration of this agent increased the time during which animals could tolerate loading before they developed respiratory arrest. Although saline-treated loaded animals had significant reductions in diaphragmatic GSH levels, the magnitude of this reduction was blunted by NAC administration. Although it produced in vivo effects on pressure and volume generation, NAC failed to prevent significant reductions in the in vitro force-generating ability of diaphragmatic muscle strips removed from animals after respiratory arrest.
Methodological issues. It is possible that the use of a decerebrate preparation may have affected our results. Decerebration eliminated cortical projections to brain stem respiratory centers, and it is conceivable that removal of these influences may have altered the relative importance of central factors in determining the course of development of respiratory failure during loading. Although this is a concern, it would have been unethical to load unanesthetized intact animals to the point of respiratory arrest, and use of an anesthetized preparation would have resulted in other artifacts. Nevertheless, it is conceivable that decerebration may have affected our results by making peripheral factors (i.e., respiratory muscle function) a more important determinant of the development of respiratory failure, thereby accentuating the effect of NAC.
Proposed mechanisms of action of NAC.The findings of a number of studies are consistent with the notion that oxygen free radicals are generated in muscle during strenuous contraction and play a role in the genesis of skeletal muscle fatigue (2, 3, 8, 14, 17, 18, 21, 22). Indexes of free radical-mediated lipid and protein oxidation have been shown to increase in the diaphragm in response to strenuous contractions elicited by electrical stimulation of the diaphragm and by increasing the diaphragmatic workload of intact spontaneously breathing animals. For example, strenuous contractions have been shown to increase diaphragmatic levels of malondialdehyde and GSSG (markers of lipid peroxidation) while simultaneously decreasing concentrations of GSH (2, 4). Other studies examining electrically induced diaphragmatic contractions have shown that the rate of diaphragmatic fatigue can be attenuated by administration of any one of several free radical scavengers. Substances previously shown to provide this protective effect include superoxide dismutase, catalase, dimethylsulfoxide, and NAC (7, 12, 17, 22, 25).
Of note, NAC is a weak free radical scavenger in its own right but it also affects cellular function by supplying cysteine for the synthesis of intracellular GSH. GSH, in turn, can work in several ways to combat the effects of free radicals (6, 10, 14, 15). This protein serves as an alternative substrate for direct reaction with free radicals; it degrades H2O2via a glutathione peroxidase-catalyzed reaction; it reacts with lipid peroxides, restoring lipid structure; and it reacts by a mixed disulfide reaction with protein disulfide bonds, reconstituting protein sulfhydryl groups.
In theory, there are several potential mechanisms by which NAC may have altered the development of respiratory failure during inspiratory loading in this experiment. For one thing, this drug has been shown to alter vascular tone and redistribute cardiac output, and it is theoretically possible that such an action might have improved respiratory muscle function (13). In addition, recent studies also suggest that this drug has significant central nervous system effects (19). Although diaphragm-integrated EMG activity was not significantly different in NAC- and saline-treated loaded animals, this is a relatively crude index of motor outflow. It is conceivable that the pattern of diaphragmatic motor unit activation could have been altered, or that drive to other respiratory muscles may have changed as a consequence of administration of this agent, and that such undetected drive alterations may have contributed to the observed findings.
Alternatively, NAC administration may have slowed the development of respiratory failure by reducing the rate of development of free radical-induced respiratory muscle dysfunction in vivo, thereby better preserving the pressure-generating capacity of these muscles. Our observation that GSH levels were increased in the diaphragms of NAC-treated loaded animals is in keeping with this possibility. Whereas NAC administration failed to diminish the loading-induced reduction of in vitro diaphragm generation observed in the present study, this finding is not inconsistent with an action of this agent to affect in vivo muscle function by slowing the rate of development of free radical-mediated muscle dysfunction. If, in fact, the level of diaphragm fatigue influences the development of respiratory failure and the point of respiratory arrest, then one would expect an agent that reduced the rate of fatigue development to simply increase the time until a critical level of fatigue developed. At the point of arrest, one would then expect this same “critical” degree of fatigue to be present, whether or not such an agent was given. Our findings are entirely compatible with such a scenario, with NAC prolonging the time required for animals to reach a critical level of muscle fatigue. If the same critical level of fatigue was achieved at the point of respiratory arrest in both saline- and NAC-treated animals, then it is not surprising that the same level of in vitro contractile dysfunction was detected in diaphragms taken from these two groups of animals at the point of respiratory arrest.
This line of reasoning requires, however, some link between the degree of peripheral respiratory muscle dysfunction present and an occurrence of a phenomenon, i.e., respiratory arrest, that is clearly centrally determined. The amount of “peripheral” respiratory muscle fatigue can, in theory, influence central events in several ways. First, a reduction in respiratory muscle pressure and volume generation due to fatigue will raise arterial CO2levels and lower oxygen tensions. Whereas these changes in blood-gas tensions initially stimulate respiration centrally, severe hypercapnia and/or sustained hypoxemia can cause ventilatory depression (9). In addition, recent work suggests that central drive may be affected during respiratory loading as the result of alterations in neural traffic carried along sensory afferents arising in the diaphragm and stimulated by metabolic substances (e.g., acid, potassium, adenosine) released by the working diaphragm (20). Reflex inhibition of central drive by this latter mechanism may contribute to respiratory depression and respiratory arrest.
Although our findings are consistent with an action of NAC to delay the development of respiratory failure by blunting the development of free radical-induced respiratory muscle dysfunction, these data alone are not sufficient to prove that this pathophysiological sequence occurred. Moreover, even if, as we have postulated, this agent acted primarily by reducing free radical-mediated respiratory muscle dysfunction, it is worth noting that there was a trend toward glutathione oxidation in loaded animals despite NAC treatment. This would suggest that this agent, in these doses, may not entirely prevent free radical-mediated modification of cellular proteins and lipids.
We should also note that we were unable to demonstrate a beneficial effect of NAC treatment on the evolution of respiratory failure in an earlier study in which this agent was given to loaded animals not given oxygen (23). In this previous study, however, animals developed extremely severe hypoxemia almost immediately after the institution of respiratory loading, and it is possible that the depressant effect of this extreme hypoxia on respiratory muscle and central nervous system function was so profound and respiratory failure developed so quickly as to mask any beneficial effect on respiratory system performance provided by NAC. Supplemental oxygen was administered in the present study, thereby avoiding the development of severe hypoxemia at the beginning of loaded breathing trials.
Implications. Regardless of the mechanism, the fact that NAC was effective in preserving volume and pressure generation over time during loading breathing in this study may have clinical implications. This drug has recently been shown to be effective in reducing the development of limb muscle fatigue in humans when administered in doses similar to those used in the present study (19). It is, therefore, possible that this drug, or an agent like it, may prove useful in reducing respiratory muscle dysfunction in patients with an increased respiratory workload due to lung disease. In view of the variable effect that this agent has already been shown to have on the response to respiratory loading (i.e., it improved load tolerance in the present study but not in a previous experiment), it is evident that substantial additional experimentation is needed to define the usage of these agents.
Address for reprint requests: G. S. Supinski, Metrohealth Medical Center, 2500 Metrohealth Dr., Cleveland, OH 44109.
- Copyright © 1997 the American Physiological Society