INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

ACETAZOLAMIDE (ACZ) INHIBITS carbonic anhydrase (CA), an enzyme widely distributed in the body and concentrated in the proximal renal tubule and erythrocytes. It is effective in preventing acute mountain sickness (3, 10). In low dose (<5 mg/kg), Acz inhibits renal CA, causing a modest metabolic acidosis, which increases ventilation at rest (32), thereby increasing arterial PO₂ (Pa_O2), which simulates acclimatization (28, 31). At higher doses (7-12 mg/kg), inhibition of red cell and tissue CA leads to respiratory acidosis at the tissue level, which also increases ventilation (32). Acz under normoxic conditions is associated with a mild (<10%) reduction in exercise capacity (27, 30) associated with lower maximal O₂ consumption (O₂; O_2 max) and respiratory exchange ratio but increased ventilation (6, 27). As dexamethasone appears equally efficacious in preventing acute mountain sickness (10) without known acute effects on exercise performance, it is of importance to define the effect of Acz on exercise performance at altitude, given the strenuous nature of exertion in mountaineering.

Acz has been reported to have varying effects on exercise capacity and ventilation under hypoxic conditions. Schoene et al. (27) found no difference in maximal power but significantly increased O_2 max on Acz at an inspired O₂ fraction (FI_O2) of 0.118 under normobaric conditions, which was associated with modest increases in ventilation and O₂ saturation measured from pulse oximetry (Sp_O2) during exercise. Increased arterial and capillary PO₂ leading to increased O₂ delivery may have accounted for these findings, although these parameters were not successfully measured (27). In contrast, Stager et al. (30) found no effect on maximum power, O_2 max, ventilation, or Sp_O2 when studied under conditions of hypobaric hypoxia with barometric pressure of 446 mmHg. Also, using simulated altitude, McLellan et al. (22) found a reduced time to exhaustion at 90% O_2 max on Acz. At a higher altitude, acclimatized mountain climbers demonstrated modest increases in ventilation, but no difference in Sp_O2, and a trend to reduced exercise capacity (13). Potential mechanisms for reduced exercise capacity include the effect of acidosis on muscle function, dehydration due to diuretic effect, and increased dyspnea due to stimulation of ventilation.

Although there may be some inconsistency in the reported effect of Acz on ventilation and arterial O₂ saturation during exercise, there is clear evidence of increased ventilatory drive at rest, as evidenced by the finding of increased hypercapnic ventilatory response (HCVR) (1, 33, 36). Hypoxic ventilatory response (HVR) was not increased (1, 36), although this conclusion was based on measurement of HVR without Acz at baseline PCO₂, whereas HVR after Acz was measured at the new lower resting PCO₂ (hypocapnic HVR). Hypocapnia attenuates HVR (29), thus possibly offsetting any augmentation due to the drug itself. Any consideration of the impact of hypoxic ventilatory drive in explaining the effect of Acz in preventing acute mountain sickness would, therefore, need to eliminate the effects of CO₂ by holding PCO₂ at the same level with and without Acz (eucapnic HVR). Eucapnic HVR on Acz has been reported to be unchanged (12) or increased (1, 36). Our primary aim was to define the effect of Acz on exercise capacity in healthy subjects under hypoxic conditions. Second, we sought to determine which of the proposed mechanisms is likely to explain any observed effect. Third, we wished to explore further the issue of dyspnea in the context of the reported effects of Acz on ventilatory drive. We, therefore, measured HCVR and HVR in the same set of subjects, in an effort to explore the links between drive and exertional dyspnea.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects. Eleven healthy, nonsmoking, male subjects were enrolled in the study, which had the approval of the institutional research ethics committee. Each subject had normal examination and spirometry before commencement and was weighed at the start of each visit. Subjects were studied on three occasions, with the first study as a familiarization study on no medication. Each visit involved measurement of HVR and HCVR followed by maximal incremental exercise at FI_O2 of 0.13 at sea level. After the familiarization study, subjects were randomized in concealed fashion to receive either Acz (500 mg twice daily for 21/2 days, including the morning of the study) or placebo capsules before the second study 1 wk later. After a 2-wk washout, subjects crossed over to be treated for a further 21/2 days with the alternative drug before the third visit. Subjects and investigators were blinded to randomization status, although subjects were aware of potential side effects of Acz. Subjects were told that the effect of Acz on hypoxic exercise was not certain and that there were sound physiological reasons to expect changes in either direction.

Ventilatory response. Resting end-tidal CO₂ concentration was measured over 5 min by using an infrared CO₂ analyzer (CD-3A, Applied Electrochemistry, Pittsburgh, PA). Subsequently, two HVR runs were performed in a modification of the technique described by Rebuck and Campbell (24) using room air as the filling gas for a 6-liter rebreathing bag. End-tidal CO₂ concentration was maintained at the resting level by using manually varied flow through a CO₂-absorbing bypass circuit. Sp_O2 was measured by using a pulse oximeter (Radiometer, Copenhagen, Denmark). Ventilation was measured with a rotating-vane ventilation monitor (PK Morgan, Chatham, UK). Borg category ratio scale scores (0-10) for dyspnea (4) were recorded each 30 s after 1 min, with each run terminated when excessive distress or O₂ saturation <70% occurred. For the second and third visits, if the baseline end-tidal CO₂ was >0.2% different from the baseline level at the first visit, a further two runs at the baseline level of the first visit were performed to enable comparison between studies under eucapnic conditions. HCVR was measured as described by Read (23), with ventilation and Borg scores for dyspnea recorded each 30 s and cessation at 4 min or when excessive distress occurred. Results of each pair of runs were averaged for comparison.

Exercise. Maximal incremental exercise was performed with the inspired gas mixture delivered from a 25-liter reservoir bag supplied by a high-flow air-nitrogen blender, allowing flow of up to 150 l/min with FI_O2 adjusted to 0.13 and monitored continuously by a paramagnetic O₂ analyzer (PK Morgan). Expired gas distal to a 10-liter mixing chamber was analyzed with a zirconia O₂ analyzer (S3A1, Applied Electrochemistry) and infrared CO₂ analyzer (CD-3A, Applied Electrochemistry). End-tidal CO₂ was measured on an infrared CO₂ analyzer (901-MK2, PK Morgan). In addition to all gas analyzer outputs, inspired ventilation and ear oximetry were recorded continuously on a multichannel chart recorder (TA2000, Gould, Cleveland, OH). After a 5-min rest period, an arterial blood-gas (ABG) sample was obtained by radial artery puncture (rest ABG) and analyzed in a blood-gas analyzer (ABL 520, Radiometer). This was followed by an arterialized capillary blood-gas (CBG) sample collected in two to three capillary tubes from a small incision in an earlobe vasodilated with nicotinic acid ointment (rest CBG). Incremental exercise was performed (10) on a calibrated, mechanically braked cycle ergometer (Monark Exercise AB, Varberg, Sweden) with 1-min increments fixed for each subject, calculated to achieve an exercise time of ~10 min (7). The increment was determined by dividing the predicted (normoxic) maximum exercise capacity by 10 and then multiplying by a further correction factor of 0.75 to allow for reduction in exercise capacity at altitude with a similar level of hypoxia (28). In practice, each subject had a calculated increment of either 17 or 20 W. Each minute, heart rate, ECG, and Borg category ratio scale scores for dyspnea and leg fatigue (4) were recorded. Near completion of exercise, a repeat arterialized CBG (peak CBG) was sampled. Exercise capacity was taken as the highest increment that the subject maintained for at least 30 s. At the completion of exercise, subjects were asked to describe the predominant limiting symptom.

Statistics. Normally distributed variables were compared by using paired t-test. Nonparametric variables were compared by using Wilcoxon's matched-pairs signed-ranks test. Limiting symptoms were compared by using the ² test. Variables are presented as means ± SD, unless otherwise stated. All statistics were performed by using SPSS statistical package, and P < 0.05 was required to reject the null hypothesis.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Two subjects withdrew from the study (one developed viral symptoms and mild neutropenia on the study drug; another withdrew for personal reasons). This left nine subjects with age of 27.9 ± 2.9 yr, height of 178.4 ± 5.0 cm, and baseline hypoxic exercise capacity of 219 ± 36 W, which was 88 ± 14% of predicted normoxic exercise capacity (18).

Ventilatory studies. See Table 1. In two subjects, HCVR could not be measured adequately due to presyncope and bag emptying, respectively. In the remaining seven subjects, Acz was associated with a 32% increase in HCVR slope and a significant left shift in x-intercept (Fig. 1). In contrast, HVR was not significantly affected by Acz under hypocapnic or eucapnic conditions.

View larger version (11K): [in this window] [in a new window]   Fig. 1.   Hypercapnic ventilatory responses on placebo and acetazolamide for subject 6. I, inspired minute ventilation.

View larger version (13K): [in this window] [in a new window]   Fig. 2.   Plot of Borg score for dyspnea vs. ventilation during hypercapnic ventilatory response measurement in subject 6.

Exercise. Exercise capacity on Acz was reduced by a mean of 10%, with eight of nine subjects having reduced maximum power and the remaining subject able to maintain the same maximum power (see Table 2). FI_O2 was similar, being 0.1292 ± 0.0017 for placebo and 0.1288 ± 0.0024 for Acz. Weight on Acz was 78.1 ± 12.0 kg, which was <78.5 ± 12.1 kg on placebo (P < 0.05). All nine subjects reported leg fatigue as the predominant limiting symptom on placebo. On Acz, one subject reported combined dyspnea and leg fatigue, with the remainder all citing leg fatigue as the predominant limiting symptom (P = 0.30). To adjust for the confounding effect of achieved exercise capacity, Table 3 compares results at the maximum power on Acz with results at the same power increment (submaximal in eight of the nine subjects) on placebo. Although Borg category ratio scale scores for leg fatigue were equivalent when compared at peak exercise, when compared at the same power, the median Borg score on Acz was 10, corresponding to the descriptor "very, very heavy," compared with the median placebo Borg score of 8, corresponding between descriptors "very heavy" and "very, very heavy" (P < 0.02).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

We found a 10% reduction in exercise capacity under hypoxic conditions on Acz. There was increased ventilatory effort on Acz, but the predominant limiting symptom was the perception of leg fatigue. Leg fatigue was the same at maximum exercise, despite the lower maximum power on Acz, and leg fatigue was increased on Acz when compared at the same power, indicating that Acz is associated with increased perception of leg fatigue during maximal exercise.

The increased leg muscle fatigue on Acz may be due to the effect of acidosis within muscle cells. Acidosis decreases endurance time during exercise and inhibits glycolytic enzymes, especially phosphofructokinase (19). We have shown reduced arterialized capillary pH at near-peak exercise on Acz. This reduction would have been greater still if it had been possible to allow for the confounding effect of the higher power increment at which placebo peak CBG was sampled. As shown previously (27), we showed reduced CO₂ production during incremental exercise, which reflects slowed CO₂ production kinetics with CA inhibition (25). CA inhibition without metabolic acidosis causes slower recovery of muscle pH after progressive exercise (21). Therefore, impaired CO₂ elimination is also likely to have contributed to decreased muscle pH in our subjects.

Diuretic effects of CA inhibition may also impair exercise. Acz decreases body weight by 2% within 24 h (30), which may be associated with impaired stroke volume, thermoregulation, and ability to maintain exercise (6). We observed a 0.5% reduction in weight on Acz, suggesting very mild dehydration, but, because maximal heart rate was lower on Acz, this is unlikely to be the limiting factor to exercise.

An additional potential mechanism of increased leg fatigue is the likely effect of increased work of breathing associated with increased ventilation to reduce O₂ delivery to working muscles. Blood flow and O₂ to the respiratory muscles approximate 14-16% of total blood flow and O₂ during maximal exercise (15). During heavy exercise, additional increases in work of breathing cause vasoconstriction in locomotor muscles (14), which can impair exercise performance (16). Although increased work of breathing might also have been expected to increase perception of dyspnea as an additional factor limiting exercise, no increase in dyspnea was observed. A further potential mechanism for increased perception of leg fatigue is the reported effect of CA inhibition to impair neural input to muscle (5) and impair muscle function (20).

The reduction in hypoxic exercise capacity on Acz contrasts with studies under sea-level and hypobaric hypoxia, in which no significant limitation of maximum exercise was seen (13, 27, 30). This may be attributed to the higher dose of Acz used in our study, which is at the upper limits of the usual clinical dosing. At doses of 7-12 mg/kg (as in this study), significant inhibition of red cell and tissue CA may develop (32) and lead to inefficient CO₂ transport during exercise.

An alternative explanation is that previous studies have used higher power increments of up to 50 W, which may have reduced their sensitivity for detecting the relatively small change in exercise capacity found in our study.

These results do not detract from the proven benefit of Acz in reducing acute mountain sickness during rapid ascent (3). Although exercise capacity was reduced under conditions of acute hypoxia at sea level, the difference is relatively small and would not necessarily occur with longer term exposure to altitude. A previous study at 4,846 m showed better exercise performance at altitude in acclimatized subjects on Acz, with less weight and muscle loss (2). Additional factors are likely to influence exercise performance in acclimatized subjects at altitude, such as higher ventilatory drive with higher Pa_O2 and polycythemia, which could increase O₂ delivery to tissues. Additional consequences of more prolonged hypoxia at altitude include high-altitude pulmonary edema (11). Acz reduces hypoxic pulmonary vasoconstriction, which may prevent high-altitude pulmonary edema (9). The modest detrimental effect on exercise capacity of Acz in the relatively high dose used in our study needs to be weighed against its proven benefits by those undertaking high-intensity exercise at altitude.

The similar peak CBG PCO₂ appears to contrast with the demonstration that Acz stimulates ventilation during exercise compared with placebo at the same power. However, inhibition of red cell CA is associated with blood-gas disequilibrium, such that the PCO₂ measured at equilibrium in the blood-gas analyzer is higher than the PCO₂ in disequilibrium within the arterial circulation (8, 34).

Acz did not increase dyspnea during exercise, which was unexpected given the increase in ventilation. We speculate that this intriguing finding, together with our demonstration of reduced dyspnea during HCVR, indicates that stimulating ventilatory drive does not come at the cost of worsening breathlessness during exercise, an observation of considerable clinical utility. It is possible that impaired neural input to respiratory muscle caused by Acz (5) may alter the perception of dyspnea.

The increase in HCVR on Acz is attributable to metabolic acidosis (33). By comparison, there was little effect on HVR. Metabolic acidosis increases HVR (33), but this is offset by direct suppression of peripheral chemoreceptors by CA inhibition (17, 26, 35). Previous studies have found significant decreases in hypocapnic HVR and unchanged or increased eucapnic HVR on Acz (1, 12, 33, 36). However, even if there is a real increase in eucapnic HVR on Acz, this would be of doubtful importance in practice, because high-altitude acclimatization is known to be accompanied by marked hypocapnia (37).

In summary, the increase in resting ventilatory drive due to Acz is associated with increased ventilatory response to hypercapnia, but not hypoxia. Acz continues to stimulate ventilation during maximal exercise at a FI_O2 of 0.13, with associated increases in Pa_O2 and arterial O₂ saturation. There is no increase in dyspnea during exercise on Acz, despite the increase in ventilation. Exercise capacity is reduced in association with an increased perception of leg effort.