Effects of acetazolamide on cerebrovascular function and breathing stability at 5050 m

Abstract

One of the many actions of the carbonic anhydrase inhibitor, acetazolamide (ACZ), is to accelerate acclimatisation and reduce periodic breathing during sleep. The mechanism(s) by which ACZ may improve breathing stability, especially at high altitude, remain unclear. We tested the hypothesis that acute i.v. ACZ would enhance cerebrovascular reactivity to CO₂ at altitude, and thereby lower ventilatory drive and improve breathing stability during wakefulness. We measured arterial blood gases, minute ventilation () and middle cerebral artery blood flow velocity (MCAv) before and 30 min following ACZ administration (i.v. 10 mg kg⁻¹) in 12 healthy participants at sea level and following partial acclimatisation to altitude (5050 m). Measures were made at rest and during changes in end-tidal and (isocapnic hypoxia). At sea level, ACZ increased resting MCAv and its reactivity to both hypocapnia and hypercapnia (P < 0.05), and lowered resting , arterial O₂ saturation () and arterial () (P < 0.05); arterial () was unaltered (P > 0.05). At altitude, ACZ also increased resting MCAv and its reactivity to both hypocapnia and hypercapnia (resting MCAv and hypocapnia reactivity to a greater extent than at sea level). Moreover, ACZ at altitude elevated and again lowered resting and (P < 0.05). Although the sensitivity to hypercapnia or isocapnic hypoxia was unaltered following ACZ at both sea level and altitude (P > 0.05), breathing stability at altitude was improved (e.g. lower incidence of ventilatory oscillations and variability of tidal volume; P < 0.05). Our data indicate that i.v. ACZ elevates cerebrovascular reactivity and improves breathing stability at altitude, independent of changes in peripheral or central chemoreflex sensitivities. We speculate that -mediated elevations in cerebral perfusion and an enhanced cerebrovascular reactivity may partly account for the improved breathing stability following ACZ at high altitude.

Key points

• Acetazolamide improves breathing stability during sleep in newcomers to high altitude, but the mechanism remains unclear.

• We examined the effects of a single i.v. dose of acetazolamide on brain vascular function and breathing at sea level and following 7 days at high altitude (5050 m).

• We demonstrated that acute i.v. acetazolamide at high altitude enhances the brain blood flow response to changes in CO₂ and improves breathing stability.

• We speculate that the enhanced brain blood flow responses following acetazolamide ingestion may account for the well-documented acetazolamide-induced improvement in abnormal breathing at high altitude.

Introduction

The stability of breathing rhythm is thought to depend strongly on the ‘CO₂ reserve’, i.e. the difference between the eupnoeic pressure of CO₂ in arterial blood () and the hypocapnic threshold for apnoea (see Dempsey (2005) for review). The CO₂ reserve is reduced when breathing hypoxic air because of increased ventilation (due to the peripheral chemoreflex activation) lowering , thereby increasing the risk of periodic breathing during sleep (Xie et al. 2001, 2006). Ascent to high altitude in newcomers often leads to unstable breathing during both wakefulness (Brusil et al. 1980; Waggener et al. 1984) and sleep (Sutton et al. 1979, 1980; Lahiri et al. 1983; Normand et al. 1990; Lahiri & Data, 1992). Typically an oscillatory pattern of breathing can be distinguished, especially during sleep, with changes in tidal volume that can develop into periodic breathing with central apnoea periods. This development of breathing instability is probably due to a greater increase in the hypoxic ventilatory response associated with high altitude-induced hypoxaemia compared with the increase in background ventilatory drive, resulting in a reduced CO₂ reserve (Khoo et al. 1982; Dempsey, 2005). The carbonic anhydrase inhibitor acetazolamide (ACZ) diminishes breathing pattern disturbances upon ascent to high altitude, especially during sleep (Weil et al. 1978; Sutton et al. 1979, 1980). At sea level, ACZ increases the CO₂ reserve in both humans (Teppema et al. 2010) and anaesthetised cats (Teppema et al. 2001). However, since both the peripheral (Hackett et al. 1987) and central chemoreflex sensitivities (Burki et al. 1992) appear to be preserved following ACZ at high altitude, it seems unlikely that a blunted ventilatory drive would account for the increase in CO₂ reserve. Accordingly, the prophylactic mechanisms of action by which ACZ improves breathing stability during wakefulness and sleep remain poorly understood.

One potential mechanism by which ACZ may improve the CO₂ reserve is via an effect on cerebral blood flow (CBF). Several studies have demonstrated a link between resting CBF and its response to CO₂ (termed cerebrovascular CO₂ reactivity), and the CO₂ reserve (Xie et al. 2005, 2009). For example, pharmacologically induced reductions in resting CBF and cerebrovascular CO₂ reactivity reduce the CO₂ reserve during sleep (Xie et al. 2009) and cause breathing instability during wakefulness (Fan et al. 2010b). It has been well documented that, at sea level, acute i.v. administration of ACZ rapidly elevates resting CBF (Ehrenreich et al. 1961; Hauge et al. 1983; Lassen et al. 1987; Jensen et al. 1990) without altering cerebral metabolism (Posner & Plum, 1960; Vorstrup et al. 1984). However, no studies have examined the acute effects of a single dose of ACZ on CBF, cerebrovascular CO₂ reactivity and breathing stability at both sea level and at high altitude. Therefore, although plausible, it remains unclear whether acute i.v. ACZ administration at high altitude alters cerebrovascular function and subsequently modulates breathing stability.

In the present study, we examined the effect of i.v. ACZ on cerebrovascular and ventilatory responsiveness to CO₂ and O₂ at sea level and following partial acclimatisation to 5050 m. We selected to use the i.v. route of administration to avoid the confounding factor of altered acid–base balance associated with oral ACZ administration (Swenson, 1998). Based on previous studies that show a worsening of breathing stability following experimental reductions in CBF and cerebrovascular CO₂ reactivity (Xie et al. 2009; Fan et al. 2010b), we tested the hypothesis that acute ACZ-induced elevations in CBF and reactivity would lead to stabilisation of breathing control in partially acclimatised newcomers to 5050 m. These experiments were part of a series investigating the cardiorespiratory and cerebrovascular adaptations and mechanisms of regulation over 14 days at 5050 m. Although some of the experiments performed during this expedition have already been published (Fan et al. 2010a,b; Thomas et al. 2010; Lucas et al. 2011), apart from the experimental design and control data at high altitude from half of this group (Lucas et al. 2011), there is no overlap or duplication of data with the present study described herein.

Methods

Participants

Twelve sea-level residents (eight male and four female) with a mean age of 30 ± 10 years (mean ± SD) and body mass index of 23 ± 2 kg m⁻² participated in this study. Participants were non-smokers, had no previous history of cardiovascular, cerebrovascular or respiratory diseases and were not taking any medication.

Ethical approval

The study was approved by the Lower South Regional Ethics Committee of Otago and conformed to the standards set by the Declaration of Helsinki. All participants were informed regarding the purposes and procedures of this study, and informed consent was given prior to participation.

Experimental design

After a full familiarisation with the experimental procedures outlined below (on the first visit), the participants underwent experimental trials at sea level and at 5050 m (the Ev-K2-CNR Pyramid Laboratory, Nepal; barometric pressure 413 ± 1 mmHg). At sea level, participants underwent three experimental trials in randomised order (ACZ, indomethacin and placebo). Due to time constraints, no placebo trial was carried out at 5050 m. The ascent profile from sea level (Dunedin, New Zealand) to the Pyramid Laboratory has been described previously (Fan et al. 2010a). To avoid any confounding influence of acute mountain sickness (AMS), experimental sessions were carried out between days 5–11 (7 ± 2 days) after arrival to 5050 m, after any symptoms from AMS had subsided. The present study was part of a larger experiment – to be reported elsewhere – in which the participants were administrated indomethacin (100 mg, oral) on a separate day (single-blinded, randomised order and separated by at least 3 days). Before each experimental session, participants abstained from exercise and alcohol for 24 h, caffeine for 12 h and a heavy meal for 4 h prior.

With the exception of the arterial blood gas sampling, which was conducted following a 10 min supine rest, all experiments were performed with participants in a semi-recumbent position. Following 10–15 min of quiet rest, each experimental testing session consisted of: (a) an arterial blood gas sample; (b) instrumentation; (c) 5 min resting baseline; (d) modified hyperoxic rebreathing and isocapnic hypoxia (see details of methods below); (e) ACZ administration (i.v. 10 mg kg⁻¹) or placebo (sea level only); (f) 30 min rest; and (g) repeat testing of a–d. The order of the modified rebreathing and isocapnic hypoxic rebreathing was randomised and 5 min recovery was permitted between each trial to restore end-tidal gases to baseline resting values.

Ventilatory stability

Details of the ventilatory stability analysis have been previously described (Fan et al. 2010b). In brief, we first established whether unstable breathing patterns were present by visually analysing the ventilation () traces obtained during spontaneous room air breathing during the final 2 min of the 5 min baseline period, based on absence or presence of Cheyne–Stokes-like respiratory patterns, characterised by distinct waxing and waning of tidal volume (V_T) and breathing frequency (f) pattern (Cherniack & Longobardo, 1973). If the trace was considered unstable, the difference between the peak and trough of any oscillation in was measured. The average of this difference (magnitude of oscillation) was subsequently used as an index of ventilatory stability. The number of oscillations during this 2 min period was also noted. In addition, we calculated the co-efficient of variation of the breath-by-breath , V_T and f data (variability) during the 2 min period.

Modified rebreathing method

The modified rebreathing method is a well-established method for assessing both ventilatory and cerebrovascular CO₂ reactivities (Ainslie & Duffin, 2009). We selected the modified rebreathing method over a steady-state method to limit the confounding influence of a gradient on ventilatory control (Fan et al. 2010b). Since hyperoxia (≥ 150 mmHg) diminishes peripheral chemoreceptors’ output (Cunningham et al. 1963; Gardner, 1980), the ventilatory response to the modified rebreathing method can be interpreted as the ventilatory CO₂ sensitivity primarily from the central chemoreflex. The details of the modified rebreathing method have been previously described in Fan et al. (2010a). In brief, the participants wore a nose clip and breathed through a mouthpiece connected to a Y-valve which allowed switching from room air to a 6 l rebreathing bag filled with 7% CO₂ and 93% O₂. Following baseline room air breathing, participants were instructed to hyperventilate for 5 min to lower and then maintain a partial pressure of CO₂ () at 22 ± 2 mmHg (sea level) or 17 ± 3 mmHg (5050 m). Participants were then switched to the rebreathing bag following an expiration and instructed to take three deep breaths to ensure rapid equalisation of in the rebreathing circuit. The rebreathing tests were terminated when either: (i) reached 60 mmHg; (ii) partial pressure of end-tidal O₂ () dropped below 160 mmHg; (iii) exceeded 100 l min⁻¹; or (iv) the participant reached the end of their tolerance.

The rebreathing data were analysed on a breath-by-breath basis using a specially designed programme (Full Fit Rebreathing programme, Version 3.1, University of Toronto, Toronto, Canada) and the analysis has been described in detail previously (Mohan et al. 1999; Duffin et al. 2000).

Isocapnic hypoxia

The soda-lime rebreathing technique was used to assess the ventilatory O₂ sensitivity as an index of peripheral chemoreflex sensitivity (Mathew et al. 1983). Participants wore a nose clip and breathed through a mouthpiece connected to a Y-valve allowing switching from room air to a circuit consisting of a 6 l rebreathing bag and a soda-lime reservoir. The protocol began with baseline room air breathing, before participants were switched to the rebreathing circuit at the end of an inspiration. Participants filled the rebreathing bag with room air drawn in through their nose and expired into the bag. Once the bag was filled (ensuring that this was at the end of expiration) the nose clip was attached and rebreathing began. The isocapnic hypoxia was terminated when either: (i) peripheral O₂ saturation () reached 80% at sea level and 70% at 5050 m; (ii) decreased to 45 mmHg at sea level and 30 mmHg at 5050 m; (iii) the exceeded 100 l min⁻¹; or (iv) the participant reached the end of their tolerance.

The breath-by-breath data during the isocapnic hypoxic rebreathing were plotted against and an inverse first-order polynomial function was used to obtain the hypoxic ventilatory response curve (Day & Wilson, 2009):

where y₀ is the y asymptote, x is the in mmHg and c is the curvature (representing the responsiveness).

Steady-state hypocapnia

Voluntary hyperventilation was used to assess cerebrovascular reactivity to hypocapnia (Xie et al. 2005). Steady-state hypocapnic cerebrovascular reactivity was estimated from the slope of the reduction in mean MCAv from the final 2 min of baseline to the voluntary hyperventilation which preceded rebreathing, relative to the reduction in (see below). We did not control for during the voluntary hyperventilation because the small increases in (by 29 ± 11 and 16 ± 12 mmHg at sea level and 5050 m, respectively) were considered unlikely to influence either MCAv or its reactivity.

ACZ administration

Slow (∼60 s) i.v. administration of ACZ (10 mg kg⁻¹; Diamox) or placebo (saline; sea level only) was achieved via an indwelling catheter located in an antecubital vein. The ACZ dose was reconstituted in approximately 5 ml of saline; a comparable volume (7–10 ml, depending on body mass) of i.v. saline was administered in the placebo trial. At sea level, the order was randomised and the participants were blinded to the condition.

Measurements

Respiratory variables

and its components of V_T and f were measured using a heated pneumotachograph (Hans-Rudolph 3813) and expressed in units adjusted to BTPS. and were measured using fast-responding gas analysers (model CD-3A, AEI Technologies, Pittsburgh, PA, USA; ML206 and ML240, ADInstruments, Colorado Springs, CO, USA). The pneumotachograph was calibrated using a 3 l syringe (Hans-Rudolph 5530) and the gas analysers were calibrated using gas mixtures of known concentrations of O₂ and CO₂ prior to each testing session.

Cerebrovascular variables

Cerebral blood flow velocity was measured in the right middle cerebral artery using a 2 MHz pulsed Doppler ultrasound system (MCAv; DWL, Compumedics Ltd, Germany). The Doppler ultrasound probe was positioned over the right temporal window and held in place with an adjustable headband. Optimal signals were obtained using search techniques described elsewhere (Aaslid et al. 1982). Beat-to-beat mean arterial blood pressure (MAP) was monitored using finger photoplethysmography (Finometer, Finapress Medical Systems, the Netherlands). Manual blood pressure measurements by auscultation were also made periodically to check and validate the automated recordings. Cerebrovascular conductance index (CVCi) was subsequently estimated by dividing mean MCAv by MAP within each breath cycle to reveal intrinsic vascular responses to CO₂.

Arterial blood gases

Arterial blood gas samples from a radial artery were obtained at rest using a 25-gauge needle into a pre-heparinised syringe. Following standardised calibration, all blood samples were analysed using an arterial blood-gas analysing system (NPT 7 series, Radiometer, Copenhagen, Denmark) for pH, partial pressure of arterial O₂ () and CO₂ (), bicarbonate concentration ([HCO₃⁻]) and arterial O₂ saturation ().

With the exception of the arterial blood gas variables, all data were acquired at 1000 Hz using an analog-to-digital converter (PowerLab; ADInstruments) with commercially available software (Chart version 5.5.6, ADInstruments), and stored on computer for later analysis.

Statistical analysis

The effects of altitude and ACZ on resting variables, cardiorespiratory and cerebrovascular responsiveness to CO₂, as well as the ventilatory variability were assessed using two-way (altitude and drug) repeated-measures ANOVA with an α-level of 0.05 (SPSS version 17.0, SPSS, Chicago, IL, USA). Pair-wise comparisons (Bonferroni corrected) were performed to isolate the effect of altitude and ACZ on the dependent measures within participants. Data are reported as mean ± SD.

Results

All 12 participants were able to complete the whole experimental protocol. Comparison of -CO₂ sensitivity could be carried out in only 11 participants, both at sea level and at 5050 m, due to a poor ventilatory flow trace during hyperoxic rebreathing. Comparisons of MAP-CO₂ reactivity were conducted in 9 and 10 participants at sea level and 5050 m, respectively, due to poor BP traces in remaining participants.

Resting variables (Table 1)

Table 1.

Resting cerebrovascular, respiratory, cardiovascular and arterial blood gas variables before and after i.v. acetazolamide (ACZ) at sea level and following partial acclimatisation to 5050 m

	Sea level		5050 m
	Control	ACZ	Control	ACZ
Cerebrovascular
MCAv (cm s−1)	70 ± 13	80 ± 16†	73 ± 12	94 ± 15†
CVCi (cm s−1 mmHg−1)	0.88 ± 0.21	0.95 ± 0.20	0.82 ± 0.17	1.08 ± 0.26†
Respiratory
(l min−1)a	14.3 ± 3.9a	13.0 ± 3.7†	17.4 ± 3.0c	16.6 ± 2.7*†a
f (breaths min−1)	15 ± 4a	14 ± 4†	18 ± 5*	17 ± 3*†
(l)	0.95 ± 0.13a	0.95 ± 0.16	1.02 ± 0.31c	0.99 ± 0.22a
(mmHg)	42 ± 5	40 ± 5†	24 ± 3*	25 ± 3*†
(mmHg)	103 ± 6	103 ± 5	47 ± 4*	49 ± 7*
Cardiovascular
MAP (mmHg)	80 ± 10	84 ± 7	90 ± 9	89 ± 15
HR (beats min−1)	68 ± 9	60 ± 8†	77 ± 11*	77 ± 11*
Arterial blood gases
pH	7.46 ± 0.04b	7.44 ± 0.04†b	7.46 ± 0.02	7.44 ± 0.02†
(mmHg)	42 ± 4b	42 ± 4b	27 ± 4*	31 ± 3*†
(mmHg)	97 ± 6b	93 ± 8†b	46 ± 4*	42 ± 3*†
(%)	98.1 ± 0.5b	97.5 ± 0.5†b	82.9 ± 3.3*	79.2 ± 2.7*†
[HCO3−] (mmol l−1)	29.8 ± 3.0b	28.7 ± 3.1b	18.7 ± 3.0*	20.7 ± 2.7*†
SBE−	6.1 ± 2.9bg	4.7 ± 3.2b	−3.9 ± 4.0*	−2.8 ± 2.7*

Values are means ± SD.

^*Different from sea level (P < 0.05)

^†different from control (P < 0.05).

n = 12, except for ^an = 11, ^bn = 9 and ^cn = 10.

Cerebrovascular variables

No differences were observed in either MCAv or CVCi following ∼7 days of living at 5050 m (P = 0.489 and 0.452 vs. sea level, respectively). The administration of ACZ increased MCAv at sea level and at high altitude, with greater increases at altitude following partial acclimatisation to 5050 m (interaction effect: P = 0.007). Specifically, ACZ administration at sea level elevated MCAv by 15 ± 15% (P = 0.003 vs. control) without increasing CVCi (P = 0.325), whereas ACZ administration at 5050 m increased MCAv by 28 ± 11% (P < 0.001 vs. control) and CVCi by 33 ± 23% (P = 0.001 vs. control; interaction effect: P = 0.046).

Respiratory variables

Residing at high altitude elevated both and f (P = 0.030 and 0.009, respectively), whereas administration of ACZ lowered these variables (P = 0.045 and 0.030), but not by a different extent at sea level vs. high altitude (interactions: P = 0.636 and 0.895). The V_T was unaffected by altitude (P = 0.509) and ACZ administration (P = 0.674). High altitude lowered (P < 0.001), whereas ACZ administration had a differential effect on at high altitude compared with at sea level (interaction effect: P = 0.010). Specifically, was lowered by ACZ at sea level (P = 0.027 vs. control) but elevated at 5050 m (P = 0.02). In contrast, ACZ administration did not alter at sea level or its reduction at high altitude (altitude, P < 0.001; drug, P = 0.427; interaction, P = 0.572).

Cardiovascular variables

High altitude tended to elevate MAP (P = 0.057), whereas no effect of ACZ on MAP was evident for pooled altitudes (P = 0.654) or differentially between altitudes (interaction: P = 0.294). Resting HR was elevated at high altitude (P = 0.017). The administration of ACZ lowered resting HR at sea level (P = 0.002 vs. control) but not at 5050 m (P = 0.650; interaction effect: P = 0.021).

Arterial blood gas variables

At 5050 m, pH was similar to sea-level values (P = 0.700), indicating complete renal correction of the hypoxia-induced respiratory alkalosis and therefore well-advanced acclimatisation to this altitude at this time point. The and SBE⁻ both remained lower than at sea level (P = 0.001). Regardless of the altitude, ACZ administration lowered pH and , while SBE-remained unchanged (drug effects: P = 0.015, 0.012 and 0.562, respectively; interaction effects: P = 0.831, 0.936 and 0.134). Ascent to 5050 m lowered resting , and [HCO₃⁻] (all P < 0.001 vs. sea level). The effect of ACZ on , and [HCO₃⁻] was greater at 5050 m compared with at sea level (interaction effects: P = 0.003, 0.018 and 0.024). Specifically, ACZ at sea level lowered (P = 0.013) without measurably changing or [HCO₃⁻] (P = 1.000 and 0.342), but at high altitude it caused a larger reduction in (P = 0.002) with elevations in resting and [HCO₃⁻] (both P < 0.001).

Following the placebo trial at sea level, cardiorespiratory, cerebrovascular and blood gas variables were unchanged (data not shown, for clarity).

Breathing stability (Table 2)

Table 2.

Effect of acetazolamide (ACZ) on breathing stability at sea level and following ascent to 5050 m

	Sea level		5050 m
	Control	ACZ	Control	ACZ
Magnitude of oscillation (Δ l min−1)	5.8 ± 5.8	2.5 ± 5.3†	7.7 ± 5.2	5.9 ± 6.1†
Frequency of oscillation (event min−1)	1.3 ± 1.2	0.3 ± 0.6†	3.0 ± 0.5*	1.8 ± 0.4*†
Variability
(%)	29 ± 17a	29 ± 17†a	36 ± 17a	22 ± 10†a
f (%)	18 ± 5a	16 ± 8†a	23 ± 10a	16 ± 7†a
(%)	26 ± 14a	27 ± 19a	37 ± 17a	21 ± 10†a

Values are means ± SD.

^*Different from sea level (P < 0.05)

^†different from control (P < 0.05).

n = 12, except for a n = 11.

Ascent to 5050 m increased the frequency of oscillation (P < 0.001; interaction effect, P = 0.551), but did not alter its magnitude, or the variability of f, V_T or (P = 0.172, 0.249, 0.076 and 0.929). At sea level and 5050 m, ACZ administration reduced both the magnitude and frequency of the oscillations (P = 0.027 and < 0.001 vs. control, respectively; interaction effect, P = 0.489 and 0.551); the variability in f and were also reduced (P = 0.033 and 0.014, respectively; interaction effect, P = 0.283 and 0.067). The ACZ administration reduced the variability of V_T at 5050 m (P = 0.011) but not at sea level (P = 0.615; interaction effect, P = 0.042).

Modified rebreathing

Cerebrovascular CO₂ reactivity (Fig. 1)

Figure 1. Effects of acetazolamide (ACZ) and high altitude (5050 m) on cerebrovascular responsivess to hypercapnia and hypocapnia.

A, individual slopes; B, group data (mean ± SD). Filled circles, sea level control; open circles, sea level ACZ; filled squares, 5050 m control; open squares, 5050 m ACZ. ACZ elevated the hypercapnic and the hypocapnic cerebrovascular reactivity at both sea level and at 5050 m. *Different from sea level (P < 0.05); † different from control (P < 0.05).

Ascent to 5050 m elevated the rebreathing hypercapnic MCAv-CO₂ reactivity by 114 ± 146% (P = 0.008 vs. sea level). ACZ administration elevated hypercapnic MCAv-CO₂ reactivity (P = 0.030) both at sea level and 5050 m (interaction effect, P = 0.355). One participant displayed a high cerebrovascular reactivity to hypercapnic following ACZ at sea level. Re-analysis of the data with this individual excluded revealed a tendency for the hypercapnic MCAv-CO₂ to be elevated with ACZ at sea level (P = 0.052). Ascent to 5050 m elevated the hypocapnic MCAv-CO₂ reactivity by 114 ± 71% (P < 0.001 vs. sea level). The administration of ACZ elevated hypocapnic MCAv-CO₂ reactivity at sea level (by 31 ± 43%; P = 0.026) and by a greater amount at 5050 m (by 69 ± 60%; P < 0.001; interaction effect, P = 0.003).

Ventilatory CO₂ sensitivity (Fig. 2)

Figure 2. Effects of acetazolamide (ACZ) and high altitude (5050 m) on ventilatory and cardiovascular responsivess to hypercapnia.

A, individual slopes; B, group data (mean ± SD). Filled circles, sea level control; open circles, sea level ACZ; filled squares, 5050 m control; open squares, 5050 m ACZ. Ascent to 5050 m enhanced the cardiovascular response to CO₂, while no changes were observed following ACZ at either sea level or 5050 m. *Different from sea level (P < 0.05).

High altitude tended to elevate the -CO₂ sensitivity (P = 0.085). Meanwhile, no changes were observed in the -CO₂ sensitivity following ACZ administration (P = 0.883), irrespective of the altitude (interaction, P = 0.127).

Cardiovascular CO₂ reactivity (Fig. 2)

Ascent to 5050 m elevated both MAP-CO₂ reactivity and HR-CO₂ reactivity (P = 0.003 and 0.006 vs. sea level, respectively). In contrast, these reactivities were unchanged following ACZ administration (P = 0.866 and 0.611), irrespective of the altitude (interaction, P = 0.105 and 0.475, respectively).

Isocapnic hypoxia

Ascent to 5050 m elevated the -O₂ sensitivity by (2420 ± 1521 vs. 1020 ± 908 units; P = 0.002 vs. sea level). No changes were observed in the -O₂ sensitivity with ACZ administration (P = 0.559 vs. control) at either sea level (981 ± 1152 vs. 1057 ± 563 units) or following ascent to 5050 m (2358 ± 1496 vs. 2482 ± 1616 units; interaction effect, P = 0.958, data not shown).

Discussion

We have examined the acute effects of a single i.v. dose of ACZ on cerebrovascular and ventilatory responsiveness to CO₂ and hypoxia in healthy resting individuals at sea level and following partial acclimatisation to 5050 m. The main novel findings were that ACZ at 5050 m: (1) elevated resting MCAv (to a greater extent than at sea level) and , while it lowered both and ; (2) improved breathing stability as reflected in a reduced variability of , V_T and incidence of breathing oscillations, while the ventilatory responsiveness to hypercapnia and hypoxia were unchanged; and (3) elevated both hypercapnic and hypocapnic cerebrovascular CO₂ reactivity, with the hypocapnia reactivity increases greater than observed at sea level. Our data thus indicate that ACZ administration at 5050 m improves breathing stability during wakefulness without concurrent improvements in arterial O₂ saturation or . Moreover, the improvement in breathing patterns following ACZ occurred independently of any measurable changes in the sensitivity of the central or peripheral chemoreflexes. We therefore speculate that -mediated elevations in cerebral perfusion and an enhanced cerebrovascular reactivity may partly account for the improved breathing stability following ACZ at high altitude

Limitations

An important limitation of the present study is the assumption that MCAv represents global CBF changes. Subudhi et al. (2011) pointed out that since CBF is markedly heterogenous during hypoxia, ACZ may have regional specific effects not detected with transcranial Doppler. ACZ administration (10 mg kg⁻¹, i.v.) in rats uniformly elevates CBF to the frontal and parietal cerebral cortex, the striatum, the hippocampus and the cerebellum, reaching a maximum at 1 h and progressively returning to baseline over 6 h (LaManna & McCracken, 1990). Likewise, ACZ (1 g) elevated CBF in numerous cortical grey matter areas (occipital lobe, superior frontal gyrus, cingulate gyrus, primary sensory-motor cortex, middle and superior temporal gyrus) and putamen and white matter, which was quantified using both MRI and PET measurements in healthy humans (Grandin et al. 2005). In addition, a MR imaging study by Schreiber et al. (2000) reported no changes in MCA diameter with ACZ administration in patients with internal carotid occlusion. Finally, a recent report showed a preserved MCA diameter at 5300 m compared with sea level (Wilson et al. 2011). Taken together, we contend that: (i) ACZ elevates CBF in all major brain regions; (ii) the ACZ-induced changes in MCAv reflect these global changes in CBF; and (iii) MCA diameter is unlikely to be altered at 5050 m.

We selected to use the i.v. route rather than oral administrations to examine the influence of the acute direct effect of ACZ on CBF without the major confounding influence of changes in acid–base balance following oral administration (Hauge et al. 1983; Vorstrup et al. 1989; see Swenson, 1998 for review). Nevertheless, upon injection of ACZ, an unexpected observation was the apparent changes in pH and at 5050 m (Table 1). Carbonic anhydrase inhibition causes disequilibrium of the CO₂ buffer system, which results from an incomplete hydration of CO₂ during the passage of circulating blood through the vascular beds. There is a progressive elevation of arterial CO₂ tension from the lungs to the peripheral tissues (Brzezinski et al. 1967). In the present study, there were several minutes of delay between the arterial blood sampling and analysis, thus allowing sufficient time for the equilibrium to complete. Therefore, it is likely that the true in vivo[H⁺] and values might be overestimated by our arterial blood gas measurements following ACZ. Nevertheless, Brzezinski et al. (1967) found a consistent relationship between with direct cerebral tissue CO₂ tension (∼6 mmHg difference) before and following ACZ in anaesthetised dogs. Accordingly, we believe that the arterial blood gas values obtained in the present study provide an insight to the differential effect of ACZ on acid–base balance at 5050 m. Our findings of a greater elevation in MCAv at 5050 m following ACZ is probably explained by the greater ACZ-induced elevations in (Table 1).

Acetazolamide elevates cerebral blood flow at high altitude

While the influence of ACZ on CBF has been studied extensively during normoxia (Posner & Plum, 1960; Ehrenreich et al. 1961; Hauge et al. 1983; Vorstrup et al. 1984), its effect on cerebrovascular function during hypoxic exposure is poorly understood and entirely based on oral administration of the drug. For example, studies have found either reduced (Subudhi et al. 2011) or preserved (Teppema et al. 2007) CBF velocity in response to short-duration hypoxia (4–8 h) following 1 and 3 day oral ingestion of ACZ, respectively (250 mg every 8 h). At 3475 m, Jensen et al. (1990) reported a 22% increase in CBF velocity 2 h following oral ACZ ingestion (1.5 g). Vuyk et al. (2006) found higher cerebral oxygenation during exercise at 3700 m and with ACZ (750 mg daily) compared with untreated controls. In the present study, acute ACZ administration at 5050 m increased resting MCAv by 28% and elevated the hypercapnic and hypocapnic MCAv-CO₂ reactivity by 51% and 69%, respectively (Table 1 and Fig. 1). Our data thereby demonstrate, for the first time, that acute i.v. ACZ administration elevates CBF and enhances the cerebrovascular response to changes in CO₂ in partially acclimatised newcomers to 5050 m. Such elevation in resting CBF with ACZ at 5050 m would also serve to improve cerebral oxygen delivery and thus potentially attenuate cerebral hypoxaemia.

Acetazolamide exacerbates hypoxaemia at 5050 m

Previous studies have found either unchanged (Teppema et al. 2006, 2007, 2010) or improved (Schoene et al. 1983; Burki et al. 1992; Mirrakhlmov et al. 1993; Subudhi et al. 2011) arterial O₂ saturation with oral ACZ – presumably related to the changes in resting associated with carbonic anhydrase inhibition. In the present study, we observed a slight reduction in both and following ACZ at both sea level and following ascent to 5050 m (Table 1). We attributed this increase in arterial hypoxaemia to hypoventilation (see below) and/or the Bohr effect, whereby ACZ-induced CO₂ retention causes less O₂ loading of the red blood cells in the lungs.

Carbonic anhydrase inhibition and breathing control

Despite the large body of literature (see Swenson, 1998 for review), much controversy still surrounds the effects of ACZ-induced carbonic anhydrase inhibition on ventilatory control. The effect of carbonic anhydrase inhibition on ventilatory control depends critically on the dose and the route of ACZ administration (Teppema et al. 2001). For example, oral administration of ACZ usually results in metabolic acidosis, which is frequently (but not always) associated with a rise in (Teppema & Dahan, 1999). Meanwhile, i.v. ACZ administration does not immediately lead to metabolic acidosis, but may be followed by inhibition of ventilatory control (Swenson, 1998).

One important factor of carbonic anhydrase inhibition on resting is the effect of elevated CBF and associated brain CO₂ washout. Hauge et al. (1983) proposed that, assuming constant cerebral CO₂ production, an increase in resting CBF and its responses to CO₂ associated with cerebral capillary carbonic anhydrase inhibition would lower brain tissue via an increase of CO₂ washout, thus attenuating central chemoreceptor activation. In support of this, they observed an initial depression of with carbonic anhydrase inhibition, which coincided with the ACZ-induced increase in resting CBF. In the present study, we also observed a reduction in resting following ACZ at both sea level and 5050 m (Table 1). Our finding contradicts previous reports of elevated resting following chronic oral ACZ administration (Swenson & Hughes, 1993; Teppema & Dahan, 1999; Teppema et al. 2007, 2010). Since we observed a ∼15% increase in resting MCAv at 30 min following acute ACZ administration at sea level, while arterial pH was lowered (Table 1), we speculate that the reduction in resting may be due to the effect of increased CBF and associated central H⁺ washout, overriding any stimulating effect of ACZ-induced CO₂ retention (Coates et al. 1991; Teppema et al. 1995, 2010). Accordingly, it appears that the ventilatory effect observed with ACZ may be closely linked with the changes in the control of CBF.

In the present study, both -CO₂ and -O₂ sensitivities remained unchanged with acute ACZ administration (10 mg kg⁻¹, i.v.) at both sea level and 5050 m (Fig. 2). However, in agreement with previous findings (Teppema et al. 1992, 2006; Swenson & Hughes, 1993; Teppema & Dahan, 2004), we did observe complete abolishment of the ventilatory response to hypoxia in some participants at both sea level (n = 3) and 5050 m (n = 1) with ACZ. In addition, there was a rightward shift of the -CO₂ slope during modified rebreathing with ACZ at 5050 m (Fig. 2). Our finding is in contrast to previous reports of a leftward shift of the -CO₂ slope with ACZ (Teppema & Dahan, 1999). However, since that study assessed the ventilatory CO₂ sensitivity following 3 days of oral ACZ ingestion (250 mg 8 h⁻¹), differences in the method of administration (oral vs. i.v.), drug dosage and assessment of ventilatory response (end-tidal forcing vs. rebreathing) probably account for these discrepant findings.

Acetazolamide and breathing instability

In the present study, we found acute ACZ administration at 5050 m improved breathing stability (Table 2), despite preserved -CO₂ and -O₂ sensitivities (Fig. 2). Our data indicate that this improved breathing stability is mediated by reductions in variability of the tidal breath volumes and the frequency of breathing oscillations (i.e. Cheyne–Stokes respiration, Table 2). Our findings provide support to the notion that ACZ ameliorates breathing pattern disturbances following ascent to high altitude (Weil et al. 1978; Sutton et al. 1979, 1980; Hackett et al. 1987), independent of changes in either peripheral (Hackett et al. 1987) or central chemoreflexes (Burki et al. 1992). Furthermore, our findings corroborate with those of Gotoh et al. (1969) who found that i.v. ACZ injection (500 mg) improved breathing stability by reducing the incidence of Cheyne–Stokes respiration cycles in patients with cerebrovascular disease. We partly attribute the improvement in breathing stability with ACZ to the elevations in CBF velocity and related elevations in cerebrovascular responsiveness to changes in CO₂ (Fig. 1). Indeed, it has been shown that reduction in resting CBF and cerebrovascular hypocapnic reactivity lowers the CO₂ reserve (the difference between eupnoeic and apnoea threshold) and increases the risk of unstable breathing during sleep (Xie et al. 2009). Xie and colleagues attributed this reduction in CO₂ reserve to an increased slope of ventilatory response to CO₂below eupnoea (i.e. ventilatory controller gain) associated with reduced cerebrovascular hypocapnic reactivity. As such, it seems reasonable that increases in CBF and cerebrovascular hypocapnic reactivity associated with ACZ would therefore blunt the ventilatory slope to CO₂ below eupnoea, thereby increasing the CO₂ reserve and improve breathing stability. Moreover, since hypoventilation lowers the CO₂ reserve (Dempsey, 2005), we speculate that the stabilising effect of ACZ must be greater than the destabilising effect of the reduced ventilation observed in the present study (Table 1). In support of this, ACZ has been found to increase the CO₂ reserve by lowering the apnoeic threshold in humans (Teppema et al. 2010) and in the anaesthetised cat models (Teppema et al. 2001). This improvement in the CO₂ reserve would account for the therapeutic effect of ACZ on breathing stability at high altitude. Together with data from the present study, these findings indicate that ACZ-induced changes in cerebrovascular function serve to modulate ventilatory controller gain below eupnoea, and thus increase the CO₂ reserve and improve breathing stability at 5050 m. However, we cannot exclude the possibility that ACZ may improve breathing stability via alterations in the peripheral chemoreceptor response, which was not detected in the present study, since carbonic anhydrase inhibition is known to (i) increase carotid body sinus activity delay (Lahiri et al. 1982; Iturriaga et al. 1991), (ii) lower carotid body activity in response to steady state CO₂in vivo (Hayes et al. 1976; Lahiri et al. 1976), and (iii) attenuate the speed of response to CO₂in vitro (Iturriaga et al. 1991). Nevertheless, the increased cerebrovascular hypocapnic reactivity with ACZ may also account for the reduction in the incidence of central sleep apnoea in patients with congestive heart failure at sea level (Javaheri, 2006; Fontana et al. 2011). Indeed, Xie et al. (2005) previously reported a lower overall cerebrovascular CO₂ reactivity, especially to hypocapnia, in congestive heart failure patients with central sleep apnoea compared with patients without. Collectively, along with the related elevations in , it appears that changes in the hypocapnic cerebrovascular CO₂ reactivity may alter the CO₂ reserve, thus influencing breathing stability during both wakefulness and sleep. The potential implications for the role of altered cerebrovascular reactivity in the pathogenesis of breathing instability certainly warrant further investigation.

Conclusions

To our knowledge, the present study is the first to examine the effects of a single i.v. dose of ACZ on cerebrovascular function and breathing control in partially acclimatised newcomers to high altitude. We demonstrated that acute ACZ at 5050 m increased resting CBF (probably via elevations in ) and enhanced the CBF responsiveness to both hypercapnia and hypocapnia, and improved breathing stability. We speculate that such elevations in CBF and cerebrovascular reactivity may account, in part, for the well-documented reductions in periodic breathing during sleep following oral ACZ ingestion at high altitude.

Glossary

Abbreviations

ACZ

acetazolamide

AMS

acute mountain sickness

CBF

cerebral blood flow

CVCi

cerebrovascular conductance index

MAP

mean arterial blood pressure

MCAv

middle cerebral artery blood flow velocity

Author contributions

P.N.A. and K.R.B. contributed to the conception and design of the experiment, the interpretation of the data and the writing of the manuscript. J.-L.F. carried out data collection, and led the analysis, interpretation and writing of the manuscript. K.C.P., K.N.T. and S.J.E.L. contributed equally in the study design and collection of data. J.D.C. and B.K. contributed to the data interpretation and manuscript preparation. All authors approved the final version of this manuscript. The sea-level work was carried out in the Human Physiology Laboratory, Department of Physiology, University of Otago, while the experiments at high altitude were carried out at the Pyramid Laboratory, Italian Research Council.