In a recent issue of The Journal of Physiology, Fan and colleagues (Fan et al. 2012) studied the effect of intravenous acetazolamide (AZ) on cerebrovascular and ventilatory O2 and CO2 sensitivity at sea level and after acclimatization to 5050 m.

In a previous study (Fan et al. 2010), resting middle cerebral artery blood flow velocity values were reported that were higher at 5050 m than at sea level by about 30%, but in their recent paper there was no significant difference although blood gases and stage of acclimatization were similar. Likewise, in the previous study hyperoxic ventilatory CO2 sensitivity rose by ∼ 42% at 5050 m, but in their last study there was no significant difference with sea level. What causes these variable results? May modified rebreathing perhaps introduce variations from unknown sources? One of the features of Duffin's modified rebreathing is the initial hyperventilation lasting 5 min. At sea level this resulted in a mean inline image and inline image of 22 and 135 mmHg, respectively; at 5050 m these values were 17 and 63 mmHg (see Fan et al. 2010). In other words, in contrast to sea level, modified rebreathing at high altitude started with a sudden transient from a hypoxic to a hyperoxic condition. Despite the low inline image, does this lead to a sudden reduction in carotid body output (and central chemoreceptor output if there is a crosstalk between them as recently suggested)? Note that in acclimatized subjects carotid body sensitivity is greatly enhanced (references in Teppema & Dahan, 2010) raising the question if hyperventilation down to a inline image of 17 mmHg simply eliminates carotid body activity in these ‘sensitized’ subjects.

During the course of modified rebreathing both the inline image and ventilation rise. A crucial assumption underlying the Duffin type of rebreathing is that, due to the hyperventilation, the inline image in all tissues decreases, which is followed by an equilibration of the arterial and central chemoreceptor inline image during the rebreathing. Brain tissue inline image (inline image) is closely linked to inline image, but if cerebral blood flow (CBF) reaches saturation at a inline image of 40–45 mmHg (a consistent observation by Battisti-Charbonney et al. 2011), the slope of the inline imageinline image relationship must change because the further rise in inline image is now solely dictated by the tissue CO2 buffering capacity. Does this lead to an altered ventilation–inline image relationship?

Fan and colleagues analysed the isocapnic hypoxic ventilatory response with a first-order polynomial function with the curvature representing hypoxic sensitivity and the hyperoxic ventilation as the y asymptote (Fan et al. 2012).The subjects rebreathed from a 6 l bag in which previously exhaled ambient air at rest was collected. So at 5050 m the subjects started to inhale air that contained much less oxygen than at sea level (mean resting inline image values were 46 and 97 mmHg, respectively). The non-linear relationship between ventilation and inline image predicts that curve fitting on data within a lower inline image range will yield a greater hypoxic sensitivity, so the reported higher hypoxic sensitivity at 5050 m is not surprising. Hypoxic sensitivities between treatments cannot be compared unless data are sampled within the same inline image range. In addition, how can curve fitting be applied without isocapnic hyperoxic ventilation data?

To examine the influence of the acute direct effect of acetazolamide (AZ) on CBF without the ‘major confounding effects of changes in acid–base balance following oral administration’ the authors used the i.v. route. This is remarkable since chronic (oral) AZ is known to decrease the loop gain and stabilise the respiratory control system by lowering the plant gain without changing ventilatory CO2 or O2 sensitivity (Kiwull-Schone et al. 2006; Teppema & Dahan, 1999; Teppema et al. 2010; Edwards et al. 2012). The AZ-induced metabolic acidosis leads to a rise in ventilation and a decrease in inline image, which has an independent stabilising influence on ventilatory control (Manisty et al. 2006). The metabolic acidosis induced by oral AZ is thought to play a crucial role in the prevention of acute mountain sickness and improvement of periodic breathing during sleep at high altitude (Swenson et al. 1991). Another disadvantage of a high intravenous dose of AZ is the ensuing chemical disequilibrium in vivo, as discussed by the authors. This leads to a smaller change in inline image induced by a given change in inline image and a reduced ventilatory CO2 sensitivity (Teppema et al. 1995) but, according to the data of Fan and colleagues (Fan et al. 2012), not to a diminished cerebrovascular CO2 sensitivity, indicating that the vascular response to CO2 may indeed be initiated at the arterial side of the blood–brain barrier.


  1. Top of page
  2. References