Reply from J.-L. Fan, K. R. Burgess and P. N. Ainslie

We thank Drs Teppema and Berendsen for their interest in our recent paper (Fan et al. 2012). If we interpret their letter correctly, they have raised three noteworthy discussion points. First, why greater elevations in middle cerebral artery blood flow velocity (MCAv) and ventilatory CO₂ sensitivities were evident during initial arrival to 5050 m (Fan et al. 2010) compared to those measured 5–12 days at this elevation. Second, they expressed concerns over curve fitting during isocapnic hypoxic rebreathing for the comparison of ventilatory O₂ sensitivity between sea level and following ascent to 5050 m. Finally, they expressed skepticism over our use of intravenous acetazolamide to examine its effect on breathing control. Although we feel many of these points are discussed in our paper (Fan et al. 2012), the salient points are outlined below.

1. The differences from during initial arrival to 5050 m (Fan et al. 2010) to those measured 5–12 days at this elevation (Fan et al. 2012) probably reflect partial acclimatization over this time frame. In our more recent study, we observed normalized pH and MCAv to sea level values, whereas our earlier report found these variables to be significantly elevated during initial arrival to 5050 m. Moreover, we found the changes in resting ventilation and with ascent to 5050 m to be greater following 5–12 days compared with initial arrival. This difference in the state of acclimatization would certainly account for the differences in the CO₂ sensitivities between our two papers. Another possibility for these apparent discrepant findings may be due to the difference in sample sizes between the two studies. In our more recent report, only 11 subjects were included in the final statistical analysis, resulting in a tendency for the ventilatory CO₂ sensitivity to be elevated at 5050 m (P= 0.085). In contrast, our earlier study reported ventilatory CO₂ sensitivity in 17 subjects. Therefore, we believe that the lack of change in ventilatory CO₂ sensitivity observed in Fan et al. (2012) could be attributed to insufficient statistical power to detect the changes in central chemoreflex following ascent to high altitude. Finally, the transition we used from room air at both altitudes to the end hyperventilation state resulted in a similar degree of mild hyperoxic exposure at both altitudes.

2. We agree that testing peripheral chemoreflex sensitivity is extremely difficult, especially in the field, and additional O₂ may be needed inside the rebreathing bag to ensure similar ranges. However, since the ranges assessed pre- and post-acetazolamide were similar at each altitude, we believe that the soda lime rebreathing method was sufficient to identify any potential changes in peripheral chemoreflex associated with our intervention. Moreover, we argue that the ranges assessed were physiologically more relevant at each altitude. We acknowledge that these points raised by Teppema and Berendsen are important considerations when assessing changes in hypoxic ventilatory sensitivities at high altitude. Nevertheless, acute gas control using end-tidal forcing is technologically challenging, especially in real field conditions as opposed to the safe confines of the laboratory (Teppema & Dahan, 2010). Also, although the authors make some interesting points about chemoreflex testing, based on their elegant review paper (Teppema & Dahan, 2010), comparisons with high altitude is almost impossible because of marked acid–base changes and the interactions between the peripheral and central chemoreflexes. Although we would suggest that no approach is perfect, we attempted to target end-tidal CO₂ levels for the rebreathing at high altitude so the degree of CO₂ washout of all tissues would be similar at the two altitudes.

3. As discussed in our paper, Teppema and Berendsen mention that metabolic acidosis associated with chronic oral acetazolamide ingestion leads to a rise in ventilation and drop in , which independently stabilizes breathing – presumably via an increased plant gain (Dempsey, 2005). However, since the majority of the work on acetazolamide has looked at oral ingestions, our reasoning for using acute intravenous administration was to examine whether acetazolamide may alter ventilatory control and breathing stability by mechanisms other than metabolic acidosis. Oral and i.v. acetazolamide adminstration is clearly physiologically very different and no similarities can be drawn between the two.

In summary, whilst thanking Drs Teppema and Berendsen for their interest, we feel their comments do not detract from our salient findings. We do agree, however, that the cerebrovascular response to CO₂ may indeed be initiated at the arterial side of the blood–brain barrier.