Defining the ventilatory response to hypoxia

In humans, the ventilatory response to hypoxia is generated at the carotid bodies and is a life-saving chemoreflex aimed at restoration of oxygen homeostasis under circumstances of reduced oxygen delivery to the brain. The response is context sensitive and displays time dependency. For example, the response magnitude is dependent on the inhaled (and thus arterial) carbon dioxide concentration, background pH and the presence of respiratory depressant agents in the body.

The response to isocapnic hypoxia (e.g. at constant end-tidal PCO2) is larger than the response to poikilocapnic hypoxia, where end-tidal CO2 is allowed to drop upon the hypoxia-induced hyperventilation.

Furthermore upon induction of hypoxia an initial hyperventilatory response (approx. 200% of baseline) is followed within 5 min by a slow decline, the hypoxic ventilatory decline. A new steady state is reached after 15 to 20 min, about 25% to 50% above resting ventilation levels.

The ventilatory response to acute hypoxia is generally calculated as the difference in ventilation between normoxia and hypoxia divided by the change in oxygen saturation as derived from pulse-oximetry at the finger tip (SpO2). Recently we observe large differences between the SpO2 and the Hb-O2 saturation measured from the arterial blood. This indicates that prior assessment of the hypoxic ventilatory sensitivity in terms of ÄVi/ÄSpO2 is severely overestimated (Vi being the measured inhaled ventilation).

Apart from the above mentioned methodological issues, there is currently a discussion in the literature (cf. Duffin 2007) on how to measure the ventilatory response to hypoxia in light of possible physiological or pharmacological changes, such as chronic hypoxia (e.g., from residence at altitude), chronic acid-base disturbances, infusion of opioids, etc. It has been argued that measurements at fixed end-tidal PCO2 values above resting as is commonly employed may not give a reliable estimate of the hypoxic sensitivity. Duffin (2007) proposes either to apply 3-min hypoxic steps (hypoxic level 6 kPa) from hyperoxia at different background of hypercapnia or to apply a modified CO2 rebreathing experiment at two levels of oxygen concentration (hypoxia and hyperoxia).

The current protocol is aimed at:

(1) assessing the ventilatory response to hypoxia while measuring Hb-O2 saturation and PO2 from the arterial blood;

(2) Comparing the ventilatory responses to hypoxia obtained using

(i) a 20-min hypoxia step input function at 3 end-tidal CO2 levels,

(b) a 3-min hypoxia step input function at three different CO2 levels and

(c) the modified rebreathing test at two background levels of oxygen.

Experiments will be performed in healthy volunteers. The study will be purely observational. The data analysis will be done in three steps.
I. Hypoxic sensitivity will be expressed as ÄVi/ÄSpO2 and ÄVi/ÄSaO2.
II. We will construct Vi-CO2 response curves at hyperoxia and hypoxia.
III. We will apply a mathematical model of the ventilatory response to hypoxia in order to get an indication of the difference between gains using end-tidal and arterial PO2 values.

In this study we will compare three method to assess the magnitude of the ventilatory response; our hypothesis is that different methods will yield different results.

All experiments will be performed during a single 3-h session.

Our primary outcome is the ventilatory response to hypoxia. To that end, hypoxia will be induced by lowering the inspired oxygen concentration using a computer-controlled dynamic end-tidal forcing system, such that the oxygen saturation will 80 ± 2%. The hypoxic sensitivity will be calculated as delta(Ventilation)/delta(desaturation) in L/min per % desaturation and as delta(Ventilation)/delta(logPaO2), where PaO2 is the arterial oxygen saturation.

We will apply three different hypoxic stimuli:

1. A 25-min hypoxic step at three end-tidal CO2 levels (total duration is 75 min),

2. A 5-min hypoxic step preceded by 5 min hyperoxia and 5 min normoxia at 3 end-tidal CO2 levels (total duration is 45 min);

3. A 10-min ramp decrease in oxygen concentration at three end-tidal CO2 levels (total duration 30 min).



Continuous measurements will be: oxygen saturation, ventilation (on a breath-to-breath basis) and arterial PO2 (arterial blood samples will be obtained prior to induction of hypoxia and at the end of hypoxia)

Each subject will participate in all three hypoxic studies on one single session. Total duration of the session (incl. rest in between studies) is 4 hours. Comparisons will be performed on two levels:



A. a comparison of the ventilatory responses calculated from Saturation (Vi/SAT) versus arterial PO2 (Vi/logPaO2)

B. a comparison of the hypoxic sensitivities (both Vi/SAT and VI/logPaO2) among the three different hypoxic studies.