Breath holding as an example of extreme hypoventilation: experimental testing of a new model describing alveolar gas pathways

Exp Physiol. 2020 Dec;105(12):2216-2225. doi: 10.1113/EP088977. Epub 2020 Oct 15.


New findings: What is the central question of this study? We modelled the alveolar pathway during breath holding on the hypothesis that it follows a hypoventilation loop on the O2 -CO2 diagram. What is the main finding and its importance? Validation of the model was possible within the range of alveolar gas compositions compatible with consciousness. Within this range, the experimental data were compatible with the proposed model. The model and its characteristics might allow predictions of alveolar gas composition whenever the alveolar ventilation goes to zero; for example, static and dynamic breath holding at the surface or during ventilation/intubation failure in anaesthesia.

Abstract: According to the hypothesis that alveolar partial pressures of O2 and CO2 during breath holding (BH) should vary following a hypoventilation loop, we modelled the alveolar gas pathways during BH on the O2 -CO2 diagram and tested it experimentally during ambient air and pure oxygen breathing. In air, the model was constructed using the inspired and alveolar partial pressures of O2 ( PIO2 and PAO2 , respectively) and CO2 ( PICO2 and PACO2 , respectively) and the steady-state values of the pre-BH respiratory exchange ratio (RER). In pure oxygen, the model respected the constraint of PACO2=-PAO2+PIO2 . To test this, 12 subjects performed several BHs of increasing duration and one maximal BH at rest and during exercise (30 W cycling supine), while breathing air or pure oxygen. We measured gas flows, PAO2 and PACO2 before and at the end of all BHs. Measured data were fitted through the model. In air, PIO2 = 150 ± 1 mmHg and PICO2 = 0.3 ± 0.0 mmHg, both at rest and at 30 W. Before BH, steady-state RER was 0.83 ± 0.16 at rest and 0.77 ± 0.14 at 30 W; PAO2 = 107 ± 7 mmHg at rest and 102 ± 8 mmHg at 30 W; and PACO2 = 36 ± 4 mmHg at rest and 38 ± 3 mmHg at 30 W. By model fitting, we computed the RER during the early phase of BH: 0.10 [95% confidence interval (95% CI) = 0.08-0.12] at rest and 0.13 (95% CI = 0.11-0.15) at 30 W. In oxygen, model fitting provided PIO2 : 692 (95% CI = 688-696) mmHg at rest and 693 (95% CI = 689-698) mmHg at 30 W. The experimental data are compatible with the proposed model, within its physiological range.

Keywords: O2-CO2 diagram; alveolar gas; apnoea; breath holding; carbon dioxide; oxygen.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adult
  • Breath Holding
  • Carbon Dioxide / metabolism
  • Exercise / physiology
  • Female
  • Humans
  • Hypoventilation / metabolism
  • Hypoventilation / physiopathology*
  • Lung / metabolism
  • Lung / physiology*
  • Male
  • Oxygen / metabolism
  • Oxygen Consumption / physiology
  • Partial Pressure
  • Pulmonary Gas Exchange / physiology*
  • Respiration
  • Rest


  • Carbon Dioxide
  • Oxygen