The basic idea is that people with pulmonary diseases that involve chronic hypoventilation rely on mild hypoxia to stimulate respiration.
To understand this, consider a basic homeostatic feedback loop that controls respiratory drive. During a breath hold, carbon dioxide levels rise and oxygen levels fall. Carbon dioxide diffuses across the blood-brain barrier and causes a decrease in CSF pH sensed at central chemoreceptors in the brainstem. The peripheral serum pH may also be depressed, stimulating peripheral chemoreceptors in the carotid and aortic bodies. In normally ventilating people, these "hypercarbic indicators" are the primary mediators of the drive to breath.
In patients with chronic hypoventilation, PaCO2 is chronically elevated. Common hypo-ventilatory diseases include:
- COPD (chronic obstructive pulmonary disease)
- obesity-hypoventilation syndrome
- neuro-muscular weakness syndromes (generally only in later phases when the diaphragm is affected: Becker/Duchenne muscular dystrophy, amyotrophic lateral sclerosis, etc.)
The old theory goes that because these folks have chronically elevated PaCO2*, the hypercarbic respiratory feedback loop described above “acclimates” and is no longer the primary driver of breathing. Instead, they rely on the hypoxic respiratory drive, which is secondary in normal people. These individuals may have a usual SpO2 ~88-90%. When a non-rebreather mask is applied and suddenly drives their SpO2 up to 100%, their respiratory drive decreases because their blunted hypercarbic response mechanisms do not respond normally to the progressive rise in PaCO2.
There are data in support of this theory. One study in COPD patients administered 100% oxygen to patients with COPD and observed a 18% decrease in ventilation after 15 minutes that returned to baseline when oxygen was removed (Aubier). However, these authors also showed that two other factors were at play explaining the observed rise in PaCO2:
About 30% of the increase in PaCO2 associated with high-flow oxygen administration was attributable to the Haldane effect, which is basically an offloading of CO2 by hemoglobin (due to a rightward shift of the carboxyhemoglobin dissociation curve), increasing the CO2 dissolved in blood but not reflecting a true change in ventilation.
An additional 48% of the increase in hypercapnia was due to dead space ventilation. The simplified idea is that the vessels in the lung compensate for hypoxia by vaso-constricting areas with low oxygen tension, so-called ventilation-perfusing matching. When high-flow oxygen is administered, the vaso-constriction is released so perfusion of poorly ventilated areas increases. Now a higher percentage of the blood is not being cleared of CO2. This is called ventilation-perfusion mismatching.
Something similar has been demonstrated in patients with obesity-hypoventilation (Wijesinghe). In light of this, a recent study investigated the best strategy for oxygen administration in patients with chronic hypoventilation (Kim). They found that a strategy titrating oxygen administration to achieve SpO2 88% to 92% minimized the risk of worsening hypoventilation while achieving adequate relief of shortness of breath and adequate oxygenation.
A helpful review of these topics provides a bit more detail for those interested (Abdo).
Abdo WF and Heunks LM. Oxygen-induced hypercapnia in COPD: myths and facts. Crit Care. 2012;16(5):323.
Aubier M, Murciano D, Milic-Emili J, Touaty E, Daghfous J, Pariente R, Derenne JP. Effects of the administration of O2 on ventilation and blood gases in patients with chronic obstructive pulmonary disease during acute respiratory failure. Am Rev Respir Dis. 1980 Nov;122(5):747-54.
Kim V, Benditt JO, Wise RA, Sharafkhaneh A.Oxygen therapy in chronic obstructive pulmonary disease. Proc Am Thorac Soc. 2008 May 1;5(4):513-8.
Wijesinghe M, Williams M, Perrin K, Weatherall M, Beasley R. The effect of supplemental oxygen on hypercapnia in subjects with obesity-associated hypoventilation: a randomized, crossover, clinical study. Chest. 2011 May;139(5):1018-24. .
*normal PaCO2 = 40 mmHg; may be 50-80 mmHg in chronic retention, possibly higher