Can You Use Oxygen Instead of CPAP?
No — and the trials that settled it are more interesting than a flat no. Oxygen fixes the low oxygen and leaves the obstruction, which turns out to tell you something about which part causes the harm.
It is a reasonable question. If the damage from sleep apnea comes from your oxygen dropping, why not just supply more oxygen and skip the mask?
The answer is no, and the reason it is no gets to the heart of what sleep apnea actually is.
What oxygen does not fix
Obstructive sleep apnea is a mechanical problem. Your airway closes. Oxygen delivered through a nasal cannula sits on the wrong side of a closed door — it raises the oxygen content of air you cannot draw in.
So the obstruction continues. The arousals continue. The effort of breathing against a shut airway continues. What changes is that when you do manage a breath, it carries more oxygen, so the desaturations are blunted.
That is not nothing, as the evidence below shows. It is also not treatment.
The trial that answered it directly
HeartBEAT randomised 318 patients with cardiovascular disease or multiple cardiovascular risk factors, with 281 (88%) evaluable for ambulatory blood pressure at both baseline and follow-up (Gottlieb et al., New England Journal of Medicine, 2014).
At 12 weeks, 24-hour mean arterial pressure was lower with CPAP than with control (−2.4 mmHg; 95% CI, −4.7 to −0.1; P = 0.04) and lower with CPAP than with supplemental oxygen (−2.8 mmHg; 95% CI, −5.1 to −0.5; P = 0.02) [Gottlieb 2014].
And the decisive comparison: there was no significant difference in 24-hour mean arterial pressure between the control group and the oxygen group [Gottlieb 2014].
The authors' conclusion is unambiguous — in patients with cardiovascular disease or multiple risk factors, treating obstructive sleep apnea with CPAP, but not with nocturnal supplemental oxygen, produced a significant reduction in blood pressure [Gottlieb 2014].
Oxygen performed like no treatment at all on that endpoint. Tier 1. How we grade evidence.
The trial that makes it interesting
A flat "oxygen does nothing" would be the wrong lesson, and a second trial shows why.
The SOX study took patients established on CPAP and withdrew it, with or without supplemental oxygen (Turnbull et al., American Journal of Respiratory and Critical Care Medicine, 2019).
Oxygen virtually abolished the blood pressure rise that normally follows CPAP withdrawal. Against air, it reduced the rise in mean systolic blood pressure by 6.6 mmHg (95% CI, −11.3 to −1.9; P = 0.008), diastolic by 4.6 mmHg (95% CI, −7.8 to −1.5; P = 0.006), and median oxygen desaturation index by 23.8 per hour (IQR, −31.0 to −16.3; P < 0.001) [Turnbull 2019].
But it changed nothing about the apnea itself. There was no significant difference between oxygen and air in AHI, subjective sleepiness or objective sleepiness [Turnbull 2019].
So oxygen did exactly what the mechanism predicts: it fixed the hypoxia and left the obstruction. The events kept happening at the same rate, and patients were just as sleepy.
What that combination reveals
The authors draw a genuinely important inference: because oxygen substantially reduced intermittent hypoxia while having minimal effect on arousal markers and sleepiness, intermittent hypoxia — not recurrent arousals — appears to be the dominant cause of daytime blood pressure increases in obstructive sleep apnea [Turnbull 2019].
That is a finding about which part of apnea does which damage. The oxygen drops drive the blood pressure. The arousals drive the sleepiness. They are separable, and they respond to different things.
It also explains why the two trials are not in conflict. HeartBEAT asked whether oxygen lowers blood pressure in people with apnea; it does not. SOX asked whether oxygen blunts the blood pressure rise when effective treatment is removed; it does. Different questions about different baselines.
This is the same logic that makes hypoxic burden a better predictor of cardiovascular mortality than event counts. What hypoxic burden measures.
Where oxygen does have a role
Not as a substitute, but it is prescribed in sleep medicine for real reasons:
Alongside positive airway pressure, where a patient remains hypoxaemic despite the airway being held open — usually because of coexisting lung disease.
In overlap syndrome, where COPD and sleep apnea occur together and the oxygenation problem is not only obstructive.
In some central sleep apnea and hypoventilation syndromes, which are different conditions from obstructive apnea.
In each case it is added to treatment by a clinician who has established why oxygen is needed. That is a different thing from swapping CPAP for a concentrator.
The specific risk of using it as a substitute
Worth stating plainly, because this is the part that makes it more than merely ineffective.
It can hide the problem while leaving it in place. Oxygen improves the number most often monitored at home — saturation — while the obstruction, the arousals and the effort continue. A reassuring pulse oximeter reading in someone with untreated apnea is a misleading reading.
In some people, oxygen can prolong events or blunt respiratory drive, which is why it is not something to self-prescribe for a breathing disorder during sleep.
And it is not a legitimate route for anyone whose treatment is documented. For commercial drivers in particular, adherence data from a positive airway pressure device is what demonstrates treatment. Oxygen produces nothing equivalent. Sleep apnea and your DOT physical.
If the real problem is that you cannot tolerate CPAP
That is the honest reason most people ask this, and it has better answers.
Most abandonment traces to fixable equipment problems — mask fit, pressure ramp, mouth leak, dry air — rather than to CPAP being intolerable. A mask refit is routine and is what most people who quit actually needed. Getting used to CPAP · Mouth leak.
If you have genuinely tried it properly, the real alternatives are a custom oral appliance, positional therapy where your study shows you are positional, weight treatment, nerve stimulation or surgery in selected cases. Those act on the obstruction. Oxygen does not. CPAP alternatives · CPAP versus an oral appliance.