Does CPAP Actually Reduce Cardiovascular Risk?
Trials have been mixed. A 2026 analysis found the benefit concentrated in high-risk sleep apnea — and reported signs of harm in low-risk patients, which is the finding nobody expected.
This is the question that should determine how hard anyone pushes CPAP on a patient without daytime symptoms, and for a decade the honest answer has been "we are not sure."
The observational evidence is strong. Untreated obstructive sleep apnea is consistently associated with hypertension, cardiovascular events and increased mortality. But randomised trials of CPAP with cardiovascular endpoints have repeatedly failed to show the benefit those associations implied.
A 2026 analysis offers the most convincing explanation yet, and it comes with an uncomfortable finding attached.
What the analysis found
The study examined CPAP's effect on major adverse cardiovascular and cerebrovascular events (MACCE), split by whether patients had high-risk obstructive sleep apnea (European Heart Journal, 2026).
Overall, there was nothing. Among 3,549 participants, 16.6% reached the MACCE endpoint on CPAP (n = 1,778) against 16.3% on usual care (n = 1,771) [Eur Heart J 2026]. Those numbers are indistinguishable, and on their own they read as another null CPAP trial.
Split by risk, they separate. The CPAP treatment effect was greater in participants with high-risk apnea than without — interaction hazard ratio 0.69 (95% CI, 0.50–0.95; P for interaction = .024), in 1,832 high-risk patients [Eur Heart J 2026].
The effect was stronger in the groups you would least expect:
- In those without excessive sleepiness: iHR 0.59 (95% CI, 0.41–0.84), n = 1,509 [Eur Heart J 2026]
- In those without increased blood pressure: iHR 0.54 (95% CI, 0.36–0.81), n = 1,244 [Eur Heart J 2026]
That is counterintuitive and important. Sleepiness is the symptom that usually drives treatment, and here the cardiovascular benefit was larger in people who did not have it.
The finding that deserves attention
The authors do not bury it: "Continuous positive airway pressure benefits in high-risk OSA were observed alongside harm in low-risk OSA" [Eur Heart J 2026].
Their conclusion is that CPAP preferentially improves cardiovascular outcomes in high-risk apnea, while harm in low-risk apnea may counteract this effect [Eur Heart J 2026].
Read carefully, that says two things at once:
- The reason overall trials looked null may be that real benefit in one group was being cancelled out by real harm in another.
- Treating low-risk apnea for cardiovascular reasons may not be neutral.
Both warrant caution rather than alarm. This is a subgroup analysis with an interaction test, not a trial designed to detect harm in low-risk patients, and a p-interaction of .024 in a study of this size is suggestive rather than definitive. The mechanism of any harm is not established by these data.
But it is the first serious account of why a decade of CPAP cardiovascular trials disappointed, and it points at patient selection rather than at CPAP not working.
Tier 2: a strong, plausible signal from a large dataset, not yet a settled finding. How we grade evidence.
Why selection would explain the earlier trials
If cardiovascular benefit is concentrated in a subgroup, enrolling by a metric that does not identify that subgroup would dilute it toward zero.
Which is exactly what most trials did — they enrolled by apnoea-hypopnoea index. And the AHI is a poor predictor of cardiovascular mortality: in two cohorts totalling nearly 8,000 people, hypoxic burden strongly predicted cardiovascular mortality while the AHI did not (Azarbarzin et al., European Heart Journal, 2019).
Recruit on a number that does not track the risk you are trying to reduce, and you fill a trial with a mixture of people who stand to benefit and people who do not. The average effect goes to nothing while both real effects exist underneath. What hypoxic burden measures.
That is a coherent story rather than a proven one, and it makes better sense of the evidence than "CPAP does not work."
What this does not mean
It is not a reason to stop using CPAP. Nothing here bears on the uses CPAP is best supported for — relieving daytime sleepiness, improving quality of life and function, and treating the apnoea itself. Those benefits are established and are why most people are prescribed it.
It is not a reason to refuse treatment. If you have moderate-to-severe apnoea with symptoms, the case for treating is intact and does not rest on cardiovascular prevention.
It is not permission to self-assess your risk. "High-risk" here is a statistical classification within a study, not something to judge about yourself from an article. Adherence decisions belong with your clinician — and for commercial drivers, with a medical examiner. Sleep apnea and your DOT physical.
What it does mean
The reason to treat matters. Treating severe apnoea with deep desaturations in someone with cardiovascular disease rests on different evidence from treating mild, asymptomatic apnoea to prevent a future heart attack. The first is well supported. The second is where this analysis urges caution.
Symptoms remain the clearest indication. If apnoea is making you tired, treating it is worth doing on its own terms and the benefit is measurable in how you feel. Getting used to CPAP · CPAP versus an oral appliance.
Ask what your own risk profile looks like. Oxygen nadir, time below 90%, desaturation index and comorbidities describe your risk far better than a severity label does. How to read your sleep study.
And the observational evidence still stands. Untreated apnoea's association with hypertension, cardiovascular events and mortality has not been overturned by any of this. What is uncertain is how much of that is reversible by CPAP, and in whom. What untreated apnea does · Blood pressure.
The line
No consumer product treats obstructive sleep apnea or affects cardiovascular risk from it. This page is about how well a real medical treatment works in which patients — not a reason to substitute anything for it. Snoring versus sleep apnea.