Does CPAP Work for Central Sleep Apnea?

CPAP partially works for central sleep apnea, but it is far less reliably effective than it is for the more common obstructive type. In heart failure patients, the largest randomized trial showed that CPAP reduced central apneas, improved oxygen levels, and boosted heart pumping function, yet it did not extend life. The effectiveness depends heavily on what is causing the central apneas in the first place. For some causes, CPAP is a reasonable first-line treatment; for others, it barely makes a dent.

Why Central Sleep Apnea Is a Different Problem

In obstructive sleep apnea, the airway physically collapses and CPAP holds it open with a steady stream of pressurized air. Central sleep apnea is a fundamentally different problem. The airway is not blocked. Instead, the brain intermittently stops sending the signal to breathe. That distinction matters because CPAP was designed to splint an airway open, not to fix a signaling problem in the brain’s respiratory control center.

The underlying mechanism in most non-hypercapnic central sleep apnea involves something called ventilatory instability. When you fall asleep, your body’s sensitivity to carbon dioxide shifts. A small drop in CO2 of just a few points below baseline can cross a threshold that triggers the brain to pause breathing entirely. The resulting apnea causes CO2 to build back up, which then triggers an exaggerated breathing response, which overshoots and drives CO2 too low again, restarting the cycle. Researchers describe this oscillation using the concept of “loop gain,” which is essentially how aggressively the respiratory control system overreacts to changes in blood gases. When loop gain is high, the system overshoots and undershoots repeatedly, producing the characteristic waxing-and-waning breathing pattern seen in central sleep apnea.¹²

CPAP can help stabilize this cycle to some degree by increasing the volume of air in the lungs, which buffers CO2 fluctuations and provides a steadier baseline for gas exchange. But it does not directly address the brain’s oversensitive respiratory controller. This is why CPAP’s track record with central sleep apnea is mixed: it helps in some patients and some subtypes, fails in others, and occasionally makes things worse.

The Landmark Heart Failure Trial

The most important piece of evidence on CPAP for central sleep apnea comes from a large Canadian trial (CANPAP) that enrolled heart failure patients with central sleep apnea. The results were telling: CPAP reduced central apnea events, improved overnight oxygen levels, raised the heart’s ejection fraction, lowered stress hormones, and improved exercise capacity. By almost every intermediate measure, CPAP was doing something beneficial. But the trial’s primary endpoint was survival, and CPAP did not improve it. The researchers concluded that their data did not support using CPAP to extend life in heart failure patients with central sleep apnea.³

A later network meta-analysis looking across multiple treatment types found that both CPAP and adaptive servo-ventilation improved left ventricular ejection fraction by roughly 4% compared to no treatment, and that the benefit was larger when heart function was more severely impaired at baseline.⁴ So CPAP does measurably help the heart pump better, which matters clinically. The disconnect between improved cardiac function and unchanged survival has puzzled researchers for years and remains unresolved.

When CPAP Is More or Less Likely to Work

Not all central sleep apnea is created equal, and CPAP’s chances of success depend substantially on the cause. A single-center study looking at predictors of CPAP response found that patients whose breathing events were less than half central apneas were considerably more likely to respond well. By contrast, opioid-related central apneas and current smoking both predicted a poor response. Interestingly, older age made doctors less likely to prescribe CPAP initially, but age itself did not predict whether the treatment would actually work once started.⁵

The practical takeaway is that CPAP tends to do better when central apneas are part of a mixed picture with some obstructive component. When central events dominate the breathing pattern completely, CPAP’s ability to control them drops off. And when opioids are involved, CPAP’s performance is particularly poor.

Opioid-Induced Central Sleep Apnea Is Resistant to CPAP

Chronic opioid use is a well-known cause of central sleep apnea, and it is one of the subtypes where CPAP performs worst. A study that put opioid-dependent patients through two rounds of CPAP titration, spaced four weeks apart, found that central apnea events remained essentially unchanged on both nights. On the first CPAP night, the central apnea index was 20 events per hour; on the second, it was 19. The researchers abandoned CPAP therapy and concluded that opioid-related central apnea is resistant to it.⁶ A systematic review of positive airway pressure treatments for opioid-related central apnea confirmed the pattern, finding CPAP mostly ineffective at reducing central events. The presence of irregular, ataxic breathing, which opioids can cause, was especially predictive of a poor response to any form of positive airway pressure.⁷

This makes physiological sense. Opioids suppress the brainstem’s respiratory drive through a mechanism distinct from the loop-gain instability that characterizes heart-failure-related central apnea. CPAP can buffer CO2 fluctuations, but it cannot restore a respiratory drive that opioids have chemically dampened. For these patients, reducing opioid doses when medically feasible is often the most effective intervention, sometimes alongside alternative ventilatory support.

Treatment-Emergent Central Apnea During CPAP

Here is a twist that catches many patients off guard: CPAP itself can cause central apneas in people who did not have them before. This phenomenon, called treatment-emergent central sleep apnea, occurs when CPAP applied for obstructive sleep apnea unmasks or triggers a central breathing pattern. In a large dataset of over 133,000 CPAP users, about 3.5% showed central apneas at either week one or week thirteen of therapy. Among those, about 55% had transient central apneas that resolved on their own, roughly 20% developed new central apneas that emerged over time, and about 25% had persistent central apneas that did not go away.⁸

The good news is that in most patients, treatment-emergent central apnea is temporary. It tends to resolve within the first few months of continued CPAP use as the respiratory system adjusts to the new baseline. Because of this, the general recommendation is a period of watchful waiting before switching therapies. Only a select group of patients whose central events persist or worsen may benefit from transitioning to a different device.⁹ Patients with emergent central apnea also tended to have higher mask leak during CPAP, which may contribute to the problem through pressure instability and CO2 washout effects.

Adaptive Servo-Ventilation Versus CPAP

Adaptive servo-ventilation, or ASV, is often presented as the step up from CPAP for central sleep apnea. Unlike CPAP, which delivers a fixed pressure, ASV continuously adjusts its pressure support breath by breath, providing more assistance when the patient’s breathing weakens and backing off when breathing strengthens. In a head-to-head comparison, ASV produced significantly greater reductions in overall apnea events, fewer arousals from sleep, and less time spent in the lightest stage of sleep compared to CPAP.¹⁰ Research has also found that while CPAP leaves substantial residual central apnea in many patients, adaptive bilevel therapy can suppress those remaining events more effectively.¹¹

However, ASV carries a critical safety caveat in one specific population. The SERVE-HF trial tested ASV in heart failure patients with reduced ejection fraction (pumping function at or below 45%) and found that ASV actually increased the risk of dying from cardiovascular causes and from any cause.¹² This finding reshaped clinical practice across Europe and beyond, leading to sharp reductions in ASV prescriptions and updated guideline warnings.¹³ The implication is that in heart failure patients with poor pump function, ASV is contraindicated, and CPAP, despite its limitations, remains a safer positive-pressure option when one is needed. In central sleep apnea patients without severely reduced heart function, ASV remains a viable and often superior alternative to CPAP.

Non-CPAP Alternatives Worth Knowing About

Given CPAP’s uneven performance, several alternatives have been studied for central sleep apnea. Understanding these helps frame what CPAP realistically offers relative to the wider treatment landscape.

Supplemental Oxygen

Nocturnal oxygen therapy works by preventing the oxygen dips that trigger arousals, which in turn drive the overshoot-undershoot breathing cycle. In a study of 20 patients with central sleep apnea from various causes, supplemental oxygen reduced the median number of central apnea and hypopnea events from about 34 per hour to 5 per hour. In 17 of the 20 patients, central events dropped by more than half, and the benefit held regardless of whether heart failure was present.¹⁴ Oxygen is less cumbersome than a PAP machine for some patients, though the network meta-analysis in heart failure patients found that oxygen did not improve ejection fraction the way CPAP and ASV did.⁴

Acetazolamide

Acetazolamide is a medication that shifts the body’s CO2 balance in a way that stabilizes breathing. It works primarily by lowering plant gain, one of the components of loop gain, making the system less prone to the overshoot-undershoot oscillations that produce central apneas.¹⁵ In spinal cord injury patients susceptible to central apnea, three days of acetazolamide widened the gap between normal breathing CO2 levels and the threshold that triggers apnea, making it harder for the system to cross into apnea territory. Central apnea events dropped substantially compared to placebo.¹⁶ Acetazolamide is also used for central apneas at high altitude, where the low oxygen drives hyperventilation that drops CO2 below the apneic threshold.¹⁷

Phrenic Nerve Stimulation

A more recent innovation is an implantable device that electrically stimulates the phrenic nerve, which controls the diaphragm. Rather than pushing air into the lungs from outside like CPAP, this approach triggers the diaphragm to contract from within when the brain fails to do so. In combined data from pilot and pivotal trials, phrenic nerve stimulation roughly halved the number of breathing events per hour and improved quality of life.¹⁸ There is also emerging evidence that it may improve heart function over time in patients with both central sleep apnea and heart failure.¹⁹ The implantation requires a procedure, which limits its accessibility, but it bypasses the adherence problems that plague mask-based therapies entirely.

Central Sleep Apnea After Stroke

Stroke is another common trigger for central sleep apnea, particularly when the stroke damages brainstem areas involved in respiratory control. Sleep-disordered breathing after stroke is widespread, and small studies have shown that treating it with CPAP can improve functional recovery outcomes. The challenge, though, is adherence. Stroke patients often have neurological deficits, cognitive impairment, or facial weakness that makes wearing a CPAP mask difficult or intolerable, and compliance in this group is notably poor.²⁰ So while CPAP may theoretically help, the practical reality is that many post-stroke patients cannot sustain the treatment.

Central Sleep Apnea in Children

The pediatric picture looks quite different from the adult one. Central sleep apnea in children often stems from neurological conditions, brainstem abnormalities, or congenital issues, and CPAP is used much less frequently than other forms of ventilatory support. In a ten-year review of children with central sleep apnea at a tertiary sleep center, bilevel positive airway pressure with a backup breathing rate was initiated in about 38% of patients, while only 4% received CPAP. A significant proportion required supplemental oxygen with or without positive pressure, and some needed invasive ventilation.²¹ The preference for bilevel devices with a backup rate makes sense in this population. Because children with central apnea often have periods where they stop breathing altogether, a device that can initiate breaths on their behalf is more protective than CPAP, which only provides continuous pressure and relies on the patient to trigger each breath.

The Diagnostic Muddle

Part of the difficulty in evaluating how well CPAP works for central sleep apnea is that the diagnosis itself is messier than most people realize. According to the international classification, central apneas and hypopneas must make up more than half of all breathing events, with a central apnea-hypopnea index of at least five events per hour. But accurately distinguishing central hypopneas from obstructive ones has been optional in every edition of the standard sleep-scoring manual, and no large population-based study has ever separately counted them.²² In practice, this means many hypopneas get lumped into the obstructive category by default, which can underestimate the true prevalence of central sleep apnea and muddy the results of treatment studies. Some patients labeled as having obstructive sleep apnea may actually have a significant central component that explains why their CPAP therapy is not working as expected. And studies of CPAP for central sleep apnea may be mixing patients with substantially different proportions of central versus obstructive events, diluting or inflating the apparent treatment effect depending on the sample. Until scoring guidelines require and standardize the separation of central from obstructive hypopneas, this uncertainty will continue to shadow the clinical evidence.