CPAP has long been described as a straightforward win for the heart. New research complicates that picture. An analysis of three randomized trials, published in the European Heart Journal in August 2025, found that continuous positive airway pressure lowered the risk of heart attack, stroke and cardiovascular death by about 17% in one group of obstructive sleep apnea patients, but was linked to a roughly 22% increase in that same risk in another group.
The dividing line was not how severe the sleep apnea looked on a standard test. It was how the body reacted to the pauses in breathing during the night.
A separate machine-learning analysis from Mount Sinai, published in Communications Medicine in April 2026, sorted patients into likely-benefit and likely-harm groups from the same underlying trial data, pointing in the same direction as the European Heart Journal paper.
What the trial analysis actually found
The European Heart Journal analysis pooled three randomized trials, ISAACC, RICCADSA and SAVE, covering 3,549 people with obstructive sleep apnea and existing cardiovascular disease. The median age was 61 years. Half of the participants used CPAP, half did not, and they were followed for about three years, according to the Mass General Brigham newsroom release on the study.
Also read: How to protect your back with simple daily habits
Overall, the two groups looked almost identical. About 16.6% of CPAP users had a major cardiac event, compared with 16.3% of non-users, per the Mass General Brigham release. The average washed out an important split underneath.
When the researchers separated patients using two physiologic markers of how the body reacts to apnea, hypoxic burden and pulse rate response, a clear pattern emerged. Patients with a large drop in blood oxygen or a large heart-rate spike during breathing pauses saw about a 17% lower cardiovascular risk on CPAP. Patients without those markers saw about a 22% higher risk.
The effect was even sharper in patients who did not feel sleepy during the day. Non-sleepy patients with high-risk markers had 24% fewer cardiovascular events on CPAP, while non-sleepy patients with low-risk markers had 30% more, per the Mass General Brigham release.
The senior author, Dr. Ali Azarbarzin of Brigham and Women's Hospital and Harvard Medical School, said the results point toward a more individualized approach. "Through our study, we found a subgroup of patients who experience cardiovascular benefits from CPAP use," Azarbarzin said in the Mass General Brigham release, calling it "the first step in making better therapeutic recommendations."
Also read: Low magnesium may be quietly pushing your blood sugar higher, even before diabetes
An AI model tried to predict who benefits
The Mount Sinai team went a step further, building a machine-learning model on the SAVE trial cohort, which included more than 2,600 participants from 89 sites in seven countries. Starting from more than 100 candidate variables, the researchers narrowed the model to 23 baseline features drawn from sleep and health data, Medical Xpress reported.
The model then estimated, for each patient, whether CPAP would likely reduce or increase the risk of major cardiovascular events. In the subgroup the model tagged as benefiting, the estimated cardiac risk fell sharply on CPAP versus usual care. In the subgroup tagged as harmed, the estimated risk moved sharply in the wrong direction. Those are modeled estimates rather than direct outcome counts, but the direction matches the trial analysis.
Dr. Neomi A. Shah of the Icahn School of Medicine at Mount Sinai said the findings "represent a significant advancement in personalized medicine, moving away from a one-size-fits-all strategy in the treatment of obstructive sleep apnea."
What this means for patients on CPAP
Neither research team is telling patients to stop using CPAP on their own. Both are questioning the idea that everyone with obstructive sleep apnea should be treated the same way when the goal is protecting the heart. The standard severity measure, the apnea-hypopnea index, does not capture how much the body strains during apnea, and it is that strain, seen as oxygen drops and heart-rate spikes, that appears to separate likely responders from likely non-responders.
Also read: A nutrient that 9 in 10 Americans miss may be quietly harming their brains
Anyone using CPAP mainly to lower cardiovascular risk should raise these findings with their sleep specialist or cardiologist before changing anything. For patients using CPAP because it helps them sleep and function during the day, the calculus is different, and the therapy remains a mainstream treatment for symptomatic sleep apnea.
Obstructive sleep apnea affects an estimated 25 million adults in the United States, according to Medical Xpress. The share of them who might fall into each of these subgroups is not yet clear, and validation in additional trials will be needed before risk-based prescribing becomes routine practice.
Also read: Study of 153,902 people challenges effectiveness of calcium and vitamin D supplements