Sleep apnea and your heart: what the evidence actually shows

Observational studies link obstructive sleep apnea with hypertension, atrial fibrillation, coronary disease, heart failure, and stroke. The largest randomized secondary-prevention trial of CPAP, SAVE, did not reduce a composite of cardiovascular events. Both of those sentences can be true. The honest version for a patient is that apnea belongs in a heart-risk conversation, that treating it reliably improves sleepiness and quality of life, and that CPAP has not been shown, in the trial that was built to ask the question, to prevent heart attacks and strokes in people who already had coronary or cerebrovascular disease and were not very sleepy.

What do the observational studies actually show?

Javaheri’s 2017 JACC review summarizes a large observational literature: OSA is common in cardiovascular clinics, largely undiagnosed, and associated with incident hypertension, AF, coronary disease, heart failure, and stroke. The authors call OSA a cause of systemic hypertension and a modifiable cardiovascular risk factor, while still having to account for trials that came out around the same time. Intermittent hypoxia, sympathetic surges, oxidative stress, and endothelial dysfunction are the mechanistic story they lay out. That is pathophysiology plus association. It is not the same design as a randomized trial of treatment.

The 2021 American Heart Association scientific statement is the document cardiologists are supposed to have read. It estimates that about 34 percent of middle-aged men and 17 percent of middle-aged women meet diagnostic criteria for OSA, and that OSA prevalence runs as high as 40 to 80 percent in hypertension, heart failure, coronary disease, pulmonary hypertension, AF, and stroke. It recommends screening for OSA in resistant or poorly controlled hypertension, pulmonary hypertension, and recurrent AF after cardioversion or ablation. It is frank that OSA is under-recognized in cardiovascular practice.

Peppard’s 2000 Wisconsin analysis of weight and AHI, and the broader Wisconsin and Sleep Heart Health observational programs, are why those associations have been hard to ignore. They are still observational. People with severe OSA differ from people with AHI under 5 in body weight, blood pressure, and a dozen other ways. Adjustment helps. It does not turn association into proof that treating the AHI will change a heart-attack rate.

What did SAVE actually test?

SAVE (McEvoy, New England Journal of Medicine, 2016) randomly assigned 2,717 adults aged 45 to 75 with moderate-to-severe obstructive sleep apnea and established coronary or cerebrovascular disease to CPAP plus usual care or usual care alone. After site exclusions, 2,687 entered the primary analysis. It was a secondary-prevention trial: everyone already had heart or brain artery disease. There was a one-week sham-CPAP run-in; people who could not manage about 3 hours a night on sham were not randomized. That design selected for people who could tolerate a mask and excluded many of the sleepiest patients, which matters when you try to apply SAVE to a sleepy truck driver in Westerville.

Mean age was 61. Eighty-one percent were men. Sleepiness was often minimal. In the CPAP arm, mean adherence over follow-up was 3.3 hours per night. Mean AHI fell from 29.0 events per hour at baseline to 3.7 on treatment, so the machine did what machines do when they are on. Mean follow-up was 3.7 years.

The primary composite (death from cardiovascular causes, myocardial infarction, stroke, or hospitalization for unstable angina, heart failure, or transient ischemic attack) occurred in 229 of the CPAP group (17.0 percent) and 207 of usual care (15.4 percent). Hazard ratio 1.10, 95 percent confidence interval 0.91 to 1.32, P = 0.34. No significant effect on the individual cardiovascular pieces. CPAP did reduce snoring and daytime sleepiness and improved quality of life and mood. Those secondary results are why the trial is not an argument against offering CPAP therapy to people who are sleepy or whose sleep is wrecked. It is an argument against promising them a lower chance of the next stent.

Why can SAVE be negative if the associations are real?

Several non-exclusive explanations sit in the paper and in the reviews that followed. Adherence of 3.3 hours may be too little if the cardiovascular injury accumulates across the whole night. The participants were not very sleepy, so the population that benefits most for symptoms was thinned out by design. Usual cardiovascular care in a secondary-prevention cohort is already aggressive, which makes an add-on harder to see. Diagnosis used a simplified ApneaLink pathway rather than full polysomnography. And the observational mortality signal may be partly residual confounding: severe OSA travels with obesity, diabetes, and blood pressure.

Javaheri’s JACC paper, written with SAVE in view, reports that CPAP lowers blood pressure by a few millimeters of mercury, more so in resistant hypertension, and that arrhythmia data are mostly observational. The AHA statement’s blood-pressure paragraph is similarly modest: meta-analyses in hypertensive OSA show roughly 2 to 3 mm Hg reductions, with about 3 mm Hg in a resistant-hypertension CPAP trial over 3 months, and a correlation with hours of use. A 2 mm Hg change is not nothing in population terms. It is not a reason to tell one patient that CPAP will replace their ACE inhibitor.

On AF, the AHA writing group says it has not been definitively proven that OSA causes AF. Observational series after ablation or cardioversion generally favor CPAP users. Those series are small or registry-based. They are not SAVE. Screening for OSA after recurrent AF is still the statement’s recommendation, because the association is strong and the treatment has other benefits. That is a reasonable clinical posture. It is not a claim that PAP is an antiarrhythmic.

So is the blood-pressure and AF story just hype?

No. It is a different kind of evidence. The hypertension signal has randomized CPAP data showing small average reductions, larger in some resistant-hypertension groups, and a mechanistic story (sympathetic activation, blunted nocturnal dipping) that is tighter than the coronary-event story. The AF signal is association plus plausible atrial strain, plus observational treatment data, minus a large randomized trial with AF burden as the primary endpoint.

What you should not hear in this office is that untreated apnea “will give you a heart attack.” What you should hear is that your cardiologist and your sleep physician are looking at overlapping risk, that treating OSA is justified for sleepiness, quality of life, and blood-pressure adjunct effect, and that we will not sell CPAP as secondary prevention of myocardial infarction after SAVE.

If you already have a cardiologist, bring the last echo, the AF history, and the medication list. Heart failure in particular changes the testing pathway, because central sleep apnea enters the differential and a home test can mis-label it. That is a Kapur 2017 point as much as a cardiology point.

If CPAP did not prevent events in SAVE, why treat?

Because SAVE’s primary endpoint is not the only outcome that counts, and because you may not be a SAVE patient. The AASM 2019 PAP guideline (Patil) gives a strong recommendation to use PAP, compared with no therapy, in adults with OSA and excessive sleepiness. It gives conditional recommendations for impaired sleep-related quality of life and for comorbid hypertension. Those recommendations do not depend on SAVE being positive. They depend on symptoms and on blood pressure as an intermediate.

You also may be sleepier, more adherent, or free of prior infarction. SAVE cannot tell us what CPAP does as primary prevention in a sleepy 48-year-old with an AHI of 40 and no known coronary disease. Absence of evidence in that group is not evidence of absence. It is a reason to be precise about which question was asked.

Dr. Qadoom’s role, described on the physician page, is to diagnose the breathing disorder and to treat it as sleep medicine, in conversation with your other physicians. We read SAVE. We still offer PAP for the indications the AASM treats as strong. We still send people back to cardiology with a diagnosis rather than a scare line.

How should a patient in Westerville use this?

If you have resistant hypertension, pulmonary hypertension, or AF that relapsed after a procedure, the AHA statement already wants a sleep evaluation. If you have sleep apnea and a heart history, treat the apnea for sleep and blood pressure, and let cardiology manage stents and anticoagulation. If someone used a heart-attack graphic to sell you a machine, ask them what they do with McEvoy 2016. If they have not read it, they are advertising.

Call (614) 898-9340 to schedule. Bring cardiac records if you have them. Expect a conversation about symptoms and testing, not a promise about mortality.