The Sleep Apnea Most People Don't Know They Have

A man in his early fifties comes for a preventive assessment. His blood pressure has not responded to two medications, his fasting glucose has drifted upward over three years, and he mentions almost as an aside that he is no longer sharp in afternoon meetings. Nothing on the standard panel explains it. The likeliest single explanation appears on no blood test: he stops breathing dozens of times an hour while he sleeps. He has obstructive sleep apnoea, and has had it for years.

Obstructive sleep apnoea is among the most common serious conditions in adult medicine and among the least often diagnosed. Prevalence estimates vary with the scoring criteria used, but every large population study agrees it is common — on the order of one in ten adults have clinically significant disease, and more among older, heavier, and post-menopausal people. The diagnostic gap is not in dispute: the large majority of those with moderate-to-severe disease have never been tested.

It is not only loud, heavy, older men

The clinical stereotype — a large, snoring, sleepy man whose wife elbows him at night — describes a real patient, but it does more diagnostic damage than any single risk factor. In women, apnoea more often presents as insomnia, unrefreshing sleep, morning headache, fatigue rather than frank sleepiness, and low mood — a picture attributed to stress, perimenopause, or depression long before anyone considers the airway. Risk rises sharply after menopause, and women are referred later and less often than men with equivalent disease.

Lean patients are missed differently: collapse is a problem of anatomy as much as adiposity, and a set-back jaw, a high narrow palate, dental crowding, large tonsils, or chronic nasal obstruction each narrow the pharynx at an unremarkable BMI.

• Witnessed pauses, choking, or gasping — the one history patients cannot supply themselves
• Waking unrefreshed, or waking to urinate twice or more without another explanation
• Blood pressure resistant to two or more agents, or not falling overnight
• Atrial fibrillation, especially when it recurs after cardioversion or ablation
• Morning headache, dry mouth, teeth grinding, or habitual mouth breathing

What actually happens during an apnoea

Sleep reduces the tone of the pharyngeal muscles that hold the upper airway open, and in a susceptible airway the negative pressure of inspiration collapses it. Effort continues against a closed tube; oxygen falls, carbon dioxide rises, and the event ends in a brief cortical arousal that restores tone and reopens the airway. Sleep resumes, tone falls, the cycle repeats — in severe disease, several hundred times a night.

Three separable insults follow. Intermittent hypoxaemia promotes oxidative stress, endothelial dysfunction, and inflammation more efficiently than sustained low oxygen would. Sympathetic surges accompany every arousal, and over years that elevated tone persists into the waking day. Sleep fragmentation dismantles the architecture of the night, so a patient spends eight hours in bed and obtains very little deep or REM sleep. Breathing against a closed airway also generates large negative intrathoracic pressure swings that stress the atria mechanically.

Sleep apnoea may be the only major cardiometabolic risk factor whose defining event can, by definition, only be witnessed by someone else. Patients do not remember the arousals. They remember the consequences, and rarely connect the two.

Why it matters downstream

Hypertension and arrhythmia. Sleep apnoea is the most common identifiable contributor to resistant hypertension, and it frequently abolishes the normal nocturnal fall in blood pressure — a non-dipping pattern that carries risk in its own right. The association with atrial fibrillation is equally consistent: untreated apnoea predicts higher recurrence after cardioversion and after ablation.

Metabolic and cognitive effects. Severity is associated with insulin resistance and incident type 2 diabetes even after adjustment for adiposity. Obesity drives both, so causality is hard to isolate, but the mechanism is not speculative. Patients also describe reduced attention and unreliable working memory, and the observational link to later cognitive decline is plausible but unproven.

Driving. Untreated moderate-to-severe apnoea carries a clearly elevated crash risk, and treatment reduces it. In someone who drives professionally, that alone justifies testing.

Screening for sleep apnoea begins with questions, not a machine

The first step is a validated questionnaire. STOP-BANG covers snoring, tiredness, observed apnoeas, high blood pressure, BMI above 35, age above 50, neck circumference above 40 cm, and male sex; three positives marks intermediate risk, five high risk. Its sensitivity for moderate-to-severe disease is good and its specificity poor — which is appropriate, since its job is to decide who gets tested, not who has the disease.

Its weakness is instructive: half the items favour the classic phenotype, which is precisely how a lean forty-year-old woman with a narrow palate scores two and is sent home. The Epworth Sleepiness Scale correlates poorly with severity and excludes nothing when normal — many patients have spent years recalibrating what counts as normal tiredness.

Home test or sleep laboratory

A home sleep apnoea test records nasal airflow, respiratory effort, and oxygen saturation. With a high pretest probability and no significant cardiopulmonary comorbidity it confirms the diagnosis cheaply, and it can be repeated over several nights — useful, since night-to-night variability is real.

Its limitations matter as much as its convenience. Without EEG there is no sleep staging and no measure of how long the patient actually slept, so events are divided by recording time rather than sleep time; someone who slept four of eight hours in bed can have severity understated by half. A negative home study where clinical suspicion remains high is not a negative result. It is an indication for polysomnography.

Severity is graded by the apnoea-hypopnoea index: 5 to 15 events per hour is mild, 15 to 30 moderate, above 30 severe. The index is crude — two patients at 20 can differ substantially in how deeply and how long they desaturate — and measures of total hypoxic exposure appear to track cardiovascular outcomes better. Ask for the oxygen data as well.

The treatments, compared honestly

CPAP remains first-line for moderate-to-severe disease; nothing else abolishes obstructive events as completely. The problem has never been efficacy but use. Real-world long-term adherence sits near half of patients, and benefit is dose-dependent, measured in hours per night. That is how to read the trial evidence: the large randomised trials of CPAP for cardiovascular prevention, of which SAVE is the best known, showed no reduction in events, with average use around three hours a night — leaving untreated the part of the night when REM-related events cluster. They show that CPAP as typically used does not prevent cardiovascular events, not that well-treated apnoea is equivalent to no apnoea. Most abandonment is fixable in the first weeks or not at all: mask fit, humidification, pressure titration, nasal obstruction.

Mandibular advancement devices — custom appliances holding the lower jaw forward — lower the apnoea-hypopnoea index less than CPAP but are worn more consistently, and head-to-head comparisons show comparable effects on sleepiness and blood pressure, because what matters clinically is efficacy multiplied by adherence. They are a reasonable first choice in mild-to-moderate disease and the main alternative in CPAP intolerance; the trade-off is gradual dental movement.

Positional therapy helps the minority whose events are markedly more frequent supine, and only where the study demonstrates that dependence. Weight loss is disease-modifying where obesity is the driver: losing roughly 10 percent of body weight typically produces a clinically meaningful fall in severity, and larger losses can resolve disease. Incretin-based weight-loss pharmacotherapy substantially reduces severity in randomised trials in obesity with apnoea — part of the discussion alongside airway therapy, not a replacement for it. Skeletal anatomy does not respond to weight loss.

Surgery is for selected patients: nasal surgery improves CPAP tolerance but rarely cures apnoea alone, tonsillectomy can be definitive when the tonsils are large, hypoglossal nerve stimulation suits CPAP-intolerant patients meeting anatomical criteria, and maxillomandibular advancement is most effective where the restriction is skeletal.

What improves, and how quickly

• Sleepiness and daytime function — days to weeks, and usually the change patients notice first.
• Nocturia and morning headache — often within the first few weeks.
• Blood pressure — a few mmHg over weeks to months, larger in resistant hypertension and with more than four hours of nightly use.
• Atrial fibrillation — observational data show lower recurrence after cardioversion and ablation in treated patients; randomised evidence is limited.
• Glucose and insulin sensitivity — inconsistent across trials: some improvement in insulin sensitivity, no reliable change in HbA1c.
• Cognition — partial improvement in attention and vigilance. Whether treatment alters long-term dementia risk is genuinely unknown.

The evidence is strongest for symptoms and daytime function, intermediate for blood pressure and arrhythmia, and unresolved for hard cardiovascular endpoints. That is an argument for diagnosing and treating the condition properly, not for leaving it undiscovered.

What to ask your physician

• Ask to be screened even if you do not snore loudly and are not overweight — particularly if you are a woman with unexplained fatigue or insomnia, or you have a set-back jaw.
• If your hypertension has not responded to two or more medications, or you have atrial fibrillation, ask directly whether sleep apnoea has been excluded. Often it has not been considered.
• Ask whether a home test is adequate for you, and if it is negative while symptoms persist, ask for polysomnography rather than accepting the result.
• Ask for lowest saturation, time below 90 percent, and whether events clustered supine or in REM — those details determine which treatment fits.
• If you start CPAP, ask how adherence and pressure will be reviewed over the first 90 days — the first three months predict long-term use better than anything else.

Most preventive medicine involves shifting risk slowly and measuring the result in years. Sleep apnoea is one of the few places where a single night of testing identifies something common, consequential, and treatable within weeks. The obstacle is not technology or cost. It is that someone has to think of it.