What Is Sleep Apnea? Types, AHI, and Severity Explained

Sleep apnea is a common but often-missed sleep disorder in which your breathing repeatedly stops or shrinks while you sleep — sometimes hundreds of times a night. The good news is that it is measurable and treatable. This guide explains what actually happens during an event, the three main types, how clinicians count those events, and how the standard severity scale (the AHI) turns a night of disrupted breathing into a single, comparable number.

This is the foundation page for our Understanding Sleep Apnea pillar. Once you grasp the vocabulary here, the rest of the hub — symptoms, risk factors, and how to read your own therapy data — will make a lot more sense.

What happens during sleep apnea

When you fall asleep, the muscles that hold your upper airway open relax along with the rest of your body. In someone with sleep apnea, that relaxation goes too far, the airway narrows or closes, and airflow drops. Your blood oxygen falls, carbon dioxide builds up, and your brain briefly wakes you — often so briefly you never remember it — to restart normal breathing.

That cycle can repeat dozens or hundreds of times per night. Each interruption fragments your sleep and stresses your cardiovascular system, even when you have no memory of waking. This is why people with sleep apnea can spend eight hours in bed and still wake up exhausted: the quantity of sleep may be fine, but the quality is shredded.

Over time, that repeated stress is why sleep apnea is more than a snoring nuisance. It is associated with — and an independent risk factor for several of — hypertension, atrial fibrillation, coronary artery disease, heart failure, pulmonary hypertension, and stroke, and it is also linked to type 2 diabetes. We cover that in depth in Sleep Apnea and Your Heart; the takeaway here is simply that the events you can't feel still matter.

The main types — obstructive, central, mixed/complex

Not all apneas have the same cause. They fall into three mechanistically distinct categories, and telling them apart drives how the condition is treated.

Obstructive sleep apnea is by far the most common form and the one most CPAP therapy is designed for. Central sleep apnea is less common and reflects a problem with respiratory drive rather than a mechanical blockage. Mixed apnea combines both, and a related pattern — treatment-emergent central sleep apnea — can appear after someone starts CPAP.

A practical note for people already on therapy: CPAP is built to splint the airway open, so it reliably treats obstructive events, but it does not directly correct the unstable breathing drive behind central events. If central (also labeled "clear airway," or CA) events rise after you start CPAP, that may be treatment-emergent central sleep apnea, which often settles on its own within weeks to a few months of continued use (reported spontaneous resolution is roughly 60–80%). If it persists, your clinician may consider a different mode such as BiPAP or ASV. The wrong move is to raise your own pressure to chase those events — higher pressure does not fix central apneas and can sometimes provoke them. We unpack this fully in Central Apneas Showing Up on CPAP and Central vs Obstructive Apnea.

How apnea and hypopnea are defined

Two specific events make up the score that defines your severity. Both have precise, time-based definitions.

• Apnea — a near-complete reduction in airflow lasting at least 10 seconds. Effectively, breathing all but stops for ten seconds or more.
• Hypopnea — a partial collapse: airflow drops by at least 30% for 10 or more seconds, accompanied by either a drop in blood oxygen or a brief arousal from sleep.

The reason a hypopnea needs that extra qualifier — an oxygen desaturation or an arousal — is that a partial airflow dip only "counts" if it actually harms your sleep or oxygenation. Here the rules get nuanced. The American Academy of Sleep Medicine (AASM) offers two scoring criteria:

• Rule 1A (recommended): a 30% airflow reduction with either a 3% oxygen desaturation or an arousal.
• Rule 1B (acceptable; the criterion Medicare/CMS often uses): a 30% airflow reduction with a 4% oxygen desaturation (arousals don't count).

Which rule a lab uses can shift your hypopnea count — and therefore your final score — which is one reason the same person can get slightly different numbers from different facilities.

This also explains an important caveat for home users: the AHI your CPAP machine reports is estimated. A sleep lab measures brain waves (EEG) to detect arousals; your machine can't, so it infers events from airflow and pressure signals alone. Device-reported AHI is genuinely useful for tracking trends, but it is not identical to a lab-scored AHI. For a deeper walk-through of what your machine actually counts, see CPAP Event Types Decoded.

AHI and the severity bands

The Apnea-Hypopnea Index (AHI) is simply the average number of apneas plus hypopneas per hour of sleep. Add up every apnea and hypopnea across the night, divide by hours slept, and you get one comparable number. The AASM sorts that number into four bands:

So an AHI of 9 is mild, 22 is moderate, and 41 is severe. These bands are the common language clinicians, insurers, and equipment suppliers use to describe your diagnosis.

A few important nuances:

• AHI isn't the whole story. It excludes RERAs (respiratory effort-related arousals) and flow limitation — subtle breathing disruptions that fragment sleep without meeting the apnea or hypopnea threshold. A related index, the RDI, captures those. See AHI vs RDI and UARS for the events your AHI can miss.
• On treatment, the goal is different from the diagnosis. Once you're on CPAP, the widely used benchmark for effective therapy is a residual AHI below 5 events per hour — the AASM defines an "optimal" titration as reducing AHI to fewer than 5. Some clinicians aim lower (for example, under 1–2) when it's comfortably achievable, but there's no formal guideline establishing a "below 2" target, and goals are individualized with your provider. What Is a Good AHI on CPAP covers this in detail.
• One night is noise; trends matter. Any single night's AHI can swing with alcohol, congestion, sleeping position, or a leaky mask. What matters clinically is the pattern over weeks. Why Does My AHI Change Night to Night explains the normal variation.

One trust-building detail worth knowing now: a large mask leak can invalidate or under-report your AHI, because the machine loses the signal it needs to detect events. A flattering low number on a high-leak night isn't reassuring — it's unreliable. This is exactly the kind of context that raw app scores hide. Once you're diagnosed and on CPAP, SomniCharts imports your data from ResMed, Philips Respironics (including the DreamStation 2, whose SD-card data is encrypted and unreadable by most tools), and Löwenstein prisma machines, then explains your AHI, leak rate, and event types in plain language — across brands, automatically, so you can have an informed conversation with your provider instead of guessing.

When to seek evaluation

Sleep apnea is diagnosed through clinical evaluation, not self-assessment. An online quiz or your watch's "sleep score" can raise suspicion, but only a clinician working from a validated sleep study can confirm the diagnosis and grade its severity.

Consider talking to a doctor if you have any of these common signs:

• Loud, habitual snoring, especially with gasping, choking, or witnessed pauses in breathing
• Waking unrefreshed despite adequate time in bed
• Excessive daytime sleepiness, brain fog, or morning headaches
• High blood pressure that's hard to control, or a heart-rhythm condition such as atrial fibrillation

Our guides on sleep apnea symptoms and when to see a doctor about snoring go deeper on the warning signs.

If your clinician decides testing is warranted, there are two main paths. Per the AASM Clinical Practice Guideline for Diagnostic Testing for Adult OSA (Kapur et al., 2017), a home sleep apnea test (HSAT) with an adequate device may be used for uncomplicated adults who show signs of an increased risk of moderate-to-severe OSA. The guideline recommends in-laboratory polysomnography (PSG) instead of an HSAT for people with significant cardiorespiratory disease, possible respiratory muscle weakness from a neuromuscular condition, awake or suspected sleep-related hypoventilation, chronic opioid use, a history of stroke, or severe insomnia — and PSG is required to diagnose central sleep apnea or hypoventilation. If a home test comes back negative, inconclusive, or technically inadequate, a full in-lab study should follow. We compare both in In-Lab PSG vs Home Test.

The bottom line: if the signs are there, get evaluated. Sleep apnea is highly treatable once it's properly diagnosed — and the numbers in this guide are the same ones you'll use, night after night, to confirm that your therapy is actually working.

Frequently asked questions

What is the difference between obstructive and central sleep apnea? In obstructive sleep apnea the airway physically collapses while your body still tries to breathe (effort against a blockage). In central sleep apnea the airway stays open but your brain briefly stops signaling the muscles to breathe, so effort stops too. CPAP reliably treats obstructive events; central events stem from breathing drive and may need a different approach under a clinician's guidance.

What AHI is considered severe sleep apnea? An AHI of 30 or more events per hour is classified as severe by the AASM. Below that, 15 to under 30 is moderate, 5 to under 15 is mild, and under 5 is normal.

Is an AHI of 5 normal? An AHI under 5 is the normal, non-apneic range. An AHI of exactly 5 or higher crosses into mild sleep apnea. On CPAP therapy, a residual AHI below 5 is the widely used benchmark for effective treatment.

Can I diagnose sleep apnea myself? No. Symptoms and screening tools can flag your risk, but a confirmed diagnosis requires a clinical evaluation and a validated sleep study (home test or in-lab PSG) interpreted by a clinician.