AHI Score Explained: What the Number Really Tells You About Your Sleep Apnea

A sleep study report is mostly one number. Your AHI score, short for apnea-hypopnea index, is the average count of breathing events per hour of sleep, and it is the number that decides which severity label goes on your chart, which treatments a physician brings up, and in most of the United States what a plan will cover.

Here is how soft that number is. In a general-population study of 2,162 adults, the share who met the AHI 15 threshold was 36.6 percent under one accepted scoring rule and 18.8 percent under another.³ Same people, same recordings, two definitions both published by the same professional body, roughly double the number of people classified as having moderate or worse sleep apnea.

The Short Answer

That is the whole answer. The rest of this article is about the four things that move an AHI, because knowing them is the difference between reading your report and over-reading it.

What an AHI Score Actually Counts

An apnea is a breathing pause lasting at least ten seconds. A hypopnea is a partial reduction in breathing, also at least ten seconds, that either drops your blood oxygen by a set amount or ends in a brief arousal from sleep. The index is plain arithmetic: apneas plus hypopneas, divided by hours of sleep. An AHI of 22 means 22 events per hour, averaged across the night.

That arithmetic throws away three things.

• Duration. A ten-second pause and a fifty-second pause each count as one event.
• Depth. A two-point oxygen drop and a twenty-point drop each count as one event.
• Timing. Because the index is an hourly average across the whole night, it flattens clustering. Events concentrated in REM sleep, or in the hours you spend on your back, can average out to a modest whole-night number while individual stretches of the night look far worse.

An AHI counts events. It does not weigh them. Two people with the same index can have measurably different nights.

AHI Severity Levels, and Where the Lines Fall

The conventional bands are straightforward:²

• Under 5: not considered sleep apnea
• 5 to 14: mild obstructive sleep apnea
• 15 to 29: moderate obstructive sleep apnea
• 30 or higher: severe obstructive sleep apnea

Sources word the boundaries slightly differently, which is itself a hint about how much weight one point is meant to carry. Our page on the clinical detail lays out the same bands.

These are conventions for grouping a continuous measurement, not biological cliffs. Nothing in your airway changes between an AHI of 14.6 and an AHI of 15.2. What changes is the label, and in the United States what often changes with the label is access to treatment.

Why a Home Test Reports an REI, Not an AHI

If your test was done at home, your report probably says REI rather than AHI, and the difference is the denominator.

An in-lab polysomnogram measures brain activity, so it knows how much of the night you were actually asleep and divides by that. A home sleep apnea test, or HSAT, records breathing and oxygen but not brain activity, so it has no way to separate sleeping from lying still. It divides by total recording time instead. The American Academy of Sleep Medicine's 2017 clinical practice guideline is explicit that this measurement limitation "results in use of the recording time rather than sleep time to define the denominator of the respiratory event index (REI)."¹

Both indexes count the same events. They divide by different things, which is why they are not interchangeable numbers.

Both differences push the same way. Time awake in bed inflates the denominator, and hypopneas that end in an arousal without a measurable oxygen drop cannot be scored at all without EEG. An REI is therefore more likely to read lower than a lab AHI on the same night than higher.

That asymmetry is why the guideline's Recommendation 3, graded STRONG, is that polysomnography be performed if a single home sleep apnea test comes back negative, inconclusive, or technically inadequate.¹ The same guideline recommends the lab rather than a home test from the start for adults with significant cardiorespiratory disease, potential respiratory muscle weakness from a neuromuscular condition, hypoventilation while awake or suspected hypoventilation during sleep, chronic opioid use, a history of stroke, or severe insomnia.¹ A breathing-only recording also cannot address narcolepsy, REM sleep behavior disorder, or periodic limb movement disorder.

Dorma Health, a sleep medicine company that runs at-home sleep apnea testing, screens for that list before shipping anything. A physician on our network reviews every order for clinical appropriateness, and if a home test is not the right tool, the order is refunded in full. You can read more about how the process works.

The Same Night Can Produce Two Different AHI Scores

This is the part almost nobody explains, and it matters more than the decimal place on your report.

Counting apneas is uncontroversial. Counting hypopneas is not, because a hypopnea has to be defined by how much your oxygen fell or whether you woke up, and there is more than one accepted definition in use. A stricter rule requires a 4 percent oxygen desaturation. A more inclusive rule accepts a 3 percent desaturation or an arousal from sleep. The stricter rule finds fewer hypopneas, so it produces a lower AHI on the same recording.

The size of that effect is not small. When researchers scored the same 2,162 general-population recordings under three published definitions, the proportion of people with an AHI of 15 or higher was 36.6 percent under the 2012 criteria and 18.8 percent under the 2007 criteria.³ The recordings did not change. The rule did.

Medicare's national coverage policy for CPAP still defines a hypopnea as requiring at least a 30 percent reduction in airflow or effort plus at least a 4 percent oxygen desaturation.⁵ So an AHI scored under the more inclusive clinical rule and an AHI scored under the coverage rule are not the same quantity, even when they are printed in the same font on the same page.

If your number sits near a threshold, the scoring rule your lab used is a fair thing to ask about.

Your AHI Moves From Night to Night

Your AHI is a measurement of one night, reported as though it were a property of you.

The largest look at this tracked 67,278 adults for an average of about 170 nights each, using an under-mattress sensor rather than a polysomnogram, which is a real caveat about absolute accuracy but not about variability.⁴ Among people whose multi-night average put them clearly in the sleep apnea range, a single night identified them correctly 79 percent of the time. Mild and moderate cases were the least stable, matching a person's overall classification on only about 54 percent and 52 percent of nights.⁴

Alcohol, congestion, sleep position, prior sleep deprivation, and how much REM you got all move the number. None of that makes a single-night study useless. It means a result near a boundary is an estimate with a range around it, not a verdict.

Why the Number Does Not Match How You Feel

Plenty of people with a severe AHI feel fine, and plenty of people with a mild AHI feel wrecked.

In a review of 200 patients, half had a sleepiness score that disagreed with their AHI category. About 29.5 percent scored in the normal range on the Epworth Sleepiness Scale despite an AHI of 15 or higher, and 20.5 percent reported excessive sleepiness despite an AHI below 15.⁶ The two measurements are answering different questions, and neither one overrides the other in a real clinical conversation.

This is the most common reason people distrust their own report. A mild number alongside real exhaustion is not a contradiction to explain away, and it is worth raising with the physician who read your study.

What Your AHI Score Actually Changes

Two concrete things.

It sets the treatment conversation. Recommendations follow severity, and the standard options include positive airway pressure (CPAP, APAP, or BiPAP), an oral appliance, and positional therapy. Which one fits is a clinical decision, not something a number decides on its own.

It sets coverage eligibility. Medicare's national coverage determination for CPAP requires an AHI or RDI of 15 or more, or 5 to 14 with a documented symptom or condition from a defined list: excessive daytime sleepiness, impaired cognition, mood disorders, insomnia, hypertension, ischemic heart disease, or a history of stroke.⁵ That policy also accepts unattended home sleep testing with a Type II, Type III, or qualifying Type IV device as the diagnostic basis.⁵ Commercial plans set their own criteria, which often resemble this one without copying it.

Dorma is cash-pay only and does not bill insurance, so none of this changes what a test costs here. It matters later, when you take a signed report to whoever will be paying for treatment.

How to Read Your Own Report

Six things worth locating on the page:

1. Whether the number is an AHI or an REI. That tells you whether the denominator was sleep time or recording time.
2. The hypopnea rule used. Usually stated as a 3 percent or 4 percent desaturation criterion in the methods or scoring notes.
3. Total recording time next to total sleep time. A large gap between them means a diluted index.
4. The supine breakdown, if reported. A whole-night average can hide a much higher rate while you are on your back.
5. The lowest oxygen saturation of the night. The index does not carry this, and it is clinically meaningful on its own.
6. The interpreting physician's written impression. This is the part that actually constitutes the diagnosis.

Your AHI score is the headline of your report and the right place to start, but it is a summary of one night, produced by one scoring rule, and it does not know how you feel. Read it as the beginning of a conversation with the physician who signed it. If something on your report is not covered here, it may be in our frequently asked questions, and if you have not tested yet, the difference between a home test and a sleep lab is the earlier decision to make.

Frequently Asked Questions

What is a normal AHI score?

An AHI below 5 events per hour is considered normal in adults. From there, 5 to 14 is classified as mild obstructive sleep apnea, 15 to 29 as moderate, and 30 or higher as severe. These bands are grouping conventions for a continuous measurement, so a result of 4.8 and a result of 5.2 are far more similar than their labels suggest.

Is an AHI of 15 bad?

An AHI of 15 is the conventional boundary between mild and moderate obstructive sleep apnea, and it is also the threshold in Medicare's national coverage policy for CPAP.⁵ Whether it warrants treatment depends on your symptoms and your other medical conditions, which is a decision for the physician who read your study. A number this close to a line is also worth reading as an estimate rather than an exact value.

Why is my AHI different from my last sleep study?

Because AHI varies from night to night and because different scoring rules produce different numbers. In a study of 67,278 adults recorded for about 170 nights each, individual nights matched a person's overall classification only about half the time for mild and moderate cases.⁴ Separately, scoring the same recordings under a 4 percent rather than a 3 percent desaturation rule roughly halved the proportion of people above the AHI 15 threshold.³

Does a home sleep apnea test give you an AHI score?

A home sleep apnea test produces a respiratory event index, or REI, which counts the same events but divides by total recording time rather than confirmed sleep time.¹ Because time spent awake inflates that denominator, and because hypopneas ending in an arousal cannot be scored without EEG, an REI tends to read lower than a lab AHI on the same night. A physician uses it to make the diagnosis, and if it is negative or inconclusive while symptoms persist, the AASM recommends in-lab polysomnography.¹

Does my AHI score decide whether insurance covers CPAP?

It is the main input, not the whole test. Medicare's coverage determination requires an AHI or RDI of 15 or higher, or 5 to 14 alongside a documented symptom or condition from a defined list.⁵ Commercial plans set their own criteria. Dorma is cash-pay only and does not bill insurance, so we cannot predict what any specific plan will approve.