Learn PPG
What Is AHI (Apnea-Hypopnea Index)?
AHI, the apnea-hypopnea index, is the number of apneas and hypopneas per hour of sleep, scored from a sleep study. In adults, AASM guidance grades obstructive sleep apnea as mild from 5 to under 15 events per hour, moderate from 15 to 30, and severe above 30. A home sleep apnea test reports the REI, which divides by recording time instead of sleep time. A watch or ring can estimate related signals. It does not score AHI, and its number is not a diagnosis.
What AHI counts
AHI is a rate. Count the apneas, add the hypopneas, and divide by the hours of sleep:
An apnea is a pause in breathing. A hypopnea is a partial drop in airflow that has to come with something else before it counts. The 2017 AASM diagnostic guideline describes the rule from the scoring manual it cites, version 2.3: the recommended definition ties the flow drop to a 3% oxygen desaturation or a cortical arousal, and an alternative definition requires a 4% desaturation and ignores arousals (Kapur et al., Journal of Clinical Sleep Medicine, 2017). The guideline adds that, depending on which definition is used, the AHI may be considerably different in the same person. The current edition is the AASM Manual for the Scoring of Sleep and Associated Events, version 3, released in February 2023. It is a paid manual, and its rule text was not opened for this page.
A worked example, using made-up numbers: 60 scored events over 6 hours of sleep is an AHI of 10.
AHI severity: mild, moderate, severe
The 2009 AASM guideline for adult obstructive sleep apnea sets the severity bands (Epstein et al., Journal of Clinical Sleep Medicine, 2009):
Mild: 5 or more, and under 15 events per hour.
Moderate: 15 or more, up to 30 events per hour.
Severe: more than 30 events per hour.
That guideline writes the bands with the respiratory disturbance index (RDI). The 2017 guideline explains that the RDI adds respiratory effort-related arousals to the apneas and hypopneas in the AHI. It graded diagnostic accuracy at the same three AHI cut points: 5, 15, and 30.
The number alone does not make the diagnosis. As the 2017 guideline summarizes ICSD-3, obstructive sleep apnea means an obstructive RDI of 5 or more per hour with typical symptoms, such as unrefreshing sleep, daytime sleepiness, waking up gasping or choking, loud snoring, or witnessed apneas, or 15 or more per hour without symptoms.
AHI on a lab study vs REI on a home test
Polysomnography records EEG, eye movements, and muscle tone, so the lab knows when you were asleep. A home sleep apnea test usually does not. The 2017 guideline says that is why a home test divides by recording time, and why its result is called the respiratory event index, or REI. A home device with conventional sensors also cannot see hypopneas that are marked only by an arousal, and a sensor can slip off. Each of those can push the REI below the AHI a lab would have scored. The 2009 guideline makes the same point about portable monitors: they are likely to underestimate severity compared with the AHI from polysomnography.
Back to the made-up night: 60 events over 6 hours of sleep is an AHI of 10. If the recording ran 7.5 hours, the same 60 events give an REI of 8.
The 2017 guideline names polysomnography as the standard diagnostic test. It recommends either polysomnography or a technically adequate home test for uncomplicated adults whose signs and symptoms point to an increased risk of moderate to severe sleep apnea, and polysomnography when a single home test is negative, inconclusive, or technically inadequate. The home sleep apnea test vs polysomnography guide covers who gets which test.
Why AHI changes between nights
One night is a sample. In the studies the 2017 guideline reviewed, 26% to 35% of people changed severity category between two nights, in either direction, and most moved by one category, such as mild to moderate. The hypopnea rule, the denominator, and the night itself all move the number.
What wearables estimate
AHI is scored on a sleep study; consumer wearables estimate related signals and are not a diagnosis. A watch or ring can track oxygen dips, pulse rate changes, and breathing-related changes in the pulse wave, then a model turns those into an AHI-like estimate. The 2017 guideline also recommends against diagnosing sleep apnea from questionnaires or prediction algorithms without polysomnography or a home test. How those estimates are built and validated is the subject of wearable AHI estimation. Published results for AHI estimation sit on the sleep apnea AHI leaderboard. The condition overview is sleep apnea, and Sensor Bio keeps a review of wearable sleep apnea studies.
Central apneas count toward AHI too; central sleep apnea explains how scorers separate them from obstructive events.
The oxygen desaturation index, ODI, counts oxygen drops rather than breathing events. It is related to AHI and is not the same number; overnight oximetry vs a sleep study compares them. For what blood oxygen usually does through the night, see normal SpO2 while sleeping.
This page explains the number. It is not a diagnosis and not treatment advice. A sleep clinician reads AHI with the rest of the study and your symptoms.
Frequently Asked Questions
What is AHI in sleep apnea?
AHI, the apnea-hypopnea index, is the number of apneas plus hypopneas per hour of sleep, scored from a sleep study. Sixty events over 6 hours of sleep is an AHI of 10. A hypopnea only counts when the drop in airflow comes with an oxygen desaturation or, under the recommended rule, an arousal (Kapur et al. 2017, doi:10.5664/jcsm.6506).
What AHI is mild, moderate, or severe?
In the 2009 AASM adult guideline (Epstein et al., doi:10.5664/jcsm.27497), mild is 5 to under 15 events per hour, moderate is 15 to 30, and severe is more than 30. That guideline writes the bands with the RDI, which also counts respiratory effort-related arousals.
Is an AHI under 5 normal?
Under 5 is below the lowest cut point in the AASM guidance. A diagnosis also weighs symptoms: ICSD-3, as summarized in the 2017 AASM guideline, uses 5 or more events per hour with typical symptoms, or 15 or more without them. The hypopnea rule used can also change the number, so a clinician reads it with the full report.
What is the difference between AHI and REI?
AHI divides events by hours of sleep on polysomnography. REI is the home sleep apnea test version, which divides by hours of recording because most home tests do not record sleep. Home tests also miss hypopneas marked only by an arousal. The 2017 AASM guideline says those factors can make the REI underestimate the true AHI.
Can a smartwatch measure AHI?
No. A watch or ring estimates related signals, such as oxygen dips and pulse changes, and some models turn them into an AHI-like number. AHI itself is scored on a sleep study, and a wearable estimate is not a diagnosis. Methods and validation are covered at /blog/wearable-ahi-estimation.
Why did my AHI change from one night to the next?
Nights differ. In studies reviewed by the 2017 AASM guideline, 26% to 35% of people changed severity category between two nights, mostly by one step. A different hypopnea rule or a home test's recording-time denominator can also change the number.