The Apnea-Hypopnea Index, or AHI, is calculated by dividing the total number of apnea and hypopnea events during sleep by the total hours of sleep. If you had 60 events over 6 hours of sleep, your AHI would be 10, meaning you averaged 10 breathing disruptions per hour. The formula itself is simple, but what counts as an “event” and how sleep time is measured can significantly change the final number.
The AHI Formula
AHI = (total apneas + total hypopneas) ÷ total sleep time in hours. That’s it. The math is straightforward division. What makes AHI tricky isn’t the arithmetic but rather the two inputs: how events are defined and how sleep time is recorded.
During a sleep study, sensors track your breathing, oxygen levels, and brain activity throughout the night. A sleep technician (or software algorithm) reviews the data and marks each time your breathing stops or significantly drops. Those marked events become the numerator. The denominator is your total sleep time, measured in hours, not the total time you spent in bed.
What Counts as an Apnea
An apnea is a complete stop in airflow lasting at least 10 seconds. There’s no partial credit here. Your breathing has to fully cease, as measured by a nasal pressure sensor, for 10 seconds or longer to be counted. Apneas can be obstructive (your airway physically collapses), central (your brain temporarily stops sending the signal to breathe), or mixed.
What Counts as a Hypopnea
Hypopneas are partial reductions in breathing, and this is where the definition gets more complicated. The American Academy of Sleep Medicine (AASM) currently uses two accepted scoring rules, and which one your sleep lab uses will affect your AHI.
Under the recommended rule (called 1A), a hypopnea is scored when airflow drops by at least 30% from your baseline for 10 seconds or more, and that drop is paired with either a 3% or greater dip in blood oxygen or a brief awakening (arousal). Under the acceptable alternative rule (1B), the same airflow reduction is required, but only a 4% oxygen desaturation qualifies. Arousals alone don’t count under 1B.
The practical difference matters. Using the 3% rule with arousals (1A) captures more events than the stricter 4% rule (1B), so the same night of sleep can produce a higher AHI under one scoring method than the other. Sleep labs are required to specify which rule they used on your report, and this is worth checking if you’re comparing results from two different studies.
AHI Severity Ranges
Once calculated, your AHI falls into one of these categories:
- Normal: fewer than 5 events per hour
- Mild sleep apnea: 5 to fewer than 15 events per hour
- Moderate sleep apnea: 15 to 29 events per hour
- Severe sleep apnea: 30 or more events per hour
These thresholds apply to adults. In children 13 and younger, an AHI of just 1 or more is considered abnormal. For teenagers between 13 and 17, doctors may apply either pediatric or adult thresholds depending on the child’s size and developmental stage.
Why Home Test Numbers Differ
If you took a home sleep test rather than an overnight lab study, your result may be reported as an REI (Respiratory Event Index) instead of a true AHI. The calculation looks the same on the surface, but the denominator is different.
Home devices typically can’t measure brain waves, which means they can’t tell exactly when you fell asleep and woke up. Instead of dividing events by actual sleep time, they divide by total recording time or total time in bed. Since you’re always awake for some portion of the recording, the denominator is larger, which pushes the score lower. REI is always less than or equal to what your AHI would have been in a lab setting. If your home test shows borderline results, this underestimation is one reason your doctor might recommend an in-lab study.
Home tests also can’t detect arousals (brief awakenings), so hypopneas that would qualify under the recommended 1A rule in a lab may go uncounted at home. Both factors pull the number in the same direction: home tests tend to underestimate severity.
AHI vs. RDI
You may also see a number called the RDI, or Respiratory Disturbance Index, on your sleep report. RDI includes everything in the AHI plus an additional category of events called respiratory effort-related arousals (RERAs). These are episodes lasting at least 10 seconds where your breathing doesn’t drop enough to qualify as a hypopnea, but your body is working harder to breathe and it wakes you up. Because RDI captures more events, it’s always equal to or higher than the AHI from the same study.
This distinction is relevant for insurance purposes. Medicare covers CPAP therapy when your AHI or RDI is 15 or higher, or when it falls between 5 and 14 and you also have symptoms like excessive daytime sleepiness, mood changes, or conditions such as hypertension, heart disease, or a history of stroke. If your AHI is borderline, an RDI that crosses the threshold may still qualify you.
Automated vs. Manual Scoring
Most modern sleep devices, especially home tests, generate an AHI automatically using software algorithms. These automated scores are a reasonable starting point, but manual review by a trained technician consistently improves accuracy. In one study comparing a portable sleep device against full lab results, manual scoring raised the correlation with lab-derived AHI from 0.65 to 0.81 and improved the ability to correctly categorize sleep apnea severity at nearly every threshold.
The improvement was especially notable for detecting breathing problems during REM sleep, the stage where apnea tends to be worst. Automated scoring overestimated REM-related AHI by about 5 events per hour on average, while manually edited scores were much closer to the lab reference. The accuracy boost from manual review was also greater in women and older adults, two groups where automated algorithms tend to perform less reliably.
If your sleep report was generated entirely by software (common with consumer-grade devices and some home tests), treat the number as an estimate rather than a precise measurement.
A Quick Example Calculation
Say your sleep study recorded 45 apneas and 75 hypopneas over a night where you slept for 6.5 hours. Your AHI would be (45 + 75) ÷ 6.5 = 18.5 events per hour, placing you in the moderate range. If that same study were done at home and the device recorded 8 hours of total monitoring time (including the 1.5 hours you were still awake), your REI would be 120 ÷ 8 = 15, right at the border of mild and moderate.
Same night, same breathing, different numbers. That gap is why understanding the denominator matters just as much as understanding the events being counted.