Apnea-Hypopnea Index (AHI): What the Numbers Mean

The Apnea-Hypopnea Index is a count of how many times per hour your breathing partially or fully stops while you sleep, and it is the primary number used to diagnose obstructive sleep apnea and gauge its severity. The standard categories are straightforward: fewer than 5 events per hour is considered normal, 5 to 14 is mild, 15 to 29 is moderate, and 30 or above is severe. But the number on your sleep study report carries more nuance than those clean cutoffs suggest, and understanding that nuance matters for treatment decisions and for knowing how seriously to take your results.

What the Standard Categories Mean

Each “event” counted in the AHI is either an apnea (a complete or near-complete pause in breathing lasting at least 10 seconds) or a hypopnea (a partial reduction in airflow that leads to a drop in blood oxygen or a brief awakening). Your AHI is the total number of these events divided by the hours you were asleep. An AHI of 22 means your breathing was disrupted about 22 times every hour throughout the night.

Treatment decisions often hinge on which category you fall into. Mild cases might be addressed with weight loss, positional changes, or a dental appliance. Moderate and severe cases more commonly call for continuous positive airway pressure (CPAP) therapy or, in some situations, surgery. Because a jump from 14 to 15 can shift you from “mild” to “moderate” and change what your insurance covers or your doctor recommends, it is worth understanding that these cutoffs are somewhat arbitrary lines drawn on a continuous spectrum of severity.

Why Your Number Depends on How It Was Measured

Not all AHI numbers are created equal. Two different labs scoring the same night of sleep can produce different results depending on which definition of “hypopnea” they use. The American Academy of Sleep Medicine recommends counting a hypopnea when airflow drops by at least 30% and is accompanied by either a 3% drop in oxygen saturation or a brief arousal from sleep. But the Centers for Medicare and Medicaid Services uses a stricter threshold, requiring a 4% oxygen drop and not counting arousals at all.¹ The practical difference is real: switching from the stricter 4% rule to the 3% rule raises the median AHI by roughly 5 to 6 events per hour, and in one study that shift moved the proportion of patients classified as moderate or severe from about 29% to nearly 74%.²

If you have had more than one sleep study, or if your doctor mentions that your AHI was scored using the “4% rule” versus the “3% rule,” that distinction alone can explain why the numbers look different. It does not mean your sleep apnea got better or worse between studies.

Lab Studies Versus Home Tests

Where you take the test also matters. An in-laboratory polysomnography (PSG) measures your actual total sleep time, because a technician monitors your brain waves and can identify exactly when you were asleep versus lying awake. Most home sleep apnea tests lack that capability. Instead, they divide the total number of events by the total recording time, which includes any period you spent awake in bed staring at the ceiling. That inflates the denominator and can make your index look lower than it would in a lab setting.³ If you spent two of your eight hours in bed awake, your home test effectively diluted your score by 25%. For someone sitting right around a diagnostic cutoff, that dilution could mean the difference between a diagnosis and a “normal” result.

Night-to-Night Variability

Your AHI can also shift substantially from one night to the next, even under identical testing conditions. A large study of over 16,000 people found that more than a third had AHI fluctuations large enough to cross a diagnostic category boundary between two consecutive nights of home testing.⁴ A separate multi-night study confirmed substantial misclassification from single-night results: roughly 20% of people whose first night registered as mild or moderate sleep apnea were reclassified after additional nights of monitoring.⁵ If your AHI came back as 13 and you feel terrible, or as 16 and you feel fine, the number from a single night may simply not have captured your typical picture. Alcohol, sleeping position, nasal congestion, and how much REM sleep you happened to get that night all feed into this variability.

What the Numbers Mean for Your Health

Even a mildly elevated AHI is associated with measurable health effects. A population-based study found a dose-response link between AHI and high blood pressure: after controlling for obesity, age, and sex, people with just 5 events per hour had roughly double the odds of having hypertension compared with those who had no sleep apnea, and people with 25 events per hour had about five times the odds.⁶ A more recent study found that this hypertension risk begins increasing from as few as 5 events per hour and continues to climb, which raises questions about whether the traditional threshold of 5 is really the point where “normal” ends and “abnormal” begins.⁷

Beyond blood pressure, sleep apnea affects thinking and memory. People with obstructive sleep apnea consistently perform worse than healthy controls on tests of attention, memory, and executive functions like planning and decision-making, and the degree of impairment tracks with the severity of the condition.⁸ The daytime fog, forgetfulness, and difficulty concentrating that many people attribute to “just not sleeping well” may be directly tied to the repetitive oxygen drops and sleep fragmentation that the AHI is measuring.

What the AHI Misses

The single biggest criticism of the AHI is that it treats all events as equal. A 15-second complete breathing pause that drops your oxygen by 12% counts the same as a 10-second partial reduction that drops it by 3%. A person with 30 short, mild hypopneas per hour gets the same AHI as someone with 30 prolonged, deep apneas, but the physiological toll of the two situations is very different.⁹ The AHI also ignores the body position during events and says nothing about how long each event lasted.

This has pushed researchers toward metrics that better capture the oxygen damage each event inflicts. One measure, called hypoxic burden, estimates the total area of oxygen desaturation associated with breathing events over the course of a night, essentially quantifying how much oxygen your body lost and for how long. Studies have found that hypoxic burden predicts cardiovascular death more accurately than AHI does.¹⁰ In two large cohort studies, people in the highest quintile of hypoxic burden had roughly two and a half times the risk of dying from cardiovascular disease, and that risk remained significant even after adjusting for AHI itself.¹¹ In plain terms, two people with the same AHI can face very different cardiovascular risks depending on how deeply their oxygen drops during each event. Hypoxic burden is not yet part of standard clinical reporting at most sleep labs, but the research behind it is growing rapidly.

Another metric you might encounter is the Respiratory Disturbance Index (RDI), which includes not only apneas and hypopneas but also subtler breathing events called respiratory effort-related arousals, where your body struggles hard enough to wake briefly even though airflow did not drop enough to qualify as a hypopnea. In about a third of sleep apnea patients, the RDI is meaningfully higher than the AHI because those extra events are being counted.¹² If your doctor uses RDI rather than AHI, your number will be the same or higher, never lower.

Position and Sleep Stage Matter More Than You Might Think

Many people have far more breathing events while sleeping on their back than on their side. The supine position narrows the airway due to gravity pulling the tongue and soft tissues backward, and it also worsens the consequences of each event: the oxygen drops tend to be deeper, the heart rate swings are larger, and the arousals are longer compared with side sleeping.¹³ For people whose apnea is strongly position-dependent, simply avoiding the supine position can cut the overall AHI roughly in half. One study of adults using positional therapy found that median AHI dropped from about 14.5 to about 6.¹⁴

Sleep stage matters too. During REM sleep, the deepest stage associated with dreaming, muscle tone drops even further, and the body’s drive to correct low oxygen weakens. The result is that apneas and hypopneas during REM tend to be longer and cause more severe oxygen desaturation.¹⁵ Some people have an AHI that looks mild overall but have severe events packed into the roughly 20-25% of the night spent in REM. Among people who already had cardiovascular disease, severe REM-specific sleep apnea was linked to roughly two and a half times the risk of a cardiovascular event compared to those without significant REM apnea.¹⁶ A standard AHI averages events across all sleep stages and all positions, which can mask a serious problem happening in just one portion of the night. If your sleep study report includes a breakdown by sleep stage or position, those numbers may be more informative than the overall AHI alone.

Sex Differences and Pediatric AHI

Women tend to have lower AHI values than men at any given level of symptoms. Part of this is due to differences in airway anatomy and where the body stores fat, but the more surprising finding is that women become symptomatic at much lower AHI thresholds. Research has shown that women with an AHI of only 2 to 5 events per hour report a similar symptom burden to men with an AHI above 15.¹⁷ Women also tend to have shorter apneas and more of the subtler upper airway resistance episodes that may not get counted in the AHI at all, and the health consequences for a given degree of severity appear to be at least as bad as in men, if not worse.¹⁸ The upshot is that a woman with an AHI of 8 and persistent fatigue should not be reassured simply because her number looks “mild.” The cutoffs were largely developed from studies of middle-aged men, and they may undercount the problem in women.

In children, the standard adult thresholds do not apply. An AHI of 1 event per hour is already considered abnormal in a child, and the clinical picture is different: instead of sleepiness, children with untreated sleep apnea often show hyperactivity, behavioral problems, and poor school performance. Yet the AHI has the same limitations in children as in adults, struggling to capture the full impact of the disease.¹⁹ Pediatric sleep apnea remains heavily underdiagnosed partly because the heterogeneity of symptoms in children is poorly represented by a single number.

What Your CPAP Machine’s AHI Is Really Telling You

If you use a CPAP or similar positive airway pressure device, it reports a nightly AHI on its display or companion app. Most sleep doctors look for a treatment AHI under 5, ideally under 2. But the number your machine gives you is not the same as the number from a formal sleep study. CPAP devices use their own algorithms to estimate events, and those algorithms have known biases.

One study found that the machine-reported hypopnea count was about double what manual scoring by a sleep technician would have found, inflating the overall AHI by roughly 50%.²⁰ This overestimation mainly comes from the device counting flow-limitation events as hypopneas even when they do not cause an oxygen drop. On the other hand, a separate real-world study found that roughly 28% of patients whose CPAP reported an AHI under 5 actually had a residual AHI above 5 on formal testing, meaning the machine was underestimating their remaining disease.²¹

So CPAP-reported AHI can err in both directions. The trend over time is usually more useful than any single night’s number. If your device consistently shows an AHI of 1 to 2 and you feel rested, that is a good sign. If the number keeps creeping up above 5 or you remain sleepy despite using the machine every night, it is worth a conversation with your sleep specialist about whether the pressure settings need adjusting or whether a follow-up sleep study on the device is warranted. The agreement between CPAP algorithms and formal polysomnography is strongest for detecting frank apneas and weakest for counting hypopneas, so a device that reports many hypopneas may be more prone to error than one reporting many apneas.²²

Oral Appliances and AHI Reduction

CPAP is the most effective treatment for bringing AHI down, but it is not the only option, and many people struggle with the mask. Mandibular advancement devices, custom dental appliances that hold the lower jaw slightly forward to keep the airway open, provide a meaningful alternative for mild to moderate cases and sometimes for severe disease in people who cannot tolerate CPAP. Across randomized controlled trials, these appliances reduced AHI by an average of about 55%, with individual reductions ranging from roughly 28% to 80% depending on the starting severity and the device used.²³ The remaining AHI after treatment varied widely, from as low as 4.5 events per hour in milder cases to 34 in more severe ones. Unlike CPAP, which can drive AHI to nearly zero when used properly, oral appliances typically reduce the number rather than eliminate it. For people in the mild to moderate range, getting from 18 down to 8 can mean fewer symptoms and lower cardiovascular risk, even though the number is not technically “normal.”

Obstructive Versus Central Sleep Apnea

Not all events on a sleep study are the same type. Obstructive apneas happen when the airway physically collapses despite continued effort to breathe. Central apneas happen when the brain temporarily fails to send the signal to breathe at all, meaning the airway is open but the breathing muscles are idle. The AHI lumps both types together, but the distinction matters because the treatments differ. CPAP helps obstructive events by physically splinting the airway open, but it may not resolve central apneas, which sometimes require a different type of device or medical workup for underlying causes like heart failure.

Standard polysomnography can distinguish between the two types because it monitors chest and abdominal effort alongside airflow. Most home sleep tests and CPAP devices have limited ability to tell obstructive events from central ones, though newer wearable devices are attempting to close that gap using pulse wave and respiratory movement analysis.²⁴ If your AHI is elevated and you are lean, do not snore loudly, or have heart disease, your doctor may want to confirm whether the events are actually obstructive before starting standard treatment.

When Symptoms and Numbers Disagree

One of the most frustrating aspects of the AHI is that it does not always match how you feel. Some people with an AHI of 40 report only mild daytime sleepiness, while others with an AHI of 8 can barely function. The underlying causes of obstructive sleep apnea vary between individuals, and the relative contribution of airway anatomy, muscle tone, arousal threshold, and respiratory control instability shifts both from person to person and across severity levels.²⁵ A person whose events are mostly brief hypopneas that barely register on the oxygen monitor may tolerate a high AHI relatively well. A person whose events cluster during REM and cause deep desaturations may suffer greatly with a technically lower AHI.

This disconnect is part of why clinicians increasingly consider symptoms, oxygen levels, and cardiovascular risk alongside AHI rather than relying on the number alone. If you feel your symptoms do not match what the number suggests, that is not in your head. It reflects a real limitation of condensing a complex, hours-long physiological process into a single average.