How AHI Differs in the Diagnosis and Treatment of Sleep Apnea

The AHI is used for the diagnosis of sleep apnea during a polysomnogram
Getty Images

For someone diagnosed with sleep apnea, the apnea-hypopnea index (AHI) becomes a familiar acronym. It is an important metric used to gauge disease severity based on diagnostic testing as well as the ultimate response to therapy.

The goal AHI in treatment may differ from the baseline scale used for diagnosis and it may be important to consider other outcomes. Review the definitions and terminology, how AHI is measured in testing and treatment, and how other considerations may be important to assess an individual’s response to interventions.

What Is AHI? 

Regardless of the type of sleep apnea (obstructive or central), the severity of the condition is determined based on the apnea-hypopnea index (AHI). The AHI is the number of obstructive apnea, hypopnea, and central apnea events noted per hour of sleep.

AHI in Diagnosis

The following ranges are used to determine the severity of the condition based on diagnostic testing:

  • Normal: AHI fewer than 5 events per hour
  • Mild: AHI 5 to 14.9 events per hour
  • Moderate: 15 to 29.9 events per hour
  • Severe: 30 events per hour and higher

It should be noted that the criteria for pediatric sleep apnea differs, with more than 1 event per hour deemed abnormal.

Confusing Terminology

The terminology used in sleep study reports may be confusing. For example, both obstructive apnea and hypopnea events occur in obstructive sleep apnea.

An obstructive apnea event means there is at least a transient 90% reduction of airflow through the upper airway (typically measured at the nose and mouth) despite continued effort to breathe. This lasts at least 10 seconds.

Hypopnea events are partial blockages of the upper airway (at least a 30% reduction of airflow) associated with an oxygen desaturation (3% or 4% drops are used) or an arousal. 

To make matters worse, there are separate rules used to assess these hypopnea events depending on medical insurance. Medicare (and those insurers who follow their rules) require hypopneas to be associated with a 4% oxygen desaturation. Other insurance may accept a 3% drop or an associated arousal.

This difference may result in a change in the diagnosis of sleep apnea among borderline cases, either affecting the severity or sometimes even whether the result is normal or abnormal.

There are also other terms that may be used interchangeably with the AHI. For example, the respiratory-disturbance index (RDI) is another term used that often includes respiratory-related arousals (RERAs). These minor breathing disturbances are associated with arousals or awakenings, but with insufficient obstruction to count as hypopneas, regardless of the insurance rules used.

Measurement Challenges

Realistically, only a small fraction of the information available from an in-center diagnostic polysomnogram (PSG) is used to generate any of these reported indices. This may present an opportunity for future refinements in testing.

Although best measured with a PSG, home sleep apnea testing can provide some of this assessment based solely on airflow, respiratory effort, and oxygen levels. This may be achieved without any traditional discernment of sleep versus wakefulness.

Recent guidance from the American Academy of Sleep Medicine recommends calling this home-based measurement an RDI, confusing matters further. This is meant to distinguish it from the AHI, based on sleep time as measured by a PSG, as the breathing disturbances noted at home are averaged per hours of recording. When wakefulness occurs, this diminishes the overall index as sleep apnea does not occur when awake, potentially affecting the diagnosis.

Does AHI Matter?

Though used for more than 50 years, there is some controversy regarding the importance of the AHI to assigning severity as well as to linking it to the associated long-term health consequences. Someone may have a high AHI, and be asymptomatic, potentially with few deleterious effects. On the other hand, someone may have a low AHI and be severely affected. 

AHI remains the primary tool for diagnostic and therapeutic decision-making despite repeated studies showing it to be inadequate in predicting clinical consequences. As noted with the definitions described above, sleep-disordered breathing is a spectrum that ranges from snoring through upper airway resistance to overt sleep apnea.

There are multiple other clinical metrics that may be useful to assess the severity and impact of sleep apnea on important outcomes. These may include:

  • Sleepiness
  • Quality of life
  • Performance
  • Medical factors (blood pressure, heart function, diabetes markers, etc.)

Unfortunately, the AHI may only partially explain the associated disease outcomes in different patients. The underlying etiology of sleep apnea (sometimes called the endotype) and ultimate clinical manifestation (known as the phenotype) in an individual may not be well described by the AHI. Genetics may further classify the disease, allowing for targeted therapy. 

Moreover, the future development of new clinical, physiological, and biomarker metrics to measure sleep apnea and associated co-morbidities may help to define treatment goals and therapy success .

Goal AHI in Treatment

It must be acknowledged that the AHI determined in the diagnosis of sleep apnea differs from that measured in the most common treatment, the use of continuous positive airway pressure (CPAP) therapy. CPAP is first-line therapy despite historically low adherence because it reliably reduces the AHI when used, and the response to other therapies is variable.

Newer CPAP devices are able to provide a surrogate measure of AHI, or events per hour, by delivering airflow and measuring resistance. This may detect persistent blockages of the airway, or even pauses in breathing that may characterize central sleep apnea.

One might wonder how accurate this measurement is as it is determined in a different way. Certainly poor mask fit and associated leak would compromise the capacity to determine this.

Assuming that it is a fairly accurate measurement, what is the goal AHI in treatment? Certainly the residual AHI should be normalized (fewer than 5 events per hour), but is less even better? Theoretically, zero would be ideal, but that rarely occurs. The unresolved AHI varies night by night based on other factors, including:

  • Nasal congestion
  • Body position
  • Amount of REM sleep
  • Alcohol consumption
  • Pressure delivered
  • Mask leak

Targeting the lowest possible AHI seems admirable, but there may be barriers to resolution. The amount of pressure needed to resolve sleep apnea when someone is on their back, in REM sleep, with a cold, and after recently drinking alcohol would undoubtedly be higher.

If this higher pressure is delivered throughout the night to cover this potential need, it may cause discomfort and intolerance, contributing to dry mouth, aerophagia, mask leak, and other problems.

Auto-titrating CPAP devices may help to address this variance, but how successful they are may depend on the range of settings prescribed and the responsiveness of the algorithm. Ideally, the device’s pressure settings should be determined by a board-certified sleep physician with refinement as needed based on the response to therapy.

Research suggests that only AHI (as measured by PSG) is a consistent predictor of all-cause mortality and cardiovascular risk, and therefore it seems like a relevant metric to gauge response to treatment. Moreover, the evidence for this association is strongest in men.

It may be more important to resolve symptoms and improve associated medical conditions rather than to target some goal AHI number. What ultimately should be tracked?

It is apparent that a more detailed understanding of how sleep apnea develops and progresses could influence clinical treatment decisions. Indeed, treatments could be individualized based on these underlying causes.

It may be possible to better understand which symptoms (and outcomes) will respond best to sleep apnea treatment (and by how much). This requires further research to elucidate these relationships, including the optimal AHI threshold values, different hypopnea definitions, and the predictors that allow for better risk stratification to guide treatment.

Finally, it should be noted that many treatments are unable to provide an ongoing measurement of AHI. For example, the use of an oral appliance may require a repeat sleep study to assess its efficacy, but this may be limited to just a single night (or, at best, several nights) of testing.

In surgery, a post-operative assessment may be performed two months following the procedure, and AHI remains the main therapeutic metric. However, research suggests it may have poor correlation to other outcomes, and additional tools based on symptoms and physiology may better assess the effects of surgical treatment.

In order to truly understand the meaning of AHI, both in terms of diagnosis and ultimately treatment, it is best to work with a sleep expert who can fully interpret its meaning in the broader clinical context. This individualized application of the metric may lead to the best possible outcome.

A Word From Verywell

This is a very complicated topic to try to understand, and one that changes as new technology is developed and new guidelines are adopted. Ultimately, more sophisticated metrics and analyses will provide optimized clinical relevance for individual patients.

State-of-the-art sleep science will rely on reinterpreting data that has been collected for decades, seeing it a new way, and moving us towards precision sleep medicine. It is a reinvention of our most trusted tools of assessment, repurposed to best serve our patients.

This is an exciting time in the field of sleep medicine. It is best to work hand-in-hand with the help of a trusted sleep professional who is updated on the latest advancements to optimize your individual response to treatment. And keep in mind the advice of one of my supervising attending physicians in the sleep center at Stanford University: Don’t let the perfect become the enemy of the good.

5 Sources
Verywell Health uses only high-quality sources, including peer-reviewed studies, to support the facts within our articles. Read our editorial process to learn more about how we fact-check and keep our content accurate, reliable, and trustworthy.
  1. Lim DC, et al. Reinventing polysomnography in the age of precision medicineSleep Med Rev. 2020 Aug;52:101313. doi:10.1016/j.smrv.2020.101313

  2. Al-Shawwa BA, et al. Defining common outcome metrics used in obstructive sleep apneaSleep Med Rev. 2008;12(6):449-61. doi:10.1016/j.smrv.2008.07.008

  3. Edwards BA, et al. More than the sum of the respiratory events: Personalized medicine approaches for obstructive sleep apneaAm J Respir Crit Care Med. 2019 Sep 15; 200(6):691-703. doi:10.1164/rccm.201901-0014TR

  4. Kendzerska T, et al. Untreated obstructive sleep apnea and the risk for serious long-term adverse outcomes: a systematic reviewSleep Med Rev. 2014;18(1):49-59. doi:10.10116/j.smrv.2013.01.003

  5. Tam S, et al. Outcome measurements in obstructive sleep apnea: beyond the apnea-hypopnea indexLaryngoscope. 2014;124(1):337-43. doi:10.1002/lary.24275

By Brandon Peters, MD
Brandon Peters, MD, is a board-certified neurologist and sleep medicine specialist.