How Sleep Apnea Is Diagnosed: What the Research Describes

Sleep apnea is a medical condition that requires clinical diagnosis. This content is educational and describes the research on how diagnostic procedures work. It is not medical advice. If you suspect sleep apnea — loud snoring, witnessed pauses in breathing, daytime fatigue despite adequate sleep duration — please consult a sleep medicine specialist or your primary care physician.

Why this matters

Sleep apnea is one of the most under-diagnosed serious conditions in adult populations. Researchers estimate that 80-90% of moderate-to-severe sleep apnea cases in adults are undiagnosed [Punjabi 2008; Peppard et al. 2013]. The condition is associated with elevated cardiovascular risk, metabolic dysfunction, daytime impairment, and higher all-cause mortality. The under-diagnosis rate matters because untreated sleep apnea has measurable health consequences over years.

This page describes how researchers and clinicians actually diagnose sleep apnea — what gets measured, why home tests miss most cases, what the diagnostic thresholds mean, and what readers should know if they suspect they or a partner have the condition.

It is not a guide to self-diagnose. The whole point of the diagnostic process is that the condition requires clinical evaluation.

What sleep apnea is

Sleep apnea is a sleep-related breathing disorder characterised by repeated interruptions of breathing during sleep. The three subtypes researchers describe:

• Obstructive sleep apnea (OSA) — the most common form. Caused by upper-airway collapse during sleep. The brain continues to signal breathing; the airway is physically blocked.
• Central sleep apnea (CSA) — the brain temporarily stops signalling the muscles that control breathing. Less common; often associated with cardiovascular conditions or specific neurological conditions.
• Complex (mixed) sleep apnea — components of both.

Each breathing interruption — an apnea — typically lasts 10-30 seconds. During severe cases, episodes can occur 30+ times per hour. Each event triggers a partial arousal from sleep that the sleeper usually doesn't remember, oxygen desaturation, and cardiovascular stress.

The cumulative effect across a night: fragmented sleep architecture, recurrent oxygen drops, sympathetic-nervous-system activation, and over time, association with hypertension, atrial fibrillation, type 2 diabetes risk, cognitive impairment, and elevated mortality [Marin et al. 2005; Yaggi et al. 2005].

The diagnostic measurements

Diagnosis requires direct measurement of breathing, oxygenation, and arousal patterns during sleep. Several metrics are central:

Apnea-Hypopnea Index (AHI)

The number of apneas (complete cessation) plus hypopneas (significant reduction in airflow) per hour of sleep. AHI is the primary diagnostic metric.

Conventional thresholds described in clinical literature:

These thresholds are diagnostic conventions used by the American Academy of Sleep Medicine (AASM) and most international sleep societies [Berry et al. 2017].

Oxygen Desaturation Index (ODI)

The number of 3% or 4% oxygen desaturations per hour. Sleep apnea routinely causes oxygen levels to drop during each event. ODI is often used alongside AHI.

Respiratory Disturbance Index (RDI)

A broader metric that includes AHI plus respiratory effort-related arousals (RERAs) — events where breathing isn't fully blocked but breathing effort increases enough to disturb sleep. Some sleep specialists prefer RDI to AHI because it captures more subtle disordered breathing.

Arousal Index

The number of brief awakenings per hour, regardless of cause. Sleep apnea fragments sleep through these arousals.

Lowest oxygen saturation

The single lowest SpO2 reading during the night. Severe sleep apnea can cause oxygen to drop into the 70s or lower during apneic events.

How diagnosis is actually performed

Polysomnography (PSG) — the gold standard

PSG is an overnight sleep study performed in a sleep lab. It measures:

• EEG (electroencephalogram) — brain electrical activity, used to score sleep stages
• EOG (electrooculogram) — eye movements, used to identify REM sleep
• EMG (electromyogram) — muscle tone, primarily chin and leg muscles
• ECG — heart rate and rhythm
• Nasal/oral airflow — direct measurement of breathing via thermistor and pressure transducer
• Respiratory effort — chest and abdomen movement via belts
• SpO2 — pulse oximetry
• Body position — most apnea is worse when sleeping supine
• Snoring sound
• Video monitoring

PSG is the comprehensive measurement standard. Studies comparing PSG against itself across repeat nights have reported good consistency for OSA diagnosis [Iber et al. 2007]. The trade-off: PSG requires an overnight stay in a sleep lab, is expensive, and has limited availability in many healthcare systems.

Home Sleep Apnea Testing (HSAT)

HSAT uses portable devices the patient takes home for one night. Most consumer-accessible HSAT devices measure airflow, respiratory effort, SpO2, and heart rate — but not EEG, EMG, or sleep stages.

The limitation: without EEG, HSAT cannot directly measure sleep itself. It records total recording time, not total sleep time. This typically results in underestimation of AHI — apneas are normalised against recording time (which includes time spent awake in bed), making the index appear lower than the actual sleep-time AHI.

Validation studies have reported:

• HSAT has reasonable sensitivity for moderate-to-severe sleep apnea (AHI > 15) in patients with a high pre-test probability — meaning patients clearly likely to have the condition based on symptoms [Collop et al. 2007]
• HSAT misses approximately 15-20% of OSA cases that PSG would identify, particularly mild cases [Rosen et al. 2018]
• HSAT cannot diagnose central sleep apnea reliably
• HSAT cannot reliably distinguish positional sleep apnea from sleep-stage-specific sleep apnea

In practice: HSAT is useful as a screening tool for high-risk patients. PSG is needed for ambiguous cases, mild presentations, suspected central sleep apnea, or when treatment isn't working as expected.

Consumer wearables

Apple Watch added a "sleep apnea notification" feature in 2024 that monitors wrist movement patterns for signs of disordered breathing. Other wearables (Withings, Garmin) have introduced similar features.

What the research describes about these features:

• They are screening tools, not diagnostic instruments
• They can flag patterns suggestive of sleep apnea over many nights
• They cannot directly measure AHI
• A flagged user should follow up with formal evaluation [Withings clinical study 2023; Apple validation summary 2024]

The right framing: consumer wearables can raise awareness for users who didn't know to investigate. They cannot substitute for clinical diagnosis.

Why home tests miss most cases

Several patterns appear in the validation literature:

1. The denominator problem. HSAT divides apnea events by total recording time; PSG divides by total sleep time. Recording time is longer than sleep time. So HSAT systematically underestimates AHI.

2. No EEG, no arousal scoring. Respiratory effort-related arousals (RERAs) require EEG to identify. HSAT can't capture them, so it misses upper airway resistance syndrome — a condition along the OSA spectrum.

3. Single-night variability. Sleep apnea severity varies night to night. A single home test on a low-AHI night can miss the diagnosis.

4. Body position effects. If the patient happens not to sleep supine during the test night, position-dependent sleep apnea may be missed.

5. Sleep efficiency assumptions. HSAT calculations typically assume the patient is asleep when in bed. Patients with insomnia (often co-occurring with sleep apnea) spend less time actually asleep.

The net effect: HSAT is more useful for ruling-in moderate-to-severe sleep apnea than for ruling-out the condition. A negative HSAT does not mean a patient does not have sleep apnea.

Symptoms researchers associate with sleep apnea

Several symptoms appear in the screening literature for sleep apnea. These do not diagnose the condition; they raise pre-test probability for clinical evaluation:

• Loud, habitual snoring (especially with witnessed pauses)
• Daytime fatigue, sleepiness, or fatigue despite adequate sleep time
• Morning headaches
• Unrefreshing sleep
• Cognitive difficulties: poor concentration, memory lapses
• Mood disturbances co-occurring with sleep symptoms
• Nighttime gasping or choking sensations
• Frequent nighttime urination (nocturia) without other obvious cause
• Treatment-resistant hypertension
• Atrial fibrillation
• Obesity (a strong risk factor, but not required — non-obese sleep apnea exists)
• Large neck circumference (>17 inches in men, >16 inches in women)
• Recessed jaw or other craniofacial features

The STOP-BANG questionnaire is a validated 8-item screening tool used in clinical settings: Snoring, Tiredness, Observed apnea, Pressure (hypertension), BMI > 35, Age > 50, Neck circumference, Gender male. A score of 3+ indicates increased pre-test probability [Chung et al. 2008].

What clinical evaluation actually looks like

The typical pathway in most healthcare systems:

1. Patient or partner notes symptoms — snoring, witnessed apneas, daytime fatigue
2. Primary care visit — symptom review, possibly STOP-BANG questionnaire, referral to sleep medicine
3. Sleep medicine consultation — clinical interview, examination of airway anatomy, decision about which test to order
4. HSAT or PSG — depending on pre-test probability and clinical complexity
5. Diagnosis and severity classification — based on AHI and clinical context
6. Treatment recommendation — most commonly CPAP for moderate-to-severe; mandibular advancement devices, surgery, weight loss, or positional therapy for specific cases

The diagnostic process is well-established and reliable when followed. The under-diagnosis problem isn't methodological; it's access and awareness.

What this page is not

This page is not a substitute for clinical evaluation. We've described how sleep apnea is diagnosed because the diagnostic process is widely misunderstood — many people believe a consumer device or a self-administered checklist constitutes diagnosis. It does not.

If symptoms suggest possible sleep apnea, the clinical pathway above is what the published research describes as appropriate. A sleep medicine specialist is the right person to assess whether testing is warranted and which test to order.

Related Proco pages

• Wearables: what they can and can't measure
• How to read a clinical trial
• The state of consumer health AI in 2026

Sources

Sleep apnea is a serious medical condition. If you suspect you or a partner may have it, consult a sleep medicine specialist for clinical evaluation. This page describes what diagnostic procedures measure — it is not a substitute for diagnosis. Proco provides educational, research-based information and does not diagnose, treat, cure, or prevent any condition.

Schema (for implementation)

{
  "@context": "https://schema.org",
  "@type": "MedicalWebPage",
  "headline": "How Sleep Apnea Is Diagnosed: What the Research Describes",
  "description": "Sleep apnea is diagnosed via polysomnography (sleep lab study) or home sleep apnea testing. This page describes what each measures, why home tests miss most mild cases, and AHI diagnostic thresholds.",
  "datePublished": "2026-06-01",
  "dateModified": "2026-05-31",
  "author": {"@type": "Organization", "name": "Proco"},
  "publisher": {"@type": "Organization", "name": "Proco", "url": "https://procohq.com"},
  "medicalAudience": "Patient",
  "lastReviewed": "2026-05-31",
  "about": {"@type": "MedicalCondition", "name": "Sleep apnea"}
}