Why the AHI Test Misses 98% of Sleep Apnea Severity in Children

Sleep-Related Breathing Disorders: Why the Standard Test Often Misses the Full Picture

A 2026 critical review from Necker-Enfants Malades hospital in Paris found that the single number doctors rely on most to diagnose sleep-related breathing disorders in children explains almost nothing about how sick those children actually are. The correlation between the apnea-hypopnea index (AHI) and validated symptom questionnaires is roughly r² ≈ 2%. In plain terms: 98% of the variation in a child’s symptoms has nothing to do with their AHI score.

That finding, published in the European Annals of Otorhinolaryngology by Dr. Vincent Couloigner, R. Luscan, and Dr. Brigitte Fauroux, is reshaping how clinicians think about sleep-related breathing disorders (SRBDs) — a family of conditions in which breathing is repeatedly disrupted during sleep. This guide covers what these disorders are, what the newest evidence says about diagnosing them, and what you can actually do about them.

What Is a Sleep-Related Breathing Disorder?

Sleep-related breathing disorders exist on a spectrum. At the mild end sits primary snoring: noisy breathing during sleep without measurable pauses in airflow or drops in blood oxygen. At the severe end sits obstructive sleep apnea (OSA), in which the upper airway partially or fully collapses during sleep, causing apneas (complete pauses) and hypopneas (shallow breaths).

The spectrum, from snoring to apnea

Between those two poles sits a poorly defined middle zone that includes upper airway resistance syndrome — a condition in which a child (or adult) must work harder and harder to breathe through a narrowed airway without technically registering apneas on a standard sleep study. These patients snore, sleep restlessly, wake unrefreshed, and can show daytime behavioral problems, yet their AHI looks “normal.”

Who is affected

In children, OSA affects an estimated 1–5% of the population, with peak prevalence between ages 2 and 8, when adenoids and tonsils are largest relative to the airway. Obesity, craniofacial differences, neuromuscular conditions, and Down syndrome all raise risk. Adults face their own version of the problem, and untreated breathing disorders feed directly into cardiovascular disease — short sleep duration has been shown to raise hypertension risk by roughly 30% in observational data, an effect explored in our piece on sleep duration and blood pressure.

Why the AHI Alone Falls Short: The 2026 Paris Review

Couloigner and colleagues performed a systematic review following SWiM (Synthesis Without Meta-analysis) methodology, grouping studies into five themes: the AHI’s methodological foundations, its clinical correlations, alternative parameters, sleep study indications, and treatment options. Their conclusions are uncomfortable for a field that has treated the AHI as an unquestioned diagnostic anchor.

The 1-event threshold rests on 1992 statistics, not clinical validation

Pediatric OSA is defined as an AHI above 1 event per hour. That threshold traces back to a 1992 study by Marcus and colleagues, which was essentially a statistical exercise — not a demonstration that children with an AHI of 1.5 are meaningfully sicker than children with an AHI of 0.8. The review found no clinical validation ever followed.

The PATS study: mild OSA is indistinguishable from “just snoring”

The randomized PATS trial compared children with primary snoring (AHI below 1) against children with mild OSA (AHI 1–5) and found no significant clinical difference between the groups. Combined with the finding that AHI explains only about 2% of variance in symptom scores on the Pediatric Sleep Questionnaire–Sleep-Related Breathing Disorder (PSQ-SRBD) scale and the OSA-18 questionnaire, the authors concluded that the AHI cannot serve as the sole guide to who needs treatment.

What correlates better: oxygen and breathing effort

According to the review, oximetric parameters — the oxygen desaturation index (ODI), the lowest recorded pulse oximetry value (SpO2 nadir), and cumulative hypoxic burden — correlate more strongly with cardiovascular morbidity than the AHI does. Caveat: validated pediatric thresholds for these measures don’t yet exist.

Markers of respiratory effort tell another important part of the story. Flow limitation, pulse transit time, and mandibular movement analysis can catch airway resistance that the AHI misses entirely. The problem is the same: no established reference values, so these remain research-grade tools rather than clinical rules.

Bruxism and Breathing: A Clue From Moscow

A separate 2026 clinical observational study by Zhou, Loktionova, and Glazachev (Sechenov First Moscow State Medical University), published in Sleep and Breathing, examined how mandibular chewing patterns relate to sleep bruxism characteristics. The connection matters for SRBD readers because teeth grinding during sleep has long been suspected of being linked to breathing disturbances — some researchers argue grinding episodes may function as an effort to reopen a collapsed airway. While the Moscow study focused on chewing-pattern associations rather than apnea outcomes, it adds to growing evidence that nighttime jaw activity is not an isolated dental issue but often a window into broader sleep physiology. If you grind your teeth loudly enough for a partner to report it, that’s a symptom worth mentioning to a physician alongside any snoring complaints.

When a Sleep Study Is Actually Needed

The Paris review ends with a proposal that could simplify care for thousands of families. For children with a typical clinical presentation — snoring, restless sleep, daytime symptoms, no significant comorbidities — the authors suggest adenotonsillectomy can be offered upfront, without a sleep study. Surgery is first-line treatment for pediatric OSA, and in straightforward cases the study adds little decision-making value.

Sleep studies remain clearly indicated when:

• Clinical picture and examination don’t match (discordance between what you see and what you hear)
• The child has comorbidities: obesity, Down syndrome, neuromuscular disease, craniofacial anomalies, or severe asthma
• The child is very young
• Adenotonsillectomy has already been performed but symptoms persist

A multiparametric approach for everyone else

For cases that do need testing, the review argues for evaluating more than a single index: overnight oximetry for desaturation burden, measures of respiratory effort, symptom questionnaires, and clinical history should be weighed together. A patient whose AHI reads 2 but who has frequent deep desaturations, heavy flow limitation, and significant daytime impairment may need more treatment than a patient with an AHI of 6 who is otherwise fine.

Actionable Steps If You Suspect a Sleep-Related Breathing Disorder

1. Record what you observe. Video or audio of sleeping breathing pauses, snoring intensity, gasping, restless sleep, and morning headaches gives clinicians more useful information than most people realize.
2. Use a validated questionnaire as a starting screen. The PSQ-SRBD and OSA-18 are not diagnostic, but they organize symptoms well enough to guide a conversation with your doctor.
3. Don’t accept the AHI as the whole answer. If you or your child has clear symptoms but a “normal” study result, ask about oxygen desaturation data, flow limitation, and respiratory effort markers — or seek a second opinion.
4. Ask about nocturnal oximetry. Home pulse oximetry is cheap, well tolerated in children, and the review suggests its parameters track real morbidity better than AHI.
5. For children with classic symptoms and no comorbidities, discuss whether surgery-first is reasonable. The evidence supports skipping the study in these cases.
6. Consider breathing training as a complement, not a cure. Slow, structured breathing practice can reduce inflammation and improve autonomic balance — see slow breathing research on breathing.monster — but it never replaces medical evaluation for confirmed apnea.
7. Follow up after treatment. Roughly 20–40% of children who undergo adenotonsillectomy have residual OSA. Persistent symptoms warrant a post-surgical sleep study, ideally one that looks beyond the AHI.

Improvements after proper treatment tend to be measurable across attention, mood, and blood pressure — the kind of changes tracked in our coverage of sleep clinic outcomes.

What the Research Still Doesn’t Settle

Honest gaps remain. Pediatric thresholds for ODI, hypoxic burden, and effort markers are unvalidated. The Paris review itself notes that respiratory-effort parameters lack reference values, meaning clinicians must interpret them case by case. And most of the strongest trial evidence (CHAT, PATS) comes from children aged 5–9, leaving younger children and adults with less direct guidance. The review’s central message is not that AHI is useless — it’s that AHI is one instrument in what should be a full orchestra.

Key Takeaways

• A sleep-related breathing disorder spans a spectrum from primary snoring to obstructive sleep apnea, with airway resistance conditions in between.
• The 2026 Necker hospital review found the AHI explains only about 2% of symptom variation in children, and the PATS trial found no clinical difference between primary snorers and mild OSA.
• Oxygen-based measures (ODI, SpO2 nadir, hypoxic burden) track cardiovascular morbidity better than AHI, though pediatric thresholds remain unvalidated.
• For typical pediatric cases without comorbidities, adenotonsillectomy can be considered without a sleep study; studies are reserved for discordance, comorbidity, young age, and post-surgical persistence.
• Respiratory-effort markers (flow limitation, pulse transit time, mandibular movement) can catch airway resistance the AHI misses.
• Sleep bruxism may signal breathing-related sleep physiology rather than a purely dental problem — report grinding alongside snoring.
• A multiparametric approach — oximetry, effort markers, symptoms, and questionnaires together — is now the evidence-based standard, not AHI alone.

Frequently Asked Questions

What counts as a sleep-related breathing disorder?

Any condition in which breathing is repeatedly disrupted during sleep, ranging from primary snoring through upper airway resistance syndrome to full obstructive sleep apnea with apneas and hypopneas.

Can my child have a normal AHI but still have a breathing problem?

Yes. The 2026 Paris review found the AHI correlates very weakly with symptoms and misses airway resistance entirely, so children with clear symptoms but a low AHI should be evaluated with oximetry and respiratory-effort measures.

Do all children with suspected OSA need a sleep study before surgery?

No. According to the review, children with typical symptoms and no comorbidities can be offered adenotonsillectomy upfront; sleep studies are reserved for unclear cases, comorbidities, very young children, and failed surgery.

Is teeth grinding at night related to sleep apnea?

Possibly. Growing evidence, including a 2026 Sechenov University study on mandibular patterns, suggests sleep bruxism is intertwined with sleep physiology and may sometimes accompany breathing disturbances, so it’s worth reporting to a physician.

This article is for informational purposes only. Consult a qualified professional for personalised advice.

---

Sources:
https://pubmed.ncbi.nlm.nih.gov/42686450/
https://pubmed.ncbi.nlm.nih.gov/42681265/
https://pubmed.ncbi.nlm.nih.gov/42677143/
https://pubmed.ncbi.nlm.nih.gov/42675482/
https://pubmed.ncbi.nlm.nih.gov/42675346/
https://pubmed.ncbi.nlm.nih.gov/42675225/
https://pubmed.ncbi.nlm.nih.gov/42673879/