The Apnea-Hypopnea Index (AHI) is considered the gold standard for diagnosis and severity classification of pediatric obstructive sleep apnea (OSA). This review critically analyzes the scientific foundations of this paradigm and its clinical implications.

A systematic literature review in the PubMed and Cochrane databases was conducted according to the SWiM methodology (Synthesis Without Meta-analysis) following EQUATOR network guidelines. Studies were grouped according to 5 themes: methodological foundations of the AHI, clinical correlations, alternative parameters (oximetry and respiratory effort), sleep study indications, natural history and therapeutic options. The narrative review focused on randomized studies (CHAT, PATS) and international guidelines.

The definition of OSA in terms of an AHI threshold of 1 event/hour is based on old statistical data (Marcus et al., 1992) without clinical validation. The PATS study found no significant clinical difference between primary snorers (AHI < 1) and mild OSA (AHI 1–5). The correlations between AHI and sleep questionnaires (Pediatric Sleep Questionnaire–Sleep-Related Breathing Disorder Scale [PSQ-SRBD], Obstructive Sleep Apnea-18 [OSA-18]) are very weak ( r ² ≈ 2%). Oximetric parameters (oxygen desaturation index (ODI), pulse oximetry SpO ₂ nadir, hypoxic burden) correlate better with cardiovascular morbidity, although pediatric thresholds remain to be validated. Markers of respiratory effort (flow limitation, pulse transit time, mandibular movement) may identify airway resistance syndrome missed by AHI alone, but are without reference values. Sleep study, with AHI assessment, is not always contributive: in typical clinical presentations without comorbidity, adenotonsillectomy may be proposed upfront.

AHI is useful but insufficient to guide therapeutic management. A personalized multiparametric approach integrating oximetric data, respiratory effort, symptoms and validated questionnaires seems more appropriate. Sleep study indications should be reserved for clinical discordance, comorbidity, young children and failure of adenotonsillectomy.