Risk of pediatric obstructive sleep apnea in preadolescent children with orthodontic anomalies: a diagnostic study.

BACKGROUND: This study aimed to investigate the relationships among the apnea-hypopnea index (AHI) measured by home sleep apnea testing (HSAT) as a screening indicator of the likelihood of pediatric obstructive sleep apnea (POSA), the presence of orthodontic malocclusions, and oral breathing. METHODS: Children aged 6 to 12 years, of (nationality) or (nationality) nationality, who were referred for orthodontic examination were included in the study. Evaluated cranio-maxillofacial features included maxillary arch constriction, anomalies in the number and position of teeth, dental arch relationship, skeletal class, mandibular growth pattern, and tongue range of motion ratio. The likelihood of POSA was screened using overnight home respiratory polygraphy (HSAT), with the apnea-hypopnea index (AHI) as the quantitative primary outcome; HSAT is a screening tool and does not replace full-night polysomnography (PSG), the diagnostic gold standard. A stepwise linear regression model was developed to evaluate the relationship between clinically significant variables and apnea-hypopnea index (AHI). RESULTS: A total of 100 children participated in the study, with 61% having an AHI consistent with mild POSA and 7% with moderate POSA on HSAT screening, respectively. AHI (treated as a continuous variable) positively correlated with maxillary arch constriction, overjet, increased ANB angle, mandibular crowding, and oral breathing preference, and negatively with the tongue range of motion ratio (correlation coefficients 0.2-0.47). The final linear regression model retaining only four parameters (maxillary arch constriction, mandibular crowding, increased ANB angle, and oral breathing preference) yielded AUC of 0.72 and 0.91 for the identification of children at risk of POSA and moderate POSA, respectively. CONCLUSIONS: Our study confirmed the multifactorial nature of POSA and the elevated risk of POSA in the orthodontic population. Including craniofacial features with small effect sizes in the model improved its performance, highlighting the importance of considering multiple contributing factors rather than isolated predictors when assessing POSA risk.