Clinical Predictors of Pediatric Obstructive Sleep Apnea Syndrome

Objective: To develop an adequate model using reliable clinical and physical factors to predict pediatric obstructive sleep apnea/hypopnea syndrome (OSAS). Methods: Complete anthropometric measurements including BMI z score, tonsil size grading, and updated Friedman tongue position (uFTP) were evaluated. Subjective assessments of clinical symptoms/signs, including snoring visual analog scale (VAS), nasal obstruction, and mouth breathing, were recorded. Results: Eighty-eight children (57 boys and 31 girls, mean age = 9.0 years) were confirmed to have OSAS by comprehensive polysomnography (PSG). When the aforementioned variables were analyzed individually, the results indicated that snoring VAS, nasal obstruction, mouth breathing, and BMI z score were reliable predictors of apnea/hypopnea index (AHI/h) values (correlation coefficient r = 0.386, P < .001; r = 0.416, P < .001; r = 0.255, P = .02; and r = 0.243, P = .02, respectively). When all significant factors were included in the stepwise multiple linear regression analysis, the final predictive model is: Pediatric AHI = 0.108 + 0.103 snoring VAS + 0.894 nasal obstruction + 0.207 BMI z score (F = 4.06, P = .01). Conclusion: The proposed noninvasive, simple, inexpensive, and easy to perform screening tool could be used to predict pediatric OSAS. An abnormal calculated AHI may prompt clinical physicians to conduct further PSG diagnostics and treatment.