Nighttime Respiratory Disturbance and Hypoxemia Associated with Impaired Linear Growth in Children with Obstructive Sleep Apnea

Objectives: Although obstructive sleep apnea (OSA) related hypoxia shows stronger associations with multiple complications than respiratory disturbance, its impact on linear growth, a major pediatric concern, remains unclear. This study investigates independent effects of hypoxia and respiratory disturbance on the height of children with OSA. Methods: This retrospective, cross-sectional study clinically evaluated children with suspicious OSA from January 2013 to June 2025. Individuals with other sleep disorders, severe neuropsychiatric disorders and physical disorders were excluded. Subjects underwent overnight polysomnography (PSG) and height measurement. OSA was defined as apnea-hypopnea index (AHI) ≥ 1 event/hour, and those with AHI < 1 event/hour served as the non-OSA reference group. AHI and percentage of total sleep time with oxygen saturation < 90% (T90%), were separately used to evaluate the severity of respiratory disturbance (mild, 1≤AHI < 5 events/h; moderate, 5≤AHI < 10 events/h; severe, AHI ≥ 10 events/h) and hypoxia (T90% quartiles). Height z-scores ≤-1 SD defined subclinical growth impairment. Linear and logistic regression models analyzed the independent associations between AHI and T90% and height and subclinical growth impairment. Nutrition, endocrine disorders, chronic illnesses, and parental height were not collected and adjusted in this study. Results: A total of 2443 children were included (348 children without OSA, 2095 children with OSA). In linear regression models, higher T90% was associated with a lower height z-score in the full sample ( B = − 0.005, 95% CI: − 0.010 to − 0.000, P = 0.033) and children with OSA ( B = − 0.005, 95% CI: − 0.010 to − 0.000, P = 0.034), whereas AHI was not (full sample: B = − 0.003, 95% CI: − 0.006 to 0.001, P = 0.111; children with OSA: B = − 0.002, 95% CI: − 0.005 to 0.002, P = 0.330).Logistic models showed higher odds of subclinical growth impairment in children with moderate (OR = 1.73, 95% CI: 1.05– 2.84) or severe OSA (OR = 1.96, 95% CI: 1.19– 3.21) and in those in the highest T90% quartile (OR = 2.17, 95% CI: 1.30– 3.60), compared with children without OSA. These associations remained stable in females and children aged 2– 9 years. Conclusions: To our knowledge, this is the first study to show that greater respiratory disturbance and nocturnal hypoxemia are associated with a higher risk of short stature in children with OSA. These findings support routine height assessment in the clinical management of pediatric OSA. Keywords: obstructive sleep apnea, children growth, height, polysomnography, apnea-hypopnea index, hypoxia