A62-15 Obstructive Sleep Apnea and Pulmonary Function Trajectories in Children With Post-prematurity Respiratory Disease

Rationale

Children with post-prematurity respiratory disease are at increased risk for obstructive sleep apnea (OSA) due to altered upper airway mechanics and abnormal respiratory control. OSA causes oxidative stress, cyclic intrathoracic pressure change and systemic inflammation that may exacerbate lung injury in developing airways. However, the longitudinal impact of OSA on pulmonary function trajectories in children with post-prematurity respiratory disease remains unknown. OBJECTIVE: To characterize sleep-disordered breathing prevalence and severity in children with post-prematurity respiratory disease, and investigate the impact of OSA on pulmonary function trajectories.

Methods

Retrospective cohort study of children born ≤32 weeks gestation with post-prematurity respiratory disease followed at a single center. OSA was defined as obstructive apnea-hypopnea index (oAHI) ≥1/hour using incident polysomnography; sensitivity analysis using polysomnography with highest oAHI showed consistent results. Spirometry z-scores (GLI-Global 2022) were analyzed using linear mixed-effects models testing age × OSA interactions, adjusted for gestational age, birth weight, and baseline lung function. Sensitivity analyses additionally adjusted for BPD severity, duration of mechanical ventilation, and history of tracheostomy.

Results

Of 1240 children with post-prematurity respiratory disease, 235 (mean GA 27 weeks; 54% male) underwent PSG; of these, 77 (33%) had OSA. PSG was performed at mean age 2.8±3.3 years. Among OSA patients, mean oAHI was 8.8±12.5/hr with 42% having moderate-to-severe OSA (oAHI≥5). OSA was associated with impaired gas exchange: higher % Total Sleep Time (TST) with SpO₂<90% (1.3% vs 0.2%, p < 0.001) and %TST with CO₂>50 mmHg (7.5% vs 2.2%, p = 0.014). In 46 patients with longitudinal spirometry, OSA patients had numerically higher baseline FEV₁, FVC, and FEV₁/FVC z-scores, possibly reflecting selection bias: non-OSA patients had lower birth weight, more ventilator days, and higher tracheostomy rates. Multivariable-adjusted mixed-effect primary model showed no significant age × OSA interactions (FEV₁/FVC: β = 0.06, p = 0.69). However, after adjustment for BPD severity, interaction effects suggested accelerated decline in OSA: FEV₁/FVC (β=-0.58, p = 0.13), FEV₁ (β=-1.08, p = 0.06), FVC (β=-1.26, p = 0.07). In the fully-adjusted model additionally controlling for ventilator days and tracheostomy, OSA was significantly associated with accelerated FEV₁/FVC decline (β=-0.98, 95% CI -1.68 to -0.28, p = 0.009).

Conclusions

In children with post-prematurity respiratory disease, OSA was prevalent and associated with accelerated FEV₁/FVC decline after adjustment for BPD severity. These findings suggest OSA may be a modifiable risk factor for impaired lung development. Larger studies are needed to confirm whether early OSA treatment could preserve long-term pulmonary function in this vulnerable population.

This abstract is funded by: T32 HL007901