Background: Intermittent hypoxemic events (IHEs) in preterm infants are thought to result from immature respiratory control systems. These are usually brief (seconds to minutes) but can occur up to thousands of times during an infant's hospitalization. Increased IHE frequency has been shown to correlate with morbidities such as retinopathy of prematurity and chronic lung disease but the association with neurodevelopmental outcomes has not been well-described.
Objective: To assess the association between IHEs during the first 12 weeks of life and the composite outcome of severe neurodevelopmental impairment (SNDI) or death in premature infants at a corrected age (CA) of 18-24 months.
Methods: This was a retrospective study of infants 23-32 weeks gestational age (GA) and birth weight (BW) <1250 grams (g) born after January 2013 and discharged from the NICU by December 2017. IHEs were defined as events with SpO2 ≤ 80% lasting 10 seconds to 5 minutes throughout the first 12 weeks of life. The primary composite outcome was death before discharge or SNDI, defined as at least one of the following: Bayley-3 motor, language or cognitive composite score ≤69, deafness requiring amplification, severe blindness or cerebral palsy. Mixed-effects regression models adjusting for GA were used to determine the relationship between mean daily IHE rate (IHE-RATE) per week and the primary outcome. Logistic regression models, stratified by GA group, were used to assess the association between primary outcome, IHE-RATE, and covariates.
Results: Of 418 infants who met both GA and BW criteria, 5% (20/418) died during initial hospitalization. Follow-up exams at 18-24 months corrected age (CA) were performed on 119 infants (mean GA of 27.2±2.2 weeks, mean BW of 846±237g, 52% females). SNDI occurred in 21% (25/119), with majority demonstrating evidence of severe motor or language delay (Table 1). IHE-RATE peaked between 4-6 postnatal weeks regardless of GA, with peak IHE-RATE successively lower as GA increased (Figure 1A). Plots derived from mixed-effects regression models (Fig. 1, panels B-D) showed that among infants 28-32 wk GA, those with SNDI had significantly higher IHE-RATE at 4-8 weeks age compared to those without SNDI. These differences were not observed among infants 23-27 weeks GA. In a logistic regression model controlling for significant covariates, each unit increase in IHE-RATE was associated with a 1.4% increase in odds of sNDI or death in the 28-32 week GA group (OR 1.014, 95% CI (1.006,1.02), p=0.008).
Conclusions: In this retrospective cohort, a higher mean daily IHE rate during the first 12 weeks of life was associated with increased risk of SNDI or death in preterm infants. In subgroup analysis this association held true for infants 28-32 weeks GA but not in infants 23-27 weeks GA.
