Predictive correlates of poor sleep associated with increased risk of severe asthma exacerbations among children with moderate-to-severe asthma

Abstract

Background: Sleep studies (polysomnography) are a diagnostic tool used to monitor various physiological parameters during sleep to diagnose and manage sleep disorders. However, the prognostic utility of sleep measures for the prediction of childhood asthma severe exacerbation (SE) risk is unknown. Methods: Retrospective cohort analysis to identify correlates and quantify the prognostic utility of poor sleep measures for the prediction of SE risk among children with moderate or severe asthma. Results: The study cohort included 161 patients (36% female, 33% African American, mean (standard deviation [SD]) age of 10 [4] years). A higher sleep arousal index (i.e., sleep fragmentation measured as disruptions in brainwave activity) was associated with increased risk of SE among male (adjusted odds ratio [aOR]: 1.13, 95% CI: 1.04, 1.23) but not female patients (aOR: 0.97, 95% CI: 0.88, 1.07). A history of SE(s) and use of inhaled glucocorticoid plus a long-acting β2-agonists (ICS plus LABA) were associated with higher odds of SE; conversely, a history of sleep latency reducing medication was associated with lower odds of SEs (p < .05). Inclusion of these sleep-related factors in the multivariable model to predict SE had higher prognostic accuracy than a model based on history of SE(s) alone (p < .01). Conclusion: In addition to prior SE(s), elevated sleep arousal index among male children, use of ICS plus LABA, and history of untreated sleep disturbance can improve the accuracy of SE risk prognosis to inform targeted preventive interventions to reduce excess acute healthcare utilization among children with comorbid sleep problems and moderate/severe asthma.