Longitudinal asthma exacerbation phenotypes in the Korean childhood asthma study cohort

Background Asthma exacerbation (AE) leads to social and economic costs and long-term adverse outcomes. We aimed to predict exacerbation-prone asthma (EPA) in children. Methods The Korean childhood Asthma Study (KAS) is a prospective nationwide pediatric asthma cohort of children aged 5-15 years followed every 6 months. Patients with AE during the 6 months prior to all three visits, with AE prior to one or two visits, and without AE prior to any visit were defined as having EPA, exacerbation-intermittent asthma (EIA), and exacerbation-resistant asthma (ERA), respectively. Risk factors and prediction models of EPA were explored. Results Of the 497 patients who completed three visits, 42%, 18%, and 15% had exacerbations prior to visits 1, 2, and 3 and 5%, 47%, and 48% had EPA, EIA, and ERA, respectively. Univariate and multivariable logistic regression revealed forced expiratory volume in 1 s (FEV1) z-score, forced vital capacity (FVC) z-score, white blood cell (WBC) count, and asthma control test (ACT) score as relevant EPA risk factors. The EPA prediction model comprised FVC z-score, WBC count, ACT score, sex, and parental education level (area under the receiver operating characteristic curve [AUROC] 0.841 [95% confidence interval (CI): 0.728-0.954]). Conclusion With appropriate management, AE decreases over time, but persistent AEs may occur. Apart from asthma control level, baseline lung function and WBC count predicted EPA.