Longitudinal lung function changes in adult eosinophilic and non-eosinophilic asthma

<bold>Introduction:</bold> Eosinophilic asthma is a well-defined asthma phenotype. Despite characterization in severe asthmatic cohorts, there is paucity of longitudinal data on eosinophilic (Ea) and non-eosinophilic (NEa) asthma among general populations. <bold>Objective:</bold> To determine spirometry changes over time in Ea and NEa asthmatics defined at baseline. <bold>Methods:</bold> Data from the observational Austrian LEAD cohort were analysed at two time-points (first visit (V1) 2011-2016, median follow-up 4.1 yrs). 6734 adults >25 yrs with current doctor’s diagnosis asthma were included. Ea had blood eosinophilic counts >= 150/ul with inhaled corticosteroid therapy (ICS) or >= 300/ul without ICS; and NEa had counts below these cut-offs. <bold>Results:</bold> Age, height and sex between the groups were not different. Ea (1.69% prevalence) had consistently lower FEV1, FEV1/FVC %pred GLI, higher bronchodilator (BD) response, lower peripheral expiratory flow (FEF25-75%) %pred GLI (61.9 vs 72.8) in V1 which remained constant in visit 2 (V2); 62.2 vs 70.1 respectively; p<0.05), but lower mean FEV1 decline (−5.7 mL/yr) compared to NEa (2.67% prevalence; -23.6 mL/yr). Airway obstruction, defined by FEV1% and FEV1/FVC %pred GLI post-BD<LLN was observed in 20.2% in Ea and 6.1% in NEa at V1 (p<0.05) and 13.5% and 7.4% at V2 (p-=0.10). Loss of obstructive lung function at follow-up is associated with the use of ICS at V1 (OR 5.1, p<0.05), while not taking ICS at V1 is associated with FEV1 decline (Wald(1)= 7.3, p<0.05). <bold>Conclusion:</bold> FEF25-75% values could be used to identify Ea asthmatics without airway obstruction. Early treatment with ICS, especially in Ea asthmatics, can indeed improve lung function of asthmatics.