Rationale:The relationship between airway inflammation and obesity in severe asthma is poorly understood. Objectives: We sought to determine the relationship between sputum mediator profiles and the distribution of eosinophilic inflammation and obesity in people with severe asthma. Methods: Clinical parameters and eight mediators in sputum were assessed in 131 subjects with severe asthma from a single center categorized into lean, overweight, and obese groups defined by their body mass index. In an independent group of people with severe asthma (n ¼ 45) and healthy control subjects (n ¼ 19) eosinophilic inflammation was enumerated in bronchial submucosa, blood, and sputum and related to their body mass index. Measurements and Main Results: Sputum IL-5 geometric mean (95% confidence interval) (pg/ml) was elevated in the obese (1.8 [1.2-2.6]) compared with overweight (1.1 [0.8-1.3]; P ¼ 0.025) and lean (0.9 [0.6-1.2]; P ¼ 0.018) subjects with asthma and was correlated with body mass index (r ¼ 0.29; P , 0.001). There was no relationship among body mass index, the sputum cell count, or other sputum mediators. In the bronchoscopy group the submucosal eosinophil number in the subjects with asthma was correlated with body mass index (Spearman rank correlation, r s ¼ 0.38; P ¼ 0.013) and the median (interquartile range) number of submucosal eosinophils was increased in obese (19.4 [11.8-31.2]) (cells per square millimeter) versus lean subjects (8.2 [5.4-14.6]) (P ¼ 0.006). There was no significant association between sputum or peripheral blood eosinophil counts and body mass index. Conclusions: Sputum IL-5 and submucosal eosinophils, but not sputum eosinophils, are elevated in obese people with severe asthma. Whether specific antieosinophilic therapy is beneficial, or improved diet and lifestyle in obese asthma has antiinflammatory effects beyond weight reduction, requires further study.Keywords: asthma; obesity; cytokines; phenotypes; eosinophil Asthma is a common, complex inflammatory disorder affecting about 5% of adults in the general population, of which approximately 5-10% suffer from severe disease (1, 2). Severe disease is often associated with comorbidities, such as obesity, and is particularly important because these patients suffer from substantial morbidity and consume a disproportionately high amount of the overall healthcare resources spent on asthma management (3-6). Asthma, and in particular severe asthma, is a heterogeneous disease as highlighted by the different phenotypes identified using cluster analysis of clinical data (7-10) and cytokine profiles of airway samples (11-13). The key benefit of dividing a multidimensional disease, such as asthma, into distinct phenotypes is expected to be more effective treatment targeting. This has been effectively shown with the success of eosinophilic airway inflammationdirected corticosteroid (14-16) and anti-IL-5 treatment (17-19) to prevent asthma exacerbations in eosinophilic subjects with asthma. The association of obesity and asthma has been evident Support...