Coexistence of obesity and asthma determines a distinct respiratory metabolic phenotype

BackgroundEpidemiologic and clinical evidence supports the existence of an obesity-related asthma phenotype. No distinct pathophysiologic elements or specific biomarkers have been identified thus far, but increased oxidative stress has been reported.ObjectiveWe aimed at verifying whether metabolomics of exhaled breath condensate from obese asthmatic (OA) patients, lean asthmatic (LA) patients, and obese nonasthmatic (ONA) subjects could recognize specific and statistically validated biomarkers for a separate “asthma-obesity” respiratory metabolic phenotype, here defined as “metabotype.”MethodsTwenty-five OA patients, 30 ONA subjects, and 30 mild-to-moderate LA age-matched patients participated in a cross-sectional study. Nuclear magnetic resonance (NMR) profiles were analyzed by using partial least-squares discriminant analysis, and the results were validated with an independent patient set.ResultsFrom NMR profiles, we obtained strong regression models that distinguished OA patients from ONA subjects (quality parameters: goodness-of-fit parameter [R2] = 0.81 and goodness-of-prediction parameter [Q2] = 0.79), as well as OA patients from LA patients (R2 = 0.91 and Q2 = 0.89). The all-classes comparison (R2 = 0.86 and Q2 = 0.83) indicated that OA patients possess a respiratory metabolic profile fully divergent from those obtained in the other patient groups. We also identified specific biomarkers for between-class separation, which are independent from clinical bias. They are involved in the methane, pyruvate, and glyoxylate and dicarboxylate metabolic pathways.ConclusionsNMR-based metabolomics indicates that OA patients are characterized by a respiratory metabolic fingerprint fully different from that of patients independently affected by asthma or obesity. Such a phenotypic difference strongly suggests unique pathophysiologic pathways involved in the pathogenesis of asthma in adult obese subjects. Furthermore, the OA metabotype could define a strategy for patient stratification based on unbiased biomarkers, with important diagnostic and therapeutic implications.

157