Severity of acute respiratory distress syndrome in trial participants with a metabolic subphenotype

Introduction: Data on obesity and diabetes in the acute respiratory distress syndrome (ARDS) are mixed. Metabolic Syndrome is a heterogeneous inflammatory state, and can be identified by metabolic risk factors. Previously, a metabolic sub-phenotype, characterized by obesity, diabetes, and hypertension, was associated with COVID-19 ARDS development and mortality. We showed this sub-phenotype was associated with lower mortality in non-COVID-19 ARDS, however, it is unclear how severity is impacted. Methods: We performed a secondary analysis of individual patient-level data from seven randomized control trials in the ARDS and PETAL Networks from the Biologic Specimen and Data Repository Information Coordinating Center database. The preliminary mortality results from this study comparing the sub-phenotype to control was previously reported showing lower mortality for this sub-phenotype. Here, we studied each criterion in isolation, as an equally-weighted contributor, and also compared the outcome of severe ARDS, defined as PaO2/FiO2 ratio of less than 100 at enrollment, between cohorts. Multivariable regression models were performed. Results: Among 4,288 ARDS trial participants, 454 (10.6%) with a metabolic sub-phenotype were compared to 3,834 (89.4%) controls. Prevalence of metabolic disease was high with 2,831 (66·0%) participants carrying at least one metric of poor metabolic health (1,457-0/3 criteria, 1,398-1/3 criteria, 979-2/3 criteria, 454-3/3 criteria). The adjusted odds of dying before days 28 or 90 were progressively lower with each metabolic criterion added as compared to a non-metabolic subgroup. Interestingly, each criterion in isolation was similarly associated with improved mortality without one particular risk factor contributing more than another, a finding supported by heterogeneity testing. Lastly, 212/454 (47%) of the metabolic sub-phenotype group met criteria for severe ARDS as compared with 1529/3834 (40%) control. The sub-phenotype was significantly associated with severe ARDS (adjusted OR 1.26 (95%CI 1.026 – 1.541). Conclusions: Obesity, diabetes, and hypertension are equal contributors to a metabolic sub-phenotype of non-COVID-19 ARDS that is significantly associated with lower mortality. However, this sub-phenotype was significantly associated with severe ARDS at enrollment.