Identification of distinct phenotypes and improving prognosis using metabolic biomarkers in COVID-19 patients

Abstract

Objective To investigate the relationship between the levels of adipokines and other endocrine biomarkers and patient outcomes in hospitalized patients with COVID-19. Methods In a prospective study that included 213 subjects with COVID-19 admitted to the intensive care unit, we measured the levels of cortisol, C-peptide, glucagon-like peptide-1, insulin, peptide YY, ghrelin, leptin, and resistin.; their contributions to patient clustering, disease severity, and predicting in-hospital mortality were analyzed. Results Cortisol, resistin, leptin, insulin, and ghrelin levels significantly differed between severity groups, as defined by the World Health Organization severity scale. Additionally, lower ghrelin and higher cortisol levels were associated with mortality. Adding biomarkers to the clinical predictors of mortality significantly improved accuracy in determining prognosis. Phenotyping of subjects based on plasma biomarker levels yielded two different phenotypes that were associated with disease severity, but not mortality. Conclusion As a single biomarker, only cortisol was independently associated with mortality; however, metabolic biomarkers could improve mortality prediction when added to clinical parameters. Metabolic biomarker phenotypes were differentially distributed according to COVID-19 severity but were not associated with mortality.