Feed

Acute hyperglycaemia is a common COVID-19 complication linked to adverse outcomes. The combined prognostic value of cytokine activation and acute insulin resistance in non-diabetic patients remains unclear. In this prospective cohort study, we enrolled 144 hospitalised adults with RT-PCR-confirmed SARS-CoV-2 infection and no prior diabetes. We aimed to characterise metabolic-inflammatory phenotypes and evaluate their association with disease severity and post-discharge glycaemic outcomes. Patients were classified as normoglycaemic or dysglycaemic based on repeated glucose profiles. Dysglycaemic patients were further phenotyped as stress hyperglycaemia (SHG) or newly diagnosed diabetes (NOD). This classification was based on post-discharge glycaemic assessment at 3 and 6 months, distinguishing transient from persistent hyperglycaemia. Admission hyperglycaemia was associated with a consistently elevated pro-inflammatory cytokine pattern. However, cytokine concentrations were comparable between stress hyperglycaemia and newly diagnosed diabetes, indicating that inflammatory burden alone does not explain metabolic persistence. In contrast, insulin resistance (HOMA-IR) was markedly higher in the newly diagnosed diabetes phenotype. Along with admission oxygenation and key cytokine signals, this contributed to risk stratification for severe disease. In conclusion, early admission assessment of glucose and insulin resistance identifies high-risk metabolic phenotypes. This enables targeted in-hospital risk stratification and post-discharge glycaemic surveillance.