Glycemic Variability and Diabetic Events in Hospitalized Patients With COVID-19 Treated With Dexamethasone: An Observational Cohort Study

Although dexamethasone reduces mortality in patients with COVID-19, plasma glucose (PG) levels increase upon initiation. In a multicenter observational cohort of 530 adults, we estimated glycemic variability based on baseline HbA1c among patients with normoglycemia (N = 238), prediabetes (N = 159), unknown (N = 63), and known diabetes (N = 159). Glycemic variability, diabetic- and hyperglycemic events (≥ 11.1 and ≥ 16 mmol/L) were analyzed using a linear mixed model and competing risks analysis adjusted for confounders. Before dexamethasone, mean PG levels were similar in those with normoglycemia (6.5 mmol/L) and prediabetes (6.6 mmol/L), but higher in unknown (8.5 mmol/L) and known diabetes (9.9 mmol/L). After treatment, PG increased across all groups. Prediabetes showed a larger increase (1.5 mmol/L) than normoglycemia (0.7 mmol/L, p = 0.002), and known diabetes had the highest increase (2.4 mmol/L, p < 0.001), reaching an average of 12.6 mmol/L. All groups except prediabetes returned to baseline after dexamethasone. The cumulative incidence of diabetic events was 98% in known diabetes, 67% in unknown diabetes, 31% in prediabetes, and 8% in normoglycemia, with significant differences between groups (p < 0.001). We conclude that dexamethasone treatment increased average PG and caused frequent hyperglycemic events in patients with prediabetes, unknown, and known diabetes, while persistent PG elevation post-treatment occurred in prediabetes.