In-Hospital Glycemic Variability and Mean Blood Glucose as Risk Markers for Long-Term Mortality in Patients with Diabetes

<b>Background/Objectives</b>: This study aimed to evaluate the influence of mean blood glucose (MG) and glycemic variability (GV) during hospitalization on the risk of mortality in patients with diabetes mellitus (DM). <b>Methods</b>: We conducted a retrospective cohort study including patients with DM admitted to the Internal Medicine ward. The dependent variable was post-discharge mortality. Capillary glucose levels were collected, and MG, standard deviation (SD), and coefficient of variation (CV = SD/MG) were calculated. The predictive value of MG and GV, expressed as CV, for mortality was assessed, adjusting for hypoglycemic episodes and comorbidities. Survival analysis and multivariate Cox regression were performed. <b>Results</b>: A total of 276 patients were included, mean age of 77.6 ± 10.2 years, 146 (52.9%) were males. Heart failure was the leading cause of admission (40.4%). During a median follow-up of 2.7 years, 249 patients (90.2%) died, corresponding to 212 deaths per 1000 patient years (289 with CV > 0.29 vs. 168 with CV ≤ 0.29). In the multivariate Cox model, CV > 0.29 (HR = 1.60; 95% CI 1.23-2.08; <i>p</i> = 0.001) and MG > 140 mg/dL (HR = 1.71; 95% CI 1.16-2.51; <i>p</i> = 0.004) were independent predictors of mortality. <b>Conclusions</b>: Elevated MG and GV measured during hospitalization may help stratify mortality risk after discharge in patients with DM.