Prevalence and prognostic impact of rhabdomyolysis in adults hospitalized with diabetic ketosis: a 10-year single-center cohort study

Objective: To determine the prevalence and clinical impact of rhabdomyolysis (RML) among adults hospitalized with diabetic ketosis. Design: Ten-year retrospective single-center cohort study. Methods: We reviewed adults admitted with diabetic ketosis, including isolated ketosis and diabetic ketoacidosis (2015-2024). RML was defined as peak creatine kinase >1,000 U/L; patients with confirmed acute myocardial infarction were excluded. The primary endpoint was a composite poor outcome, defined as in-hospital death or discharge against medical advice (DAMA) due to critical illness with a grave prognosis. Predictors of poor outcome were assessed using Firth-corrected multivariable logistic regression. Results: Of 920 eligible patients, 18 (1.96%) developed RML. Compared with controls, patients with RML were older and more likely to have neurological comorbidities, recent falls, impaired consciousness, infection, effective serum osmolality >320 mOsm/kg, leukocytosis, higher C-reactive protein, elevated creatinine, acute kidney injury, markedly raised myoglobin, and elevated troponin I. RML was associated with a longer hospital stay, higher costs, a higher in-hospital mortality (11.1 vs 0.67%), and a higher rate of the composite poor outcome (16.7 vs 1.11%). Leukocytosis (>9.5 × 109/L; adjusted OR: 4.29, 95% CI: 1.35-14.37, P = 0.014) and troponin I > 0.03 ng/mL (adjusted OR: 5.79, 95% CI: 1.56-21.86, P = 0.009) independently predicted poor outcome. Conclusions: RML is an uncommon but important complication of diabetic ketosis, conferring a higher short-term mortality, healthcare use, and composite poor outcome. Routine screening for RML and concurrent inflammatory or cardiac injury provides a crucial window for early risk stratification, enabling clinicians to optimize therapeutic strategies and resource use for high-risk individuals.