Comorbidity phenotypes and the incidence and severity of acute kidney injury in older critically ill patients: a retrospective cohort study

The impact of specific comorbidity combinations on the risk and severity of acute kidney injury (AKI) in older critically ill patients remains unclear. This study aimed to identify comorbidity phenotypes using latent class analysis (LCA) and quantify their associations with hospital-acquired AKI. We conducted a retrospective study using the MIMIC-IV database on first-time ICU patients (> 65 years) with a length of stay ≥ 48 h. Patients with baseline chronic kidney disease (CKD) or early-onset AKI were excluded. LCA was applied to baseline chronic conditions to identify distinct comorbidity phenotypes. Subsequently, we employed multivariable Cox proportional hazards models to evaluate AKI incidence, and logistic regression to assess AKI severity (KDIGO Stage ≥ 2), adjusting for potential confounders. Of 5,721 eligible patients, 2,202 (38.5%) developed hospital-acquired AKI. LCA identified five phenotypes: Class 1 (Low Comorbidity), Class 2 (Respiratory & Nicotine), Class 3 (Isolated Metabolic & Atherosclerosis), Class 4 (Atrial Fibrillation & Complex Cardiac), and Class 5 (Ischemic Heart Failure). Compared to Class 1, Class 4 showed the highest AKI risk (HR 1.52, 95% CI 1.29–1.79, P < 0.001), followed by Class 5 (HR 1.41, 95% CI 1.20–1.65, P < 0.001). Class 2 (HR 1.34) and Class 3 (HR 1.14) also had elevated risks. Notably, secondary analysis revealed that Class 4 (OR 1.87) and Class 5 (OR 1.51) were strongly associated with progression to moderate-to-severe AKI, whereas Class 3 was not. Furthermore, in a Fine-Gray competing-risk model treating in-hospital death before AKI occurrence as the competing event, these associations remained generally consistent. Distinct comorbidity phenotypes significantly influence the incidence and severity of hospital-acquired AKI. Phenotypes characterized by Atrial Fibrillation and Ischemic Heart Failure represent the highest-risk groups for severe renal outcomes, highlighting the potential value of targeted renal surveillance and individualized hemodynamic management in older critically ill patients.