Abstract
BACKGROUND: Delirium frequently occurs in individuals with acute kidney injury (AKI), leading to serious adverse outcomes. However, there are currently no predictors of early intervention for delirium in clinical practice. This study aims to investigate whether a correlation exists between TyG indices and the clinical symptoms of delirium in patients with AKI. METHODS: Eligible participants diagnosed with AKI from the Medical Information Mart for Intensive Care-IV (MIMIC-IV) database were categorised based on their TyG index. The primary outcome of this study was the incidence of delirium. The TyG indices were quartile and Kaplan-Meier (K-M) cumulative curve was conducted to compare the consequence of each group. Cox proportional hazards and restricted cubic spline (RCS) analyses were employed to explore the associations between TyG indices and outcomes. To mitigate potential biases, a no-replacement propensity score matching (PSM) approach was employed. Subgroup analyses were conducted to explore differences across various demographic and clinical categories. RESULTS: A positive correlation between the quartile groupings of TyG-AVG and an increased cumulative incidence of delirium in individuals with severe AKI, as demonstrated through K-M cumulative curves and Cox regression analysis. Regarding the TyG index, patients in the 4th group displayed the highest hazard of delirium in both of the methods mentioned above. Furthermore, RCS analysis indicated that the interaction between the two variables is approximately linear. Subgroup analyses revealed that the effects of both metrics remained consistent across most examined subgroups. CONCLUSION: Higher TyG indices were clearly associated with the incidence of delirium in patients with severe AKI. These indices could serve as valuable tools for identifying delirium-prone individuals with AKI.