Association of creatinine-to-albumin ratio and short-term mortality in critically ill patients with acute myocardial infarction: a threshold effect analysis

The ratio of creatinine-to-albumin (CAR) has been linked to adverse outcomes in various clinical settings. However, its relationship with mortality in critically ill patients with acute myocardial infarction (AMI) remains uncertain. The eICU Collaborative Research Database was used in this retrospective cohort study to analyze 2,726 AMI patients admitted to intensive care units. Measurements from the first 24 h post-ICU admission were used to calculate CAR. The outcome assessed was 28-day mortality in the ICU. Cox proportional hazards models, nonlinear relationship analysis, subgroup analysis, and mediation analysis were performed. Missing data was addressed using multiple imputation. The 28-day ICU mortality rate was 8.034%. In fully adjusted multivariate analysis, each unit increase in CAR correlated with a higher risk of mortality (adjusted HR = 1.945). The high CAR group exhibited a significantly increased mortality risk compared to the low CAR group. These associations remained consistent in multiple imputation analyses. Generalized Additive Model analysis indicated a nonlinear relationship between CAR and mortality, with a threshold effect at CAR = 1.089. Before this threshold, CAR showed a significant positive association with mortality (HR = 3.969, P < 0.001). Subgroup analyses demonstrated the consistency of these associations. Mediation analysis revealed that heart rate partially mediated (4.81%) the relationship between CAR and mortality. The high CAR group demonstrated a significantly decreased survival probability. Elevated CAR levels were linked to higher 28-day mortality rates in critically ill patients with AMI, showing a nonlinear relationship with a threshold effect. This association remained robust and was partially mediated by heart rate.