Association between blood urea nitrogen to serum albumin ratio trajectories and mortality in patients with acute kidney injury: a retrospective cohort study

Abstract Background Although previous studies have identified the blood urea nitrogen to serum albumin ratio as a risk factor for acute kidney injury (AKI), they only evaluated baseline levels, overlooking dynamic changes during intensive care unit (ICU) hospitalization. This study aimed to examine blood urea nitrogen to serum albumin ratio trajectories, assess their relationship with mortality in AKI patients, and identify factors that might influence this association. Methods Critically ill patients with acute kidney injury were identified from the Medical Information Mart for Intensive Care (MIMIC-IV) database. The latent growth curve model was utilized to assess average blood urea nitrogen to serum albumin ratio trajectories and inter-individual variability. The group-based trajectory model identified subgroups exhibiting similar blood urea nitrogen to serum albumin ratio patterns. Cox regression analysis assessed the risk of mortality across different blood urea nitrogen to serum albumin ratio trajectory groups. Subgroup analyses were performed to validate the robustness of the results. Results Among 3013 eligible patients, latent growth curve modeling revealed significant heterogeneity in blood urea nitrogen to serum albumin ratio trajectories. Group-based trajectory modeling identified three trajectories: high level of decrease (22.4%), medium level of stability (42.4%), and low level of increase (35.2%). Multivariate Cox regression confirmed a higher mortality risk in the “high level of decrease” group at both 28 days (Hazard Ratio [HR] 2.42, 95% Confidence Interval [CI] 1.95–3.01; P < 0.001) and 360 days (HR 1.84, 95% CI 1.56–2.17; P < 0.001). Subsequently, subgroup analyses were performed for potential confounders identified by the causal directed acyclic graphs to investigate their influence on the association between blood urea nitrogen to serum albumin ratio trajectories and mortality. Conclusion Blood urea nitrogen to serum albumin ratio longitudinal trends are positively correlated with mortality. Interventions aimed at reducing blood urea nitrogen to serum albumin ratio may improve survival in acute kidney injury patients.