Acute kidney injury is associated with less reversal of urinary metabolomic changes after coronary artery bypass grafting

Abstract Introduction: Urinary metabolomics can identify differences associated with acute kidney injury (AKI), but group comparisons do not show whether postoperative changes reverse within individual patients. Objectives: We examined whether urinary profiles moved toward each patient’s preoperative profile between approximately 20 and 44 h after coronary artery bypass grafting (CABG). Methods: This secondary analysis used public data from 145 on-pump CABG patients (50 AKI and 95 non-AKI), sampled before surgery and approximately 6, 20 and 44 h afterward, with 4,048 features. Residual ratio measured the remaining distance from baseline relative to that at 20 h, reversal fidelity measured alignment toward baseline, and trajectory distance compared patients with a leave-one-patient-out cohort path; inference used patient-level permutations and bootstraps. Results: AKI was associated with higher residual ratio (medians 1.035 versus 0.980; difference 0.0548; 95% confidence interval 0.0334–0.0790) and lower reversal fidelity (0.321 versus 0.413; difference −0.0927; 95% confidence interval −0.1206 to −0.0534; both one-sided p = 1.00×10⁻⁵). Trajectory distance was greater in AKI but closely tracked overall change magnitude; adjusted evidence was inconclusive, and separate urine and tissue datasets could not test longitudinal return. Conclusion: AKI was associated with less reversal of urinary metabolomic changes from approximately 20 to 44 h after CABG. These profile changes do not establish kidney-function recovery and require independent longitudinal replication.