Disentangling Uric Acid and Renal Pathways in SGLT2 Inhibitor Effects After Acute Myocardial Infarction: A Retrospective Mediation Analysis

Background/Objectives: Sodium–glucose cotransporter-2 (SGLT2) inhibitors have demonstrated cardiovascular benefits beyond glycemic control, yet the specific biological pathways potentially linking SGLT2 inhibitor exposure to cardiovascular outcomes after acute myocardial infarction (AMI) remain incompletely characterized. Two biologically plausible pathways, serum uric acid (SUA) reduction and renal functional preservation, have been proposed, but not directly compared in a unified analytical framework. This study aimed to explore whether associations between SGLT2 inhibitor exposure and recurrent post-AMI outcomes may be more strongly linked to SUA reduction and to renal functional changes, using a hypothesis-generating causal mediation analysis. Methods: This retrospective observational cohort study included 142 consecutive patients hospitalized for AMI who underwent percutaneous coronary intervention (PCI) during the index hospitalization, reflecting standard-of-care management for AMI in this tertiary center. Patients were categorized by SGLT2 inhibitor exposure (n = 57) vs. controls (n = 85). Both diabetic (47.2%) and non-diabetic (52.8%) patients were included. The primary endpoint was change in SUA (ΔUA); the secondary endpoint was myocardial infarction (MI) recurrence. Causal mediation analysis with nonparametric bootstrap simulation tested both mechanistic pathways. Results: SGLT2 inhibitor therapy was associated with significant SUA reduction (ΔUA = −0.99 mg/dL vs. +0.56 mg/dL in controls; p < 0.001), consistent across diabetic and non-diabetic subgroups and independent of AMI recurrence. Each 1 mg/dL decrease in SUA was associated with lower odds of recurrent MI in the initial model (β = −0.25; p = 0.041). However, after incorporation of renal functional change, the uric acid-mediated pathway lost significance (ACME p = 0.462), whereas the renal-mediated pathway remained significant (ACME p = 0.038). Serum creatinine change emerged as the strongest independent predictor of MI recurrence (β = 2.22; p = 0.015). Conclusions: The findings are more consistent with a renal-mediated pathway than with an independent uric acid-mediated pathway in explaining the observed associations between SGLT2 inhibitor exposure and recurrent post-AMI outcomes. These hypothesis-generating results from a retrospective design warrant prospective validation.