Background: Animal studies showed that the use of metformin after myocardial infarction (MI) resulted in a protective effect on cardiac myocytes. In this study, we examined the effect of metformin in patients with diabetes mellitus (DM) on left ventricular ejection fraction (LVEF) and post-MI mortality.
Methods: We reviewed charts of patients with MI admitted to the UAMS medical center. Baseline characteristics and 12-month follow up data were collected. Patients were classified into three groups: Control group- no DM (n = 464), Metformin group- DM + MI (n = 88) and No-Metformin group- DM + MI (n = 168). First, we compared Metformin and No-Metformin groups to the Control group. Second, we performed propensity-score matching in patients with DM, and compared Metformin to No-Metformin groups.
Results: All-cause 30-day and 12-month mortality was significantly higher in the No-Metformin group compared to controls (13.5 vs 9.3% p = 0.03 at 30 days, 23.7 vs 15.9 % p = 0.03 at 12 months). However, all-cause 30-day and 12-month mortality were similar in the Controls and Metformin group (9.3 vs 6.8 % p = 0.93 at 30 days, 15.9 vs 11.4 % p = 0.97 at 12 months). Mean LVEF on presentation (45 % in the three groups) and at follow up (47.84, 46.38 and 43.62 % in Control, Metformin, and No-Metformin groups, respectively) were not statistically different. There were no significant differences in regard to re-hospitalization, re-intervention, new stroke, CHF development, new MI, or identifiable arrhythmias. Metformin was an independent predictor of lower 30-day and 12-month all-cause mortality in patients with DM (HR 0.25, p = 0.02 and HR 0.32, p = 0.01, respectively). In the matched analysis, 30-day all-cause mortality was significantly higher in the No-Metformin compared to the Metformin group (21.1 vs 8.8 %, p = 0.05). However the difference in 12-month all-cause mortality did not reach statistical significance (24.6 vs 15.8 %, p = 0.15).
Conclusion: This proof-of-concept study shows that use of metformin in patients with DM is associated with lower 30-day all-cause mortality and tendency for a lower 12-month all-cause mortality following MI without discernible improvement in LVEF.