Metformin and Myocardial Injury in Patients With Diabetes and ST‐Segment Elevation Myocardial Infarction: A Propensity Score Matched Analysis

BACKGROUND Although animal studies have documented metformin's cardioprotective effects, the impact in humans remains elusive. The study objective was to explore the association between metformin and myocardial infarct size in patients with diabetes presenting with ST-segment elevation myocardial infarction. METHODS AND RESULTS Data extraction used the National Cardiovascular Data CathPCI Registry in all patients with diabetes aged >18 years presenting with ST-segment elevation myocardial infarction at 2 academic medical centers from January 2010 to December 2013. The exposure of interest was ongoing metformin use before the event. Propensity score matching was used for the metformin and nonmetformin groups on key prognostic variables. All matched pairs had acceptable D scores of <10%, confirming an efficient matching procedure. The primary outcome was myocardial infarct size, reflected by peak serum creatine kinase-myocardial band, troponin T, and hospital discharge left ventricular ejection fraction. Of all 1726 ST-segment elevation myocardial infarction cases reviewed, 493 patients had diabetes (28.5%), with 208 metformin users (42.1%) and 285 nonusers. Matched pairs analysis yielded 137 cases per group. The difference between metformin and nonmetformin groups was -18.1 ng/mL (95% CI -55.0 to 18.8; P=0.56) for total peak serum creatine kinase-myocardial band and -1.1 ng/mL (95% CI -2.8 to 0.5; P=0.41) for troponin T. Median discharge left ventricular ejection fraction in both groups was 45, and the difference between metformin and nonmetformin users was 0.7% (95% CI -2.2 to 3.6; P=0.99). CONCLUSIONS No statistically significant association of cardioprotection was found between metformin and myocardial infarct size in patients with diabetes and acute ST-segment elevation myocardial infarction.