A history of diabetes predicts outcomes following myocardial infarction: an analysis of the 28 771 patients in the High-Risk MI Database

Aims: To examine the impact of diabetes mellitus on long-term clinical outcomes in patients with myocardial infarction (MI) complicated by clinical signs of heart failure (HF) or left ventricular dysfunction (LVD). Methods and results: The High-Risk MI Database consists of individual data from 28 771 patients and was created by merging four large recent randomized clinical trials (VALIANT, EPHESUS, OPTIMAAL, and CAPRICORN) that each examined the impact of pharmacological interventions following MI in patients with evidence of HF or LVD. The mean age of patients was 65 years, 70% were male, and almost 94% Caucasian. Overall, 7368 (26%) had a history of diabetes. All the major outcomes were adjudicated by independent end-point committees. Strong and highly significant associations were found with all major clinical outcomes. Diabetes was associated with an increased risk for all-cause death [adjusted hazard ratio (HR) 1.37; confidence interval (CI) 1.28–1.46; P < 0.001]. The higher risk for all-cause death was largely mediated by higher risk for cardiovascular death (adjusted HR 1.38; CI 1.27–1.48; P < 0.001) predominantly due to a substantially increased risk for fatal re-infarction (adjusted HR 1.78; CI 1.42–2.23; P < 0.001). Additionally, diabetes was associated with an increased risk for hospitalizations, particularly HF hospitalization (adjusted HR 1.50; CI 1.39–1.63; P < 0.001). There were also elevated risks for composite outcomes, particularly death or hospitalization due to HF (adjusted HR 1.48; CI 1.38–1.59; P < 0.001). Conclusion: The risk for adverse outcomes associated with diabetes remains elevated even after debut of coronary artery disease in patients with MI complicated by clinical signs of HF or LVD. This association is particularly strong for HF-related outcomes.