We attempted to describe the risk of heart failure (HF) occurrence according to diabetes mellitus (DM) status in patients with coronary artery disease (CAD) over time, from acute myocardial infarction (MI) to the chronic stable phase.

For the acute and subacute MI phases, we analysed the FAST-MI cohort restricted to patients without history of HF (n = 12,473). The analysis on 1-year outcomes after MI was further restricted to patients who were discharged alive and without history of HF and/or HF symptoms during the index hospitalisation for MI (n = 9181). To analyse the chronic phase, we analysed the CORONOR cohort restricted to patients without history of HF (n = 3871). The primary endpoint was HF occurrence according to DM status. We also analysed the composite of all-cause death or HF.

Killip-Kimball class ≥II during the index MI hospitalisation was more frequent in DM patients compared to non-DM patients (29% vs. 15.3%, adjusted OR = 1.60). At one year after MI, hospitalisation for HF was more frequent in DM patients (3.3% vs. 1.2%, adjusted HR = 1.73). At the chronic phase (5-year outcomes), hospitalisation for HF was more frequent in DM patients (8.5% vs. 4.3%, adjusted HR = 1.70). Results focusing on the composite endpoint (all-cause death or HF) were consistent.

DM was associated with a very constant near 2-fold increase in the risk of HF whatever the presentation of CAD. Avoiding the risk of HF occurrence in CAD patients with DM is critical in daily practice and should be a constant life-long endeavour.