BackgroundThe prognostic role of diabetes mellitus (DM) in heart failure (HF) patients is undefined, since DM is outweighed by several DM-related variables when confounders are considered. We determined the prognostic role of DM, treatment, and glycemic control in a real-life HF population. Methods3927 HF patients included in the Metabolic Exercise Cardiac Kidney Index (MECKI) score database were evaluated with a median follow-up of 3.66 years (IQR 1.70–6.67). Data analysis considered survival between DM (n = 897) vs. non-DM (n = 3030) patients, and, in diabetics, between insulin (n = 304), oral antidiabetics (n = 479), and dietary only (n = 88) treatments. The role of glycemic control was evaluated grouping DM patients according to glycated hemoglobin (HbA1c): <7% (n = 266), 7.1–8% (n = 133), >8% (n = 149). All analyses were performed also adjusting for ejection fraction, renal function, hemoglobin, sodium, exercise peak oxygen uptake, and ventilation/carbon dioxide relationship slope. Study primary endpoint was the composite of cardiovascular death, urgent heart transplantation, or left ventricular assist device implantation. Secondary endpoints were cardiovascular and all causes death. ResultsFor all endpoints, upon adjustment for confounders, DM status and insulin treatment or dietary regimen were not significantly associated with adverse long-term prognosis compared to non-DM and oral antidiabetic treated patients, respectively. A worse prognosis was observed in HbA1c >8% patients (Log-Rank p < 0.001), even after correction for confounding factors. All results were replicated by hazard ratio analysis. ConclusionIn HF patients, DM, insulin treatment and dietary regimen are not adverse outcome predictors. The only condition related to long-term prognosis, considering potential confounders, is poor glycemic control.