In five large-scale trials involving >40 000 patients, sodium-glucose cotransporter 2 (SGLT2) inhibitors decreased the risk of serious heart failure events by 25-40%. This effect cannot be explained by control of hyperglycaemia, since it is not observed with antidiabetic drugs with greater glucose-lowering effects. It cannot be attributed to ketogenesis, since it is not causally linked to ketone body production, and the benefit is not enhanced in patients with diabetes. The effect cannot be ascribed to a natriuretic action, since SGLT2 inhibitors decrease natriuretic peptides only modestly, and they reduce cardiovascular death, a benefit that diuretics do not possess. Although SGLT2 inhibitors increase red blood cell mass, enhanced erythropoiesis does not favourably influence the course of heart failure. By contrast, experimental studies suggest that SGLT2 inhibitors may reduce intracellular sodium, thereby preventing oxidative stress and cardiomyocyte death. Additionally, SGLT2 inhibitors induce a transcriptional paradigm that mimics nutrient and oxygen deprivation, which includes activation of adenosine monophosphate-activated protein kinase, sirtuin-1, and/or hypoxia-inducible factors-1 /2 . The interplay of these mediators stimulates autophagy, a lysosomally-mediated degradative pathway that maintains cellular homeostasis. Autophagy-mediated clearance of damaged organelles reduces inflammasome activation, thus mitigating cardiomyocyte dysfunction and coronary microvascular injury. Interestingly, the action of hypoxia-inducible factors-1 /2 to both stimulate erythropoietin and induce autophagy may explain why erythrocytosis is strongly correlated with the reduction in heart failure events. Therefore, the benefits of SGLT2 inhibitors on heart failure may be mediated by a direct cardioprotective action related to modulation of pathways responsible for cardiomyocyte homeostasis.In four large-scale clinical trials in patients with type 2 diabetes followed for 2-5 years who largely did not have a diagnosis of heart failure at the time of study entry, patients assigned to treatment