Empagliflozin, a sodium–glucose cotransporter type 2 (SGLT2) inhibitor, has been associated with a remarkable reduction in cardiovascular and all-cause mortality in patients with type 2 diabetes and antecedents of cardiovascular disease. This effect was attributed to a diuretic (haemodynamic) rather than metabolic (antiatherogenic) effect. The aim of this review is to offer arguments that either support or challenge this ‘diuretic hypothesis’.

The literature was scrutinized to: (1) examine the diuretic effects of SGLT2 inhibitors vs. hydrochlorothiazide as the reference diuretic; (2) analyze the effects of classic diuretics on cardiovascular outcomes and mortality in diabetic patients; and (3) reconsider some of the specific analyses of the EMPA-REG OUTCOME trial possibly related to a diuretic effect.

The diuretic effect of empagliflozin has so far been poorly investigated, although SGLT2 inhibitors have actions distinct from those of classic diuretics. The effects of thiazide-like diuretics on cardiovascular and overall mortality have been limited in diabetic patients with hypertension, whereas the effects of mineralocorticoid receptor antagonists in subgroups of diabetic patients with heart failure were more impressive, but still largely inferior to those reported in EMPA-REG, where relative reductions in mortality with empagliflozin were observed in diabetic patients with or without heart failure, arterial hypertension, renal impairment or diuretic background therapy.

Although the diuretic hypothesis was put forward to explain the remarkable reduction in mortality with empagliflozin in EMPA-REG, the available results do not support a major contribution of this mechanism, unless the specific diuretic effect of SGLT2 inhibitors turns out to be markedly different from those of classic diuretics.