SGLT2 inhibitors (SGLT2i) improve glycemic control by inhibiting renal proximal tubular glucose reabsorption. SGLT2i may also elevate ketones. Fasting is expected to increase endogenous ketones, and SGLT2i may enhance ketosis. We asked if SGLT2i alters circulating ketones (β-hydroxybutyrate; BOHB) and disrupts glycemia during prolonged fasting. Dogs (n=7) were studied with a crossover design. Dogs received either canagliflozin (CANA; 300 mg) or placebo (PL) once/day for 7 d, after which a 48 h fast was performed while treatment was continued. To begin the fast, after morning feeding (9 am to noon) , remaining food was removed (“t=0”) . Blood was sampled at regular intervals over a 48 h period of total fasting (urine collected over final 3 h) . After the fast, dogs were fed over 3 h. Treatment was then discontinued for 2 wk washout, and dogs were then assigned to the alternate treatment (CANA or PL) for 7 d, after which fasting experiment was repeated. Glycemia was sustained in both treatment groups during initial 22 h of fasting, presumably due to reduction of insulin (↓50.7±13.1% and ↓49.5±21.5% for CANA and PL, respectively; p<0.035 for both) . But marked effects of CANA alone were revealed after 22 h of fasting, when glucose declined markedly in CANA (52.9±3.4 mg/dl at t=48 h) , but not PL (86.6±2.5 mg/dl at 48 h; p=0.0002 vs. CANA) , due to drug-induced glycosuria (13948±543 vs. 19.6±1.5 mg/dl; p<0.0001) . After 22 h of fasting, CANA induced ∼5-fold increase in FFA (p<0.03) and marked increase in plasma BOHB that persisted for duration of fast (CANA: 0.560±0.065 mM, PL: 0.048±0.006 mM; p<0.0001) . After refeeding, glucose and BOHB normalized, although BOHB remained higher in CANA (p<0.0007) .

In conclusion, effect of SGLT2i on ketones is manifest after 24 h of total food restriction. CANA treatment induces glycosuria, after which reduction in carbohydrate stores switches animals to fat-based metabolism, characterized by increased FFA and ketosis. CANA appears to induce ketosis when carbohydrate stores are depleted.


M. Ader: None. R. Sedighi: None. M. Kabir: None. R. N. Bergman: Consultant; Fractyl Health, Inc., Novo Nordisk, Research Support; AstraZeneca, Janssen Research & Development, LLC.


Janssen Pharmaceutica, Inc.; NIH (DK29867, DK27619)

Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at