Energy Balance After Sodium–Glucose Cotransporter 2 Inhibition

OBJECTIVE: Sodium-glucose cotransporter 2 (SGLT2) inhibitors cause substantially less weight loss than expected from the energy excreted via glycosuria. Our aim was to analyze this phenomenon quantitatively. RESEARCH DESIGN AND METHODS: Eighty-six patients with type 2 diabetes (HbA1c 7.8 ± 0.8% 62 ± 9 mmol/mol, estimated glomerular filtration rate eGFR 89 ± 19 mL ⋅ min(-1) ⋅ 1.73 m(-2)) received empagliflozin (25 mg/day) for 90 weeks with frequent (n = 11) assessments of body weight, eGFR, and fasting plasma glucose (FPG). Time-dependent glucose filtration was calculated as the product of eGFR and FPG; time-dependent glycosuria was estimated from previous direct measurements. The relation of calorie-to-weight changes was estimated using a mathematical model of human energy metabolism that simulates the time course of weight change for a given change in calorie balance and calculates the corresponding energy intake changes. RESULTS: At week 90, weight loss averaged -3.2 ± 4.2 kg (corresponding to a median calorie deficit of 51 kcal/day interquartile range (IQR) 112). However, the observed calorie loss through glycosuria (206 kcal/day IQR 90) was predicted to result in a weight loss of -11.3 ± 3.1 kg, assuming no compensatory changes in energy intake. Thus, patients lost only 29 ± 41% of the weight loss predicted by their glycosuria; the model indicated that this difference was accounted for by a 13% (IQR 12) increase in calorie intake (269 kcal/day IQR 258) coupled with a 2% (IQR 5) increase in daily energy expenditure (due to diet-induced thermogenesis). This increased calorie intake was inversely related to baseline BMI (partial r = -0.34, P < 0.01) and positively to baseline eGFR (partial r = 0.29, P < 0.01). CONCLUSIONS: Chronic glycosuria elicits an adaptive increase in energy intake. Combining SGLT2 inhibition with caloric restriction is expected to be associated with major weight loss.