Glycemia and Gluconeogenesis With Metformin and Liraglutide: A Randomized Trial in Youth-onset Type 2 Diabetes

Objective: Elevated rates of gluconeogenesis are an early pathogenic feature of youth-onset type 2 diabetes (Y-T2D), but targeted first-line therapies are suboptimal, especially in African American (AA) youth. We evaluated glucose-lowering mechanisms of metformin and liraglutide by measuring rates of gluconeogenesis and β-cell function after therapy in AA Y-T2D. Methods: In this parallel randomized clinical trial, 22 youth with Y-T2D—age 15.3 ± 2.1 years (mean ± SD), 68% female, body mass index (BMI) 40.1 ± 7.9 kg/m², duration of diagnosis 1.8 ± 1.3 years—were randomized to metformin alone (Met) or metformin + liraglutide (Lira) (Met + Lira) and evaluated before and after 12 weeks. Stable isotope tracers were used to measure gluconeogenesis [²H₂O] and glucose production [6,6-²H₂] glucose after an overnight fast and during a continuous meal. β-cell function (sigma) and whole-body insulin sensitivity (mS_I) were assessed during a frequently sampled 2-hour oral glucose tolerance test. Results: At baseline, gluconeogenesis, glucose production, and fasting and 2-hour glucose were comparable in both groups, though Met + Lira had higher hemoglobin A1C. Met + Lira had a greater decrease from baseline in fasting glucose (−2.0 ± 1.3 vs −0.6 ± 0.9 mmol/ L, P = .008) and a greater increase in sigma (0.72 ± 0.68 vs −0.05 ± 0.71, P = .03). The change in fractional gluconeogenesis was similar between groups (Met + Lira: −0.36 ± 9.4 vs Met: 0.04 ± 12.3%, P = .9), and there were no changes in prandial gluconeogenesis or mS_I. Increased glucose clearance in both groups was related to sigma (r = 0.63, P = .003) but not gluconeogenesis or mS_I. Conclusion: Among Y-T2D, metformin with or without liraglutide improved glycemia but did not suppress high rates of gluconeogenesis. Novel therapies that will enhance β-cell function and target the elevated rates of gluconeogenesis in Y-T2D are needed.