Metformin improves blood glucose by increasing incretins independent of changes in gluconeogenesis in youth with type 2 diabetes

Aims/hypothesis: Metformin is the only approved oral agent for youth with type 2 diabetes but its mechanism of action remains controversial. Recent data in adults suggest a primary role for the enteroinsular pathway, but there are no data in youth, in whom metformin efficacy is only ~50%. Our objectives were to compare incretin concentrations and rates of glucose production and gluconeogenesis in youth with type 2 diabetes before and after short-term metformin therapy compared with peers with normal glucose tolerance (NGT). Methods: This is a case–control observational study in youth with type 2 diabetes who were not on metformin (n = 18) compared with youth with NGT (n = 10) who were evaluated with a 2 day protocol. A 75 g OGTT was administered to measure intact glucagon-like 1 peptide (iGLP-1), gastric inhibitory polypeptide (GIP) and peptide YY (PYY). Insulinogenic index (IGI) and whole-body insulin sensitivity were calculated using glucose and insulin levels from the OGTT. Basal rates of gluconeogenesis (²H₂O), glucose production ([6,6-²H₂]glucose) and whole-body lipolysis ([²H₅]glycerol) were measured after an overnight fast on study day 2. Youth with type 2 diabetes (n = 9) were subsequently evaluated with an identical 2 day protocol after 3 months on the metformin study. Results: Compared with individuals with NGT, those with type 2 diabetes had higher fasting (7.8 ± 2.5 vs 5.1 ± 0.3 mmol/l, mean ± SD p = 0.002) and 2 h glucose concentrations (13.8 ± 4.5 vs 5.9 ± 0.9 mmol/l, p = 0.001), higher rates of absolute gluconeogenesis (10.0 ± 1.7 vs 7.2 ± 1.1 μmol [kg fat-free mass (FFM)]⁻¹ min⁻¹, p < 0.001) and whole-body lipolysis (5.2 ± 0.9 vs 4.0 ± 1.4 μmol kg_FFM⁻¹ min⁻¹, p < 0.01), but lower fasting iGLP-1 concentrations (0.5 ± 0.5 vs 1.3 ± 0.7 pmol/l, p < 0.01). Metformin decreased 2 h glucose (pre metformin 11.4 ± 2.8 vs post metformin 9.9 ± 1.9 mmol/l, p = 0.04) and was associated with ~20–50% increase in IGI (median [25th–75th percentile] pre 1.39 [0.89–1.47] vs post 1.43 [0.88–2.70], p = 0.04), fasting iGLP-1 (pre 0.3 ± 0.2 vs post 1.0 ± 0.7 pmol/l, p = 0.02), 2 h iGLP (pre 0.4 ± 0.2 vs post 1.2 ± 0.9 pmol/l, p = 0.06), fasting PYY (pre 6.3 ± 2.2 vs post 10.5 ± 4.3 pmol/l, p < 0.01) and 2 h PYY (pre 6.6 ± 2.9 vs post 9.0 ± 4.0 pmol/l, p < 0.01). There was no change in BMI, insulin sensitivity or GIP concentrations pre vs post metformin. There were no differences pre vs post metformin in rates of glucose production (15.0 ± 3.9 vs 14.9 ± 2.2 μmol kg_FFM⁻¹ min⁻¹, p = 0.84), absolute gluconeogenesis (9.9 ± 1.8 vs 9.7 ± 1.7 μmol kg_FFM⁻¹ min⁻¹, p = 0.76) or whole-body lipolysis (5.0 ± 0.7 vs 5.3 ± 1.3 μmol kg_FFM⁻¹ min⁻¹, p = 0.20). Post metformin iGLP-1 and PYY concentrations in youth with type 2 diabetes were comparable to levels in youth with NGT. Conclusions/interpretation: Overall, the improved postprandial blood glucose levels and increase in incretins observed in the absence of changes in insulin sensitivity and gluconeogenesis, support an enteroinsular mechanistic pathway in youth with type 2 diabetes treated with short-term metformin. Graphical abstract: [Figure not available: see fulltext.]