Long‐term glycaemic improvement after addition of metformin to insulin in insulin‐treated obese type 2 diabetes patients

SUMMARY Aim   To assess the adjunct effect of metformin to insulin in type 2 diabetes. Methods   Obese and overweight type 2 diabetes patients treated with insulin for at least 1 year, and with poor glycaemic control (HbA1c > upper reference level + 2%), were included in a randomised, double-blind, placebo-controlled study. Patients were treated for 12 months with either metformin (850 mg b.i.d.) or placebo added to their usual insulin, which was stabilized during a 3-month placebo run-in period, but thereafter attempted to be unchanged. Results   Thirty-seven patients were included. Two patients dropped out during run-in. There were no differences between the metformin (n = 16) and placebo (n = 19) group at baseline. Most patients received multiple insulin injections. Metabolic control was improved by addition of metformin. Mean change in HbA1c from baseline showed highly significant difference between groups at 3, 6, 9 and 12 months. Mean change (percentage units ± s.d.) at 12 months was −1.1 ± 0.7% vs. + 0.3 ± 0.8% (p < 0.001) for HbA1c and −1.4 ± 2.1 mmol/l vs. + 0.6 ± 2.2 mmol/l (p = 0.025) for fasting blood glucose. Mean low density lipoprotein (LDL) cholesterol change differed slightly at 6 months, but not at 12 months. There were no changes in insulin dose, blood pressure, body weight, triglycerides, total- and high density lipoprotein (HDL) cholesterol, fibrinogen, C-peptide and laboratory safety variables, including serum B12. Combination therapy was well-tolerated with the same adverse event rate as insulin alone, but more patients with diarrhoea. Conclusion   Addition of metformin to insulin induced and maintained clinically significant and consistent long-term reduction of hyperglycaemia in obese, insulin-treated type 2 diabetes patients.