Effect of metformin on glycaemic control in patients with type 1 diabetes: A meta-analysis of randomized controlled trials.

Background: For type 1 diabetes (T1D) patients, adding metformin to insulin therapies is thought to improve blood glucose levels, but current evidence does not support this clinical benefit. Additional data from large clinical trials are now available; therefore, we conducted a meta-analysis of studies on assessing the efficacy and adverse effects of metformin.

Methods: We searched the MEDLINE, EMBASE, and Cochrane Library databases for data from randomized controlled trials. We performed statistical analyses by using Review Manager 5.2.

Results: Thirteen randomized controlled trials that compared metformin versus placebo met our inclusion criteria and were included in the study. The final meta-analysis included a total of 1183 participants with T1D. Metformin was associated with reductions in BMI (-1.14, 95% CI -2.05 to -0.24, P = .01), insulin requirements (-0.47, 95% CI -0.70 to -0.23, P = .0001), total cholesterol (-0.23, 95% CI -0.34 to -0.12, P < .0001), and low-density lipoprotein cholesterol (-0.20, 95% CI -0.29 to -0.11, P < .0001) in T1D patients. No clear evidence indicated that metformin improved HbA1c, triglyceride, or high-density lipoprotein cholesterol levels. A safety analysis showed that metformin slightly increased the risk of severe hypoglycaemia (1.23, 95% CI 1.00 to 1.52, P = .05) and mainly gastrointestinal adverse events (2.67, 95% CI 2.06 to 3.45, P < .00001). No evidence showed that metformin increased diabetic ketoacidosis events.

Conclusions: Compared with placebo, metformin was not associated with glycaemic control in T1D patients. Although it exhibited other benefits, such as lower BMI and reduced insulin requirements, total cholesterol, and low-density lipoprotein cholesterol, negative outcomes, such as gastrointestinal adverse effects and severe hypoglycaemia, should also be considered in the use of metformin for T1D patients.