An Updated Systematic Review and Meta-Analysis on the Efficacy and Safety of Metformin as Add-on Therapy to Insulin in Patients With Type 1 Diabetes

Abstract

Introduction: This study aims to perform an updated meta-analysis evaluating the efficacy and safety of metformin adjunct therapy in type 1 diabetes mellitus (T1DM) patients.

Method: Cochrane, PubMed and Embase were searched for randomised controlled trials (RCTs) that reported the efficacy and safety of metformin in T1DM patients. Statistical analyses were performed using STATA software.

Results: Twenty-nine placebo-controlled RCTs enrolling 2051 T1DM patients were included. Adolescents experienced a notable reduction in total insulin daily dose (TIDD) (mean difference [MD] = -0.61 [95% confidence interval (CI): -1.02, -0.20] units/kg per day) and levels of haemoglobin A1c (HbA1c) (MD = -0.45 [95% CI: -0.79, -0.11]), total cholesterol (TC) (MD = -0.78 [95% CI: -1.54, -0.02]), and low-density lipoprotein (LDL) (MD = -0.69 [95% CI: -1.36, -0.02]) at 3 months of follow-up with metformin. In adults, metformin significantly reduced Body Mass Index (BMI) (MD = -0.71 [95% CI: -1.23, -0.19]), TIDD (MD = -0.44 [95% CI: -0.73, -0.16]), and levels of HbA1c (MD = -0.70 [95% CI: -1.10, -0.30]) and TC (MD = -0.60 [95% CI: -1.09, -0.10]) at 6 months. The risk of gastrointestinal adverse events (GIAEs) was significantly higher in both adolescents (Relative Risk [RR] = 1.74 [95% CI: 1.38, 2.21]) and adults (RR = 3.24 [95% CI: 1.49, 7.02]). All of the above had p-values less than 0.05. The metformin group showed no differences in BMI Z-score, high-density lipoprotein (HDL) level, or diabetic ketoacidosis (DKA) risk. No statistical difference was identified for any of the outcomes at other follow-up endpoints.

Conclusions: Metformin may reduce TIDD and levels of HbA1c, TC, triglycerides (TG), and LDL in T1DM adolescents. BMI, TIDD, and levels of HbA1c and TC may decrease in adults. Moreover, it may raise the risk of GIAEs in both age groups.

Keywords: meta‐analysis; metformin; type 1 diabetes mellitus.

FIGURE 1

PRISMA flowchart. Study selection process for the articles in the systematic review and meta‐analysis.

FIGURE 2

Forest plots showing the BMI of adolescents. Forest plots illustrating the meta‐analysis results comparing BMI between adolescents with T1DM treated with a combination of metformin and insulin versus placebo plus insulin.

FIGURE 3

Forest plots showing the BMI of adults. Forest plots illustrating the meta‐analysis results comparing BMI between adults with T1DM treated with a combination of metformin and insulin versus placebo plus insulin.

FIGURE 4

Forest plots showing the TIDD of adolescents. Forest plots illustrating the meta‐analysis results comparing TIDD between adolescents with T1DM treated with a combination of metformin and insulin versus placebo plus insulin.

FIGURE 5

Forest plots showing the TIDD of adults. Forest plots illustrating the meta‐analysis results comparing TIDD between adults with T1DM treated with a combination of metformin and insulin versus placebo plus insulin.

FIGURE 6

Forest plots showing the HbA1c of adolescents. Forest plots illustrating the meta‐analysis results comparing HbA1c between adolescents with T1DM treated with a combination of metformin and insulin versus placebo plus insulin.

FIGURE 7

Forest plots showing the HbA1c level of adults. Forest plots illustrating the meta‐analysis results comparing HbA1c level between adults with T1DM treated with a combination of metformin and insulin versus placebo plus insulin.

FIGURE 8

Forest plot showing the GIAEs risk of adolescents. Forest plot illustrating the meta‐analysis results comparing the risk of GIAEs between adolescents with T1DM treated with a combination of metformin and insulin versus placebo plus insulin.

FIGURE 9

Forest plot showing the GIAEs risk of adults. Forest plot illustrating the meta‐analysis results comparing the risk of GIAEs between adults with T1DM treated with a combination of metformin and insulin versus placebo plus insulin.