[Metformin in adolescents and adults with type 1 diabetes mellitus: not evidence-based]

Adolescents with type 1 diabetes mellitus (DM1) often have problems in achieving optimal glycaemic control. We investigated whether there is evidence of the beneficial effect of the addition of metformin to insulin therapy in adolescents with DM1.

Systematic literature study.

Medline and Embase were searched for randomised double-blind trials in adolescents with DM1 up to May 2011 inclusive. Two reviewers selected relevant articles based on title, summary and, if necessary, the full text. The quality of the methodology was also assessed.

We found 2 studies in adolescents, of limited scope and duration. On this basis, it was decided that the search of the literature should be extended to adults with DM1, whereby 4 studies were found. All six trials were of good methodological quality, and included 196 patients in total. Clinical and statistical heterogeneity precluded pooling the results in a meta analysis. In one study in adolescents metformin treatment showed a reduction of HbA1c by 0.6% (95% CI: -1.16-0.04) and a slight decrease in daily total insulin dose. However, the treatment groups were not comparable at baseline. In the other studies, no significant changes in HbA1c were found. All studies showed decreased daily insulin dose; in four studies this was significant. Two studies showed a beneficial effect on weight or BMI. No serious side effects were recorded. One study showed an increase in hypoglycaemic episodes during metformin treatment.

The possible benefit of adding metformin to insulin in adolescents and adults with type 1 diabetes remains unclear. A well-designed double-blind randomised trial carried out over a longer time period is required to assess whether metformin is of added value.