This review concluded that there may be useful benefits of metformin therapy in patients with overweight or insulin insensitive type 1 diabetes. Multiple issues with the conduct and reporting of the review made the reliability of the conclusions very unclear.

To assess interventions to reduce insulin resistance in people with type 1 diabetes with particular emphasis on use of metformin (which is the focus of this abstract).

PubMed was searched without language restrictions. Search terms were reported, but search dates were not. References of selected reviews in peer-reviewed journals were checked.

Prospective controlled clinical trials that had a minimum of three months follow-up and assessed adjuvant metformin in patients with type 1 diabetes who had a body mass index of at least 25kg/m² or were insulin insensitive (total daily insulin dose >0.9U/kg) were eligible for inclusion.

Included trials had a duration of between three and six months. Mean age of patients ranged from 15.9 to 48 years. Patients' mean body mass index ranged from 22.8kg/m² to 31.3kg/m². Mean HbA_1c ranged from 7.6 to 9.4%. Mean insulin dose ranged from 0.7U/kg to 1.2U/kg. Control interventions and doses or schedules of metformin were not reported.

The authors did not state how the papers were selected for the review.

Study validity was assessed using the Jadad scale of up to 5 points for criteria of randomisation, blinding and treatment of withdrawal and dropouts.

The authors did not state how many reviewers carried out the assessment.

Data were extracted on the mean and standard deviation of glycated haemoglobin (HbA_1c), insulin dose, weight and lipid profiles.

The authors did not state how many reviewers performed data extraction.

Studies were combined using a fixed-effect meta-analysis to calculate standardised mean differences (SMDs) for HbA_1c and weighted mean difference (WMD) for total cholesterol outcomes. The estimated treatment effect on HbA_1c in absolute units was calculated by multiplication of the standardised mean difference by the mean pooled standard deviation. Statistical heterogeneity was assessed using the I² statistic (threshold of 50%). Publication bias was assessed using funnel plot inspection.

Five studies (n=149) were included in the review. Three trials had a parallel design and two a crossover design.

Metformin treatment showed a small but statistically significant improvement in HbA1c (SMD -0.346, 95% CI -0.052 to -0.640), which was equivalent to a reduction of 0.27%. There was also a moderate reduction in insulin dose (SMD -0.422, 95% CI -0.148 to -0.735). Three studies also found improvements in total cholesterol (WMD -0.433, 95% CI -0.812 to -0.054). There was no significant change in weight.

No adverse events that required cessation of therapy were reported. There was no evidence of statistically significant heterogeneity in any of the analyses. There was also no evidence of publication bias.

Despite its limitations, this analysis suggested that there may be useful benefits of metformin therapy in patients with overweight or insulin insensitive type 1 diabetes.

The review question and the inclusion criteria were unclear for comparators and outcomes. Only one database was searched, which increased the chances that some relevant studies were omitted from the review. An assessment of publication bias was conducted but, as the authors acknowledged, was unlikely to be informative with so few studies. The authors did not report using methods designed to reduce reviewer bias and error at any stage of the review process. Validity was assessed using an appropriate scale, but the results of the assessment were not reported or used to inform the synthesis. The use of meta-analysis appeared reasonable, although the treatment of crossover trials was unclear. Multiple issues with the conduct and reporting of the review made the reliability of the conclusions very unclear.

Practice: The authors stated that patients should be counselled about the risk of lactic acidosis attached to metformin therapy before treatment began.

Research: The authors stated that studies of adequate power and follow-up were required in order to assess the clinical benefits of reduced insulin resistance in type 1 diabetes.

Clinical research fellowship from Novo Nordisk.

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.