Metformin vs. insulin in gestational diabetes. A randomized study characterizing metformin patients needing additional insulin

Aims We compared metformin with insulin as treatment of gestational diabetes mellitus (GDM). Furthermore, we aimed to characterize metformin-treated patients needing additional insulin to achieve prespecified glucose targets. Methods We conducted a single centre randomized controlled study with non-inferiority design comparing metformin and insulin in the treatment of 217 GDM patients having birth weight as primary outcome variable. Results There were no significant differences in mean birth weight expressed in grams [+15 (90% confidence interval (CI): -121 to 89)] or SD units [+0.04 (90% CI: -0.27 to 0.18)] between the metformin and insulin groups. There were no significant differences in neonatal or maternal data between the groups. Only 23 (20.9%) of the 110 patients in the metformin group needed additional insulin. Compared with the patients on metformin only, those needing additional insulin were older (p?=?0.04), their oral glucose tolerance test had been performed earlier and diabetes therapy started earlier in gestation (p?=?0.01 and p?=?0.004, respectively). The risk for additional insulin was 4.6-fold in women with baseline serum fructosamine concentration above median compared with those below median. Conclusions Metformin is an effective alternative to insulin in the treatment of GDM patients. Serum fructosamine may help in predicting the adequacy of metformin treatment alone.