A randomised study of metformin versus insulin in gestational diabetes: early childhood outcomes.

Abstract

Metformin is a possible alternative to insulin in the treatment of gestational diabetes once it has been established that diet and exercise alone have failed to maintain glycaemic control. The largest randomised trial to date of metformin in women with gestational diabetes (MiG) (751 participants) showed greater acceptability to women, a lower risk of hypoglycaemia and less weight gain than insulin (Rowan et al. N Engl J Med 2008;358:2003–15). A meta-analysis of six randomised trials of metformin versus insulin (1420 participants) confirmed the lower weight gain, and a lower risk of neonatal hypoglycaemia in the metformin group (Su et al. Diabetes Res Clin Pract 2014;104:353–7). However, both the MiG trial and the meta-analysis showed a 60% increased risk of preterm birth that cannot be explained. In addition, metformin is known to cross the placenta, although the safety profile so far is relatively reassuring and the medication is used in pregnancy for other conditions. Therefore any information from the randomised trials regarding effects on the children after the neonatal period is valuable. The authors of this report describe follow up at 6, 12 and 18 months of 93 children whose mothers took part in a randomised trial of metformin versus insulin for gestational diabetes in Finland (Ijäs et al. BJOG 2011;118:880–5). Anthropometric measurements were taken at each child's paediatric visits by clinic staff. A brief developmental evaluation was also conducted at 18 months as part of routine care. The children in the metformin group were found to be heavier but also taller at 12 and 18 months. In addition, at 12 months significantly more children were greater than the 95th centile for height in the metformin group and at 6 months significantly more children in the insulin group were less than or equal to the fifth centile. Ponderal index was not different between the two groups at any visit. The developmental examination did not reveal any differences; however, the assessment was basic, and the power to detect any but extreme differences was limited. There was no difference in the age at walking or standing. Previously, children from the MiG study followed up at 2 years were shown to have higher subscapular and biceps skinfold measurements in the metformin group than the insulin group (Rowan et al. Diabetes Care 2011;34:2279–84). All other measurements, including body fat percentage, were not significantly different, which may signify a more favourable distribution of body fat. However, this study was marred by the large lost to follow up rate (44%) at the three sites involved. The results of trials to date suggest that metformin use in pregnancy is beneficial and apparently safe, apart from the still open question of preterm birth. The Finnish study gives some reassurance that the longer-term consequences of exposure in-utero are benign. Although more research is needed, this study still provides a useful piece of the puzzle. I have no interests to declare.