To Name :
To Email :
From Name :
From Email :
Comments :

Examining the Evidence

Metformin for gestational diabetes: As safe and as effective as insulin?

July 2008 · Vol. 20, No. 07
This week's quiz:
More »

YES. In this open-label randomized trial comparing metformin, with or without supplemental insulin, with insulin alone, metformin did not increase the risk of perinatal complications and was preferred by a majority of women.


Rowan and colleagues add to the data on the potential benefits of oral hypoglycemic agents, compared with insulin, in managing gestational diabetes. The presumption was that dietary treatment alone would not result in adequate glycemic control.

In the study, women assigned to metformin were given a starting dosage of 500 mg once or twice daily, which was then increased to a maximum daily dosage of 2,500 mg. According to the authors, women assigned to insulin were prescribed the drug “according to usual practice,” although that practice was never defined. In addition, if adequate glycemic control was not achieved in the metformin group, insulin was added.

Overall, 363 of the women who received metformin completed the study, with 195 receiving metformin alone and 168 ultimately receiving metformin plus insulin. In the other arm, 370 of the women assigned to insulin completed the study. Maternal baseline characteristics were the same for both groups, except that a statistically greater number of patients in the metformin group had had three or more pregnancy terminations or miscarriages.

The primary outcome of this study was a composite of various neonatal outcomes. Of the variables analyzed, significant differences were found only for prematurity (delivery <37 weeks), which was greater in the metformin group, and neonatal hypoglycemia (any blood glucose level <28.8 mg/dL), which occurred more frequently in the insulin group.

A variety of secondary outcomes were also analyzed, with no meaningful differences. The authors conclude that metformin with or without supplemental insulin is “effective and safe” for women with gestational diabetes. In the next sentence, however, they observe that “follow-up data are needed to establish long-term safety.”


All the attention to gestational diabetes has yet to significantly improve obstetric outcomes such as birth injury, C-section, or serious short-term neonatal morbidity. Nor is it any surprise that women in this study preferred metformin to insulin; most people would prefer a pill to a “shot.” However, nearly half of the pill group ended up needing a shot anyway.

Metformin is pregnancy category B and should not be used by nursing women. Rowan and colleagues acknowledge that long-term safety data are insufficient to recommend the use of oral hypoglycemic agents to manage diabetes in pregnancy.

This trial was well designed and executed, but insulin remains, in my opinion, the standard of care. Oral hypoglycemic agents just are not “ready for prime time” when it comes to gestational diabetes.—JOHN T. REPKE, MD

Did you miss this content?
Vaginal hysterectomy 
with basic instrumentation