271: The effect of metformin on insulin requirements in pregnancies complicated by type 2 diabetes

Abstract

To investigate the effect of metformin on gestational insulin requirements in pregnancies complicated by type 2 diabetes (T2DM) in obese and non-obese populations. A retrospective cohort study was performed on women with singleton term pregnancies with T2DM enrolled in the U.C. San Diego Diabetes and Pregnancy program from February 2008 to October 2011. Exclusion criteria included use of sulfonylureas, delivery at outside institutions, or insufficient prenatal care. Comparisons of patients using insulin alone versus those on insulin/metformin therapy were made using the t-test for continuous variables and the chi-squared or fisher's exact tests for categorical variables. A total of 69 women met inclusion criteria, 40 in the insulin-only cohort and 19 in the insulin/metformin cohort. Demographic variables, including age, ethnicity, parity, initial BMI and HbA1c were not significantly different between the two groups. 75% of patients in the insulin-only cohort and 53% in the insulin/metformin cohort were obese in the first trimester. In the obese population, patients on metformin had significantly higher total and short-acting insulin requirements at 36 weeks compared to the insulin-only cohort (table 1). In the non-obese population, there were no significant differences in insulin requirements between the metformin/insulin and insulin-only cohorts. Gestational age at delivery, infant birthweight, frequency of macrosomia, mode of delivery, and Apgar scores were similar in both groups.Tabled 1Insulin requirements for the insulin-only and metformin/insulin combination therapy cohorts Obese patients with T2DM on metformin as an adjunct to insulin required higher insulin doses in the third trimester as compared to patients on insulin only. Use of metformin did not reveal any negative effect on birth outcomes. A key mechanism of action of metformin, increasing insulin sensitivity, would purport that decreased insulin doses would be required, contrary to our conclusions. Prospective trials are needed to elucidate whether empiric metformin can truly decrease insulin needs.