Gestational diabetes mellitus is increasingly common and in some populations occurs in up to 14 percent of mothers. Successful treatment of gestational diabetes is associated with reduced serious maternal and fetal morbidity. Oral glyburide (Micronase) is as effective as insulin in achieving glycemic control and is associated with birth weights similar to infants whose mothers were treated with insulin. Many patients prefer oral glyburide to insulin treatment. Glyburide also is less expensive and less likely to cause hypoglycemia. However, some women do not achieve adequate glycemic control with glyburide. Kahn and colleagues studied predictors of failure of glyburide treatment with the aim of developing a tool to identify those women in whom glyburide therapy is less likely to be successful, thereby avoiding weeks of inadequate glycemic control during pregnancy.

The study included women attending a university medical center for pregnancy care. Those with morbid obesity, a history of gestational diabetes, a strong family history of diabetes, or characteristics of insulin resistance were screened for gestational diabetes early in pregnancy; all others were screened for gestational diabetes at 24 to 28 weeks’ gestation. Mothers with gestational diabetes in whom dietary interventions failed after two weeks were offered glyburide therapy unless they had a fasting blood glucose level of 125 mg per dL (6.9 mmol per L) or more or other indications of undiagnosed pre-pregnancy diabetes. Patients were informed that glyburide is not approved for use in pregnancy, but only two patients refused glyburide over insulin therapy (both had used insulin in previous pregnancies). The initial dose of glyburide was calculated for each patient based on weight and level of hyperglycemia, and the dosage was titrated to a maximum of 20 mg daily. Failure of therapy was defined as fewer than 80 percent of fasting and one-hour postprandial glucose levels in the target range. Secondary outcomes included maternal hypoglycemia, preeclampsia, method of delivery, birth weight, and neonatal complications.

About one half of the 95 mothers were Hispanic, and about one third were white. Seventy-three percent were obese (i.e., body mass index of 26 kg per m² or greater), about 8 percent smoked, and 15 percent had a history of gestational diabetes. The average age was 30 years. The mean fasting blood glucose level was 102 mg per dL (5.7 mmol per L), and the average gestation at the start of glyburide therapy was 30 weeks. Glyburide therapy was successful in 77 mothers (81 percent). Of the 18 women in whom glyburide treatment failed, 55 percent also did not experience control with insulin therapy. Several factors were significantly associated with failure of glyburide therapy (see accompanying table). In statistical analysis, after adjusting for older maternal age and multiparity, glyburide therapy was 8.3 times more likely to fail in women who had gestational diabetes diagnosed before 25 weeks’ gestation than in those with a later diagnosis. The overall neonatal outcomes were good; the average birth weight was 3,309 g, and the rate of macrosomia was 7 percent. Twelve percent of babies required intravenous glucose for hypoglycemia.

The authors conclude that glyburide is an effective and attractive treatment option for many mothers with gestational diabetes. Although 80 percent of mothers overall are successful with this therapy, it is most likely to fail in older, multiparous women with higher blood glucose levels early in pregnancy. The authors speculate that these women have increased insulin resistance and may be better treated with early initiation of insulin therapy.