To determine whether the use of oral hypoglycemic agents during early pregnancy is associated with a risk of congenital malformations in infants of mothers with non-insulin-dependent diabetes mellitus (NIDDM) independent of maternal metabolic control.

From a prospectively collected data-base of pregnancies complicated by diabetes at a large urban medical center, we identified 332 consecutive infants born to women with NIDDM who did not participate in a preconceptional diabetes care program. Stepwise logistical regression was used to identify maternal characteristics that were independently associated with risks of major and minor congenital malformations in infants.

Overall, 56 (16.9%) of the 332 infants were born with congenital anomalies (11.7% major anomalies and 5.1% minor anomalies). Analysis of data from subgroups of women who were treated with diet therapy, exogenous insulin, or sulfonylurea compounds during the first 8 weeks of gestation did not reveal statistically significant differences in major or minor malformation rates among the three groups. Stepwise logistic regression analysis revealed two maternal characteristics that were independently associated with major malformations in infants: maternal HbA_1c at initial presentation for care (direct relationship; P = 0.0007) and the maternal age at onset of diabetes (inverse relationship; P = 0.009). The risk of major malformations was unrelated to the mode of antidiabetic therapy during early pregnancy. No relationship was found between maternal glycemia or treatment modality and rates of minor congenital anomalies.

These data indicate that, in the absence of special preconceptional care, NIDDM is associated with a risk for major congenital anomalies that is in the range reported for pregnancies complicated by insulin-dependent diabetes mellitus. Moreover, the risk in individual patients appears to be related to maternal glycemic control rather than to the mode of antidiabetic therapy during early pregnancy.