Objective To determine whether the glycemic status of pregnant women with a normal 3-h 100-g oral glucose tolerance test (OGTT) is related to outcome with respect to large for gestational age (LGA) infants.

Research Design and Methods A prospective study of 2631 women was conducted. One hundred seventy-six women had an OGTT based either on a 1-h 50-g OGTT (n = 105) or clinical risk factors (n = 71). Thirty-three women were diagnosed as having gestational diabetes mellitus.

Results Negligible discriminatory capacity for the variables with respect to prediction of LGA infants was indicated by the areas under the receiver operating characteristic (ROC) curves for fasting blood glucose, 2-h OGTT blood glucose, and the OGTT response curve area for women with a normal OGTT (n = 143). However, a statistically significant increased incidence of LGA infants was established for both the OGTT-positive and normal OGTT groups (P < 0.0001). Multiparity, a maternal weight > 70 kg, and birth of a male infant were other factors associated with a significantly increased frequency of LGA infants.

Conclusions The results may be interpreted as either indicating a role for confounding variables, i.e., maternal weight, multiparity, and birth of a male infant, or the imprecision of the OGTT in assessing physiologically important changes in maternal hyperglycemia.