RESUMO
ABSTRACT Objective Our goal was to investigate which glucose measurement from the 75-g oral glucose tolerance test (OGTT) has more capability of predicting large for-gestational-age (LGA) newborns of mothers with gestational diabetes mellitus (GDM). Subjects and methods The study group consisted of 118 consecutively pregnant women with singleton pregnancy, patients of Outpatients Department of the Endocrinology, Diabetes, and Metabolic Disorders Clinic. All were prospectively screened for GDM between 24th and 28th week of pregnancy and followed to delivery. Outcome measures included: patients’ ages, pre-pregnancy BMI, BMI before delivery, FPG, 1 and 2 hour OGTT glucose values, haemoglobin A1c at third trimester, gestational week of delivery, mode of delivery and baby birth weight. Results From 118 pregnancies, 78 (66.1%) women were with GDM, and 40 (33.9%) without GDM. There were statistically significant differences (30.7 versus 5.0%, p < 0.01) between LGA newborns from GDM and control group, respectively. Gestation week of delivery and fasting glucose levels were independent predictors for LGA (Beta = 0.58 and Beta = 0.37 respectively, p < 0.01). Areas under the receiver operator characteristic curve (AUC) were compared for the prediction of LGA (0.782 (0.685-0.861) for fasting, 0.719 (0.607-0.815) for 1-hour and 0.51 (0.392-0.626) for 2-hour OGTT plasma glucose levels). Conclusion Fasting and 1-hour plasma glucose levels from OGTT may predict LGA babies in GDM pregnancies.