Background Gestational diabetes mellitus (GDM) remains a major obstetric concern, yet the optimal screening strategy and the prognostic value of oral glucose tolerance test (OGTT) parameters remain debated. We aimed to compare the diagnostic yield and clinical outcomes of a two-step OGTT strategy (50 g glucose challenge followed by 100 g OGTT) versus a one-step 75 g OGTT approach, and to evaluate the predictive performance of individual OGTT time points for pregnancy complications and treatment requirement. Methods In this prospective randomized controlled trial, 1,439 pregnant women undergoing routine screening at 24–28 weeks of gestation were randomized to either a two-step OGTT strategy (n=719) or a one-step 75 g OGTT strategy (n=720). GDM was classified as diet-controlled or insulin-requiring. Maternal risk factors, obstetric outcomes, and neonatal outcomes were recorded. Receiver operating characteristic (ROC) analyses assessed the predictive ability of OGTT parameters for polyhydramnios and insulin requirement. Results Overall GDM prevalence was 12.3%, including 8.4% diet-controlled and 3.9% insulin-requiring cases. The one-step strategy identified a numerically higher proportion of GDM without significant differences in maternal or neonatal outcomes compared with the two-step approach. Rates of polyhydramnios, hypertensive disorders, macrosomia, cesarean delivery, preterm birth, neonatal intensive care admission, small for gestational age (7.4%), and intrauterine growth restriction (4.2%) were comparable between groups. ROC analyses demonstrated that 2-hour OGTT values showed the strongest predictive performance for polyhydramnios (AUC up to 0.816) and insulin requirement (AUC up to 0.808), whereas the 50 g screening test showed only moderate discrimination. Conclusion The one-step 75 g OGTT increases diagnostic labeling without improving short-term clinical outcomes. Post-load OGTT values—particularly 2-hour glucose levels—provide the most clinically meaningful prognostic information and may support a risk-stratified approach to GDM management rather than expansion of diagnostic thresholds alone.
Turgay et al. (Wed,) studied this question.