Background: An oral glucose tolerance test with a result that is negative but close to the diagnostic cut-off in early pregnancy was hypothesized to serve as a predictor of subsequent gestational diabetes in a high risk group. The aim of the study was to determine those cut-off values of OGTT at gestational weeks <= 16, which can predict or exclude subsequent onset of GDM in a high risk group. Methods: Pregnant women at high risk of gestational diabetes (n = 163) underwent a 2-h, 75-g oral glucose tolerance test at gestational weeks <= 16 were analyzed in this study. In the event of a negative result, subsequent oral glucose tolerance tests were performed at gestational weeks 24-28 and 32-34. The sensitivity, the specificity, the positive and negative predictive values and the Odds ratio of the best cut-off values of fasting and postload glucose levels were calculated. Results: The best cut-off values to exclude subsequent GDM for fasting and postload glucose were 5.0 and 6.2 mmol/l, respectively. In combination, the best cut-off values were 5.3 mmol/l for fasting and 6.8 mmol/l for postload glucose, with negative predictive values of 0.97 and 0.71 and sensitivities of 96.9 and 86.3 at gestational weeks 24-28 and 32-34, respectively. Combination of these cut-off values with obesity proved to be very predictive for gestational diabetes by gestational weeks 32-34, with an Odds ratio of 6.0 [95% confidence interval: 1.7-21.0]. Conclusions: With regard to the very high negative predictive value of the method, pregnant women with glucose levels of <= 5.3 mmol/l at fasting and of <= 6.8 mmol/l at postload in gestational weeks <= 16 should undergo subsequent oral glucose tolerance testing merely at gestational weeks 32-34. Approximately a quarter (24.5%) of the pregnant women at risk of gestational diabetes satisfied these criteria. (c) 2004 Elsevier Ireland Ltd. All rights reserved.
