Early prediction of gestational diabetes mellitus: the role of the pregnancy-specific triglycerides-glucose index and other fasting parameters in combination with dynamic testing.

The identification of mothers at risk for gestational diabetes mellitus (GDM) at start of pregnancy may be beneficial to improve perinatal outcomes. This study aims evaluating the predictive performance of fasting and dynamic indices of glucose metabolism at first trimester and their association with later GDM development. A cohort of 198 women received detailed metabolic assessment at median gestational age (13 weeks) including 75-g oral glucose tolerance test (OGTT) with assessment of glucose, insulin and C-peptide, and biochemical markers (including triglycerides) to calculate different indices of insulin sensitivity either at fasting and in the OGTT dynamic conditions. Moreover, parameters of β-cell function were assessed. A second OGTT was performed between 24 and 28 gestational weeks (GW) to identify women with GDM. We found that 28 women developed GDM, and, in univariable analysis, this was fairly predicted by several first trimester indices, both at fasting and in dynamic conditions. However, fasting indices containing maternal triglycerides showed better accuracy as compared to traditional indices (even the dynamic ones). In multivariable analysis, the best predictive model of GDM development included fasting and OGTT glucose values, HbA1c, and an insulin sensitivity marker that includes triglycerides (e.g. the improved triglyceride-glucose index, TyGIS). β-Cell function was not included in such predictive model, but at 24-28 GW it showed remarkable impairment in women with GDM. In conclusion, both fasting and dynamic parameters of glucose homeostasis at early pregnancy showed fair predictive accuracy for later GDM, with TyGIS showing excellent performance. β-Cell dysfunction role needs being further elucidated.