Predicting gestational diabetes mellitus early: insights from biochemical and molecular predictors

This study aimed to assess the predictive capability of a combination of fasting plasma glucose (FPG) and plasma exosomal miRNAs in determining the occurrence of gestational diabetes mellitus (GDM) during the first trimester. We conducted a cross-sectional study involving randomly recruited 193 pregnant women based on exclusion criteria. Blood samples were collected at 10–14 weeks of gestation for analysis of miR-16-5p, miR-29a-3p, and FPG levels. The pregnant women were categorized into 60 GDM cases and 103 without GDM, based on a 75-g oral glucose tolerance test performed between 24 and 28 weeks of pregnancy, and excluded 30 cases. The predictive values of miR-16-5p, miR-29a-3p, and FPG for GDM were assessed using a receiver operating characteristic curve. Three databases, miRDB, TargetScan, and miRWalk, were used to predict the functions of target genes miR-16-5p and miR-29a-3p. The areas under the curve (AUC) of FPG, plasma exosomal miR-16-5p, and miR-29a-3p in predicting GDM were 0.698, 0.743, and 0.727, respectively. Combining the three factors resulted in an AUC of 0.874, with a sensitivity of 0.867 and specificity of 0.767. The selected miRNAs, miR-16-5p and miR-29a-3p, were significantly enriched in carbohydrate metabolism and glucose homeostasis, involved in the P53 signaling pathway in GDM. The combination of FPG with miRNAs demonstrated a larger AUC, along with higher sensitivity and specificity in predicting GDM.