Diagnostic Performance of the Triglyceride-Glucose Index in Screening for Gestational Diabetes Mellitus at 24–28 Weeks of Gestation

<b>Background/Objectives</b>: The objective of this study was to assess the diagnostic accuracy of the Triglyceride-Glucose (TyG) index for screening gestational diabetes mellitus (GDM) at 24-28 weeks of gestation, to determine its optimal diagnostic threshold, and to compare its predictive performance with conventional lipid ratios (LDL/HDL, TG/HDL, and TC/HDL). <b>Materials and Methods</b>: We retrospectively analyzed 440 pregnant women with singleton pregnancies who underwent a 75 g oral glucose tolerance test (OGTT) between January and July 2025. The TyG index and lipid ratios were calculated, and their associations with GDM were evaluated. Subgroup analyses were conducted to assess the efficacy of the TyG index in predicting GDM, using logistic regression to estimate odds ratios (ORs) with 95% confidence intervals (CIs), and receiver operating characteristic (ROC) curve analysis along with restricted cubic spline modeling to evaluate diagnostic performance and determine the optimal cutoff value. <b>Results</b>: The overall prevalence of GDM, as defined by the IADPSG (International Association of the Diabetes and Pregnancy Study Groups) criteria, was 22.7%. The median TyG index was significantly higher in the GDM group compared with the non-GDM group (9.1 vs. 8.9, <i>p</i> = 0.001). The TyG index was a significant predictor of GDM (<i>p</i> < 0.05), with each one-unit increase associated with significantly higher odds of GDM (OR = 12.29), after adjusting for covariates. ROC analysis demonstrated an AUC of 0.716 (95% CI: 0.627-0.793, <i>p</i> < 0.001) for the TyG index, and the optimal cut-off value was identified as 9.35, yielding a sensitivity of 38.5% and a specificity of 96.5% and a negative predictive value of 83.7%. Subgroup analyses indicated that the TyG index had limited discriminative ability for predicting GDM in both the post-load and insulin-requiring groups. Among conventional lipid ratios, TG/HDL demonstrated the highest predictive performance (AUC = 0.587), while LDL/HDL (AUC = 0.483) and TC/HDL (AUC = 0.509) demonstrated low predictive accuracy. Compared with conventional lipid ratios, the TyG index demonstrated superior predictive performance. <b>Conclusions</b>: A higher TyG index was positively associated with the development of GDM and showed better predictive ability than conventional lipid ratios. However, its low sensitivity limits its use as a standalone diagnostic tool, suggesting it may be most useful when combined with other clinical parameters.