Abstract:Aim To investigate the association of triglyceride-glucose index-body mass index (TyG-BMI) trajectories with the incidence of gestational diabetes mellitus (GDM) and adverse pregnancy outcomes, integrating data mining from the National Health and Nutrition Examination Survey (NHANES) database with real-world case validation, to provide a clinical basis for early screening, risk stratification of GDM, and improvement of pregnancy outcomes. Methods This study employed a two-stage design combining public database mining with real-world clinical validation:first, data from six cycles of the NHANES from 2007 to 2020 were extracted, including 153 reproductive-age pregnant women. Multivariate Logistic regression was used to analyze the association between TyG-BMI levels and the risk of GDM, and restricted cubic spline was employed to assess the dose-response relationship. Subsequently, clinical data were retrospectively collected from 486 pregnant women with GDM at Baoji Maternal and Child Health Hospital. Group-based latent growth mixture model (LGMM) was used to identify TyG-BMI trajectories from early to mid-pregnancy. Univariate and multivariate Logistic regression analyses were performed to examine the association between these trajectories and adverse pregnancy outcomes, and ROC curves were used to evaluate predictive performance. The primary outcome was the incidence of GDM, and secondary outcomes were composite adverse pregnancy outcomes, including hypertensive disorders of pregnancy and cesarean section. Results The analysis results of the NHANES database showed that the BMI and TyG-BMI levels in the GDM group were significantly higher than those in the non-GDM group (P＜0.05). The results of univariate Logistic regression analysis showed that for every 1 unit increase in TyG-BMI, the risk of GDM increased by 0.7% (OR＝1.7,5%CI:1.001～1.013, P＝0.015); after adjusting for age and BMI, the association was no longer statistically significant (P＞0.05). Restricted cubic spline analysis confirmed a linear positive correlation between TyG-BMI and GDM (overall correlation P＝0.043), with no clear risk threshold or inflection point. Clinical cohort validation identified three distinct TyG-BMI trajectory patterns:the stable group (n＝125), the moderate-increase group (n＝293), and the rapid-increase group (n＝68). The proportion of family history of obesity was significantly higher in the rapid-increase group compared to the other two groups (P＝0.028), while there were no significant differences in age or pre-pregnancy BMI among the three groups (P＞0.05). Among the 486 GDM women, the overall incidence of composite adverse pregnancy outcomes was 65.8%, and which increased across the three TyG-BMI trajectory groups, being 56.00%, 66.89% and 79.41% in the stable, moderate-increase, and rapid-increase groups, respectively (P＜0.05). Similar trends were observed for individual adverse outcomes such as cesarean section and macrosomia. The results of multiple Logistic regression analysis showed that, with the stable group as a reference and after adjusting for confounders, the risk of composite adverse pregnancy outcomes increased by 57.1% (OR＝1.1,5%CI:1.018～2.423) and 195.7% (OR＝2.7,5%CI:1.514～6.064) in the moderate-increase group and rapid-increase group, respectively. ROC curve analysis indicated that the TyG-BMI trajectory had good predictive performance for composite adverse pregnancy outcomes (AUC＝0.8,5%CI:0.863～0.932), significantly better than single TyG-BMI measurements in early pregnancy (AUC＝0.566) and mid-pregnancy (AUC＝0.614, P＜0.05), and comparable to ΔTyG-BMI (AUC＝0.905, P＝0.460). After constructing a predictive model combining TyG-BMI trajectory with indicators such as age and pre-pregnancy BMI, the AUC reached 0.872 (95%CI:0.832～0.912), with sensitivity and specificity superior to single indicators. Conclusions The moderate and rapid increasing trends of TyG-BMI can disrupt the homeostasis of glucose and lipid metabolism as well as body weight during pregnancy, and are independently associated with adverse pregnancy outcomes in women with GDM. As a reflection of these metabolic changes, the TyG-BMI dynamic trajectory demonstrates predictive performance superior to single-point measurements. Dynamic monitoring of TyG-BMI trends during pregnancy can facilitate early identification of high-risk populations with metabolic homeostasis imbalance, providing a novel intervention target for improving pregnancy outcomes.
