TPOAb

This prospective cohort study aimed to develop a predictive model for gestational diabetes mellitus (GDM) in early pregnancy using blood biomarkers. Among 413 pregnant women, 116 were diagnosed with GDM. Key differences between GDM and normal controls included age, pre-pregnancy BMI, fasting plasma glucose (FPG), HbA1c, triglycerides, progesterone, IgA levels, percentage of B lymphocytes, and thyroid peroxidase antibodies (TPOAb). A predictive nomogram incorporating these risk factors was developed and evaluated. The model achieved an AUC of 0.772, demonstrating good predictive accuracy. Calibration curves indicated strong agreement between predicted and observed outcomes, and decision curve analysis (DCA) confirmed the model's clinical utility. This nomogram provides a practical tool for early GDM risk assessment, potentially improving patient counseling and management during early pregnancy.