HbA1c

This study aimed to develop a predictive model for early identification of gestational diabetes mellitus (GDM) based on clinical factors and biomarkers. Using data from a cohort of 7929 pregnant women, a nested case-control analysis was conducted, including 264 women with GDM and matched controls. HbA1c and hsCRP were elevated, while SHBG was reduced in women who developed GDM (p<0.001). The predictive model for GDM, incorporating HbA1c, SHBG, BMI, history of GDM, family history of diabetes, and pre-pregnancy soft drink intake, achieved an AUC of 0.79. For predicting GDM requiring insulin therapy, the model had an AUC of 0.88 and a sensitivity of 68.9% at a 10% false-positive rate. This simple model could enable early identification of women at risk of GDM, particularly those requiring insulin therapy, facilitating timely interventions.