Combined predictive value of early pregnancy fasting plasma glucose and pre-pregnancy body mass index for gestational diabetes mellitus

Abstract

This single-center retrospective study explored the predictive value of early pregnancy fasting plasma glucose (FPG) combined with
pre-pregnancy body mass index (BMI) for gestational diabetes mellitus (GDM). Healthy singleton pregnant women with natural
pregnancy who delivered at the Second People's Hospital of Wuhu City between January 2023 and January 2025 were included.
Blood glucose indicators were measured at 11–13 weeks of gestation, and a 75 g oral glucose tolerance test was performed at 24–
28 weeks. Participants were classified into GDM and non-GDM groups. After Z-score normalization, univariate and multivariate
Logistic regression analyses identified pre-pregnancy BMI, gestational weight gain and FPG as independent risk factors for GDM.
The established Logistic regression model showed moderate predictive efficiency, with an AUC of 0.75, an accuracy rate of 0.82, a
sensitivity of 0.48, a specificity of 0.92, a PPV of 0.63 and a NPV of 0.86. Bootstrap internal validation and Hosmer–Lemeshow
test indicated acceptable model calibration, while decision curve analysis showed clinical net benefit within a threshold probability
range of 0.10–0.50. We conclude that early pregnancy FPG combined with pre-pregnancy BMI may help identify women at high
risk of GDM and support personalized intervention strategies.