Submitted
Abstract Submission
Predicting Postpartum Dysglycaemia Following GDM: Clinical and Biochemical Drivers from a Real-World Cohort
Poster Presentation
Scientific Research Abstract
Diabetes
Author's Information
5
No more than 15 authors can be listed (as per the Good Publication Practice (GPP) Guidelines).
Please ensure the authors are listed in the right order.
Ryan Tan ryan.tan121@imperial.ac.uk Imperial College London Medicine London United Kingdom -
Rui Cheng Luo RLUO007@e.ntu.edu.sg Nanyang Technological University Medicine singapore Singapore -
Amal Al-Aamri smart6doctor@gmail.com Department of Internal Medicine, Nizwa Hospital, Sultanate of Oman, Medicine oman Oman -
Chin Meng Khoo mdckcm@nus.edu.sg National University Hospital Endocrinology Singapore Singapore -
Pei Chia Eng janoviee@gmail.com National University Hospital Endocrinology Singapore Singapore *
-
-
-
-
-
-
-
-
-
-
Abstract Content
Women with prior gestational diabetes mellitus (GDM) have a 6–12× higher lifetime risk of type 2 diabetes. Postpartum risk stratification typically relies on a 75-g oral glucose tolerance test (OGTT) at 6–12 weeks after delivery. We sought to identify antepartum predictors of postpartum dysglycaemia to enable targeted follow-up after delivery.
We assessed women with prior GDM who completed a 75-g OGTT at 6-12 weeks postpartum (1/10/2023–31/07/2025). The outcome was dysglycaemia versus normoglycaemia on postpartum OGTT. Univariate and multivariate logistic analyses were performed; multicollinearity was assessed by variation inflation factors and missing data were multiply imputed. Model performance was evaluated using AUC and Brier score with 1,000 bootstrap iterations for internal validation. Predictor importance was evaluated via dominance analysis. Optimal thresholds for important continuous variables were identified using Youden’s index.
Of 409 women, 140 (34%) had postpartum dysglycaemia and 269 (65.8%) had normoglycaemia. In the final multivariate model, the predictors were maternal age (OR 1.05), higher antepartum 2-hrs OGTT glucose threshold (OR 2.03), higher HbA1c at GDM diagnosis (OR 3.08), and family history of diabetes (OR 2.29). Compared to Malays, Chinese ethnicity had a 1.70-odds of developing postpartum dysglycaemia. In-vitro fertilisation (IVF) (OR 0.19) and pre-pregnancy BMI showed an inverse association (OR 0.95). Model performance was acceptable (AUC 0.765, 95% CI 0.721–0.807; Brier 0.174). Dominance analysis highlighted the 2-hrs OGTT as the largest contributor (32.5% of pseudo-R²), followed by family history (16.3%), ethnicity (15.4%), HbA1c (13.0%), IVF (8.5%), age (4.6%), and pre-pregnancy BMI (5.7%). Youden-optimised thresholds for postpartum dysglycaemia were an antenatal HbA1c 5.45% and a 2 hours OGTT of 8.45mmol/L.
Readily available antepartum metrics could identify women at highest risk of postpartum dysglycaemia and could inform targeted prevention within postpartum diabetes care pathways.
Gestational diabetes, predictors
 
 
314
Pei Chia
Eng
janoviee@gmail.com
 
Presentation Details