Poster Presentation Australian Diabetes Society and the Australian Diabetes Educators Association Annual Scientific Meeting 2016

First trimester screening for gestational diabetes: A risk prediction model combining maternal clinical and biomarker parameters      (#314)

Arianne N Sweeting 1 2 , Jencia Wong 1 2 , Glynis P Ross 1 2 , Paul F Williams 1 2 3 , Heidi Appelblom 4 , Heikki Kouru 4 , Mikko Sairanen 4 , Jonathan Hyett 2 5
  1. Royal Prince Alfred Hospital, Diabetes Centre, Sydney, Australia
  2. Discipline of Medicine, University of Sydney, Sydney, Australia
  3. Department of Clinical Biochemistry, Royal Prince Alfred Hospital, Sydney, Australia
  4. Diagnostics, Perkin Elmer, Turku, Finland
  5. Royal Prince Alfred Hospital, Department of High Risk Obstetrics, Sydney, Australia

Introduction: Gestational diabetes (GDM) can be prevented in high-risk women1, however the best test in early pregnancy to predict GDM is unknown. A multivariate prediction model combining maternal risk factors may be superior for early detection2, but is yet to be assessed in an Australian population.

Aim: 1.Develop a first trimester risk prediction model for GDM combining maternal clinical and biomarker parameters in a large multi-ethnic cohort. 2.Compare the performance of this model to other published prediction models2-4 within our cohort.

Methods: Retrospective case control study of 224 women who developed GDM (ADIPS 1998 diagnostic criteria5) and 718 controls who did not, undertaken at Royal Prince Alfred Hospital, Sydney. Clinical parameters (ethnicity, age, parity, body mass index (BMI), previous GDM, family history of diabetes, mean arterial pressure (MAP) and uterine artery dopplers) were prospectively obtained at 11-13+6 weeks’ gestation. Biomarkers (β-HCG, PAPP-A, adiponectin, leptin, PAI-2, lipocalin-2, non-fasting glucose, lipids) were measured on retrieved samples. A predictive logistic regression model for GDM was developed, calculating areas under the receiver-operating characteristic (AUROC) curve. The performance of our model was compared to derived AUROCs for the published models within our cohort.

Results: The AUROC for a model based on weighted clinical factors alone (age, BMI and ethnicity) was 0.72. The addition of PAPP-A+MAP and triglycerides+lipocalin-2 to these clinical factors improved the AUROC to 0.75 and 0.78, respectively. The best performing model combined all these risk factors, achieving an AUROC of 0.93 and detection rate of 87% for a 20% false-positive rate. This AUROC was also higher than the other models (derived AUROCs ranged from 0.64-0.87).

Conclusion: Our multivariate model combining maternal clinical and biomarker risk factors accurately predicts GDM in early pregnancy, out-performing other prediction models. Although prospective validation is required, an effective early screening and intervention strategy may prevent subsequent GDM.

  1. Koivusalo et al. Gestational Diabetes Mellitus Can Be Prevented by Lifestyle Intervention: The Finnish Gestational Diabetes Prevention Study (RADIEL): A Randomized Controlled Trial. Diabetes Care. 2016;39(1):24-30.
  2. Syngelaki et al. First-Trimester Screening for Gestational Diabetes Mellitus Based on Maternal Characteristics and History. Fetal Diagn Ther. 2015;38(1):14-21.
  3. Nanda et al. Prediction of gestational diabetes mellitus by maternal factors and biomarkers at 11 to 13 weeks. Prenat Diagn. 2011;31(2):135-141.
  4. van Leeuwen et al. Estimating the risk of gestational diabetes mellitus: a clinical prediction model based on patient characteristics and medical history. BJOG. 2010;117(1):69-75.
  5. Hoffman et al for the Australasian Diabetes in Pregnancy Society (ADIPS). Gestational diabetes mellitus – management guidelines. Med J Aust 1998;169:93–7.