ECE2022 Poster Presentations Diabetes, Obesity, Metabolism and Nutrition (202 abstracts)
Development and validation of a new gestational diabetes mellitus (gdm) risk algorithm
Leire Mendizabal 1 , Maddi Arregui 1 , Johanna Valerio 2 , Ana María Ramos 2 , Ana Barabash 2 , Nuria Garcia de la Torre 2 , mirella zulueta 3 , Eunate Arana 4 , Eva Saez 1 , Igotz Aránbarri 1 , Juan Martinez-Lage 1 , Alicia Cortazar 4 , Sonia Gaztambide 4 , Luis Castaño 4 , Laureano Simon 1 & Alfonso Luis Calle-Pascual 2
1Patia Europe, San Sebastian, Spain; 2Hospital Clínico San Carlos, Madrid, Spain; 3Patia, Boston, United States; 4Hospital de Bilbao Biocruces, Bilbao, Spain
Background and Objective: Gestational Diabetes Mellitus (GDM) is associated with life-long adverse outcomes for the mother and the baby. To date there is no rigorous clinical test for the assessment of GDM risk, since estimation of GDM risk is currently primarily based on clinical risk factors. Additional markers are needed to identify women at high risk. Our aim was to develop and validate a risk assessment model to identify women at high risk of GDM through an algorithm that integrates genetic and clinical variables.
Methodology: We analyzed a retrospective cohort of 711 women with 425 control pregnancies and 286 GDM cases. The entire cohort was randomly divided into a training/development dataset (70% of the cohort) for algorithm development and a test dataset (30% of the cohort) for validation. A total of 112 SNPs (Single Nucleotide Polymorphisms) were selected for this analysis after exhaustive exploration of the databases published to date of SNPs associated with GDM. The SNPs were selected based on their predictive power and population frequency, with the following criteria: OR>1.2, RAF>0.20, P<1×10-5. SNPs were grouped into glycemic traits categories. Genotyping was performed using iPlex Gold-MassARRAY from Agena Bioscience. In the clinical and genotype data set of the development/training group, significant attribute selection was performed using Sequence Feature Selection (SFS) techniques. Logistic regression analysis was then applied to obtain prediction coefficients for the selected attributes in the training group data set. Discrimination and calibration of risk scores were evaluated using the receiver operating characteristic (ROC) curve in the training and the validation dataset.
Results: An algorithm was developed on the training dataset that provides a risk score for GDM. The algorithm includes 10 SNPs, maternal age, pregestational body mass index, and number of previous pregnancies. In the training dataset the AUC was 0.7420. The AUC of the 10 SNPS alone (0.6981) and the clinical variables alone (0.6133) were significantly lower than their combination. AUC in the validation set was 0.7139.
Conclusions: a new tool for GDM risk assessment is presented, which suggests that the utilization of genetic markers in combination with clinical characteristics may improve accuracy of GDM risk evaluation and reinforce the adoption of preventive intervention as early as possible. Further clinical validation studies in different patient cohorts are ongoing.
Article tools
My recent searches
My recently viewed abstracts
Authors
Endocrine Abstracts
ISSN 1470-3947 (print) | ISSN 1479-6848 (online)
© Bioscientifica 2024 |
Privacy policy |
Cookie settings
BiosciAbstracts
Bioscientifica Abstracts is the gateway to a series of products that provide a permanent, citable record of abstracts for biomedical and life science conferences.
Find out more