Poster Session 2
Xiaoqing He, MS
Medical Doctoral Candidate
Ministry of Education-Shanghai Key Laboratory of Children's Environmental Health, Xinhua Hospital, affiliated with School of Medicine, Shanghai Jiao Tong University, Shanghai, China
Shanghai, Shanghai, China
Jun Zhang, PhD
Shanghai Jiao Tong University School of Medicine
Shanghai, Shanghai, China (People's Republic)
Yun Huang, PhD
Shanghai Jiao Tong University School of Medicine
Shanghai, Shanghai, China (People's Republic)
To integrate various factors to predict FGR, enabling early identification and risk stratification of high-risk populations.
Study Design:
This study was conducted based on a prospective birth cohort. A total of 1,739 mother-child pairs were included. Three nested case-control studies were further conducted: 1) 58 FGR cases (< 3rd percentile); 2) 78 X5th cases (< 5th percentile); 3) 180 SGA cases (< 10th percentile). The control group included 679 healthy controls (BW in the 25th to 75th percentile, without major complications). We developed 6 risk prediction models for birth outcomes (FGR, X5th, and SGA) using clinical risk factors, life factors, maternal and fetal polygenic risk scores (PRSs), and maternal plasma metabolic profiles in early pregnancy.
Results:
Eight clinical factors and two life factors (health behaviors and diet class) were included. Both maternal and fetal PRSs were derived from the 2019 GWAS meta-analysis for birth weight from the EGG consortium. Univariate analysis and the Unbiased Variable selection in R (MUVR) package identified potential metabolomic biomarkers. For the three case-control groups, 6, 7, and 10 metabolites were identified as potential biomarkers for FGR, X5th, and SGA, respectively.
The final model based on 8 clinical features, 2 life factors, maternal and fetal PRSs, and the 6 metabolite sets demonstrated the best predictive ability, achieving a maximum AUC of 0.826 (95% CI: 0.763-0.879) for predicting FGR. A model built solely on early pregnancy data achieved a maximum AUC of 0.744 (95% CI: 0.681-0.806) for predicting FGR.
Conclusion:
The inclusion of metabolic and genetic information significantly enhanced predictive performance compared to traditional models. Integrating genetic and metabolic data enhances the accuracy of fetal growth assessments, enabling earlier and more precise identification of high-risk pregnancies and facilitating timely interventions to improve outcomes.
