Location: Responsive Agricultural Food Systems Research Unit
Project Number: 3093-10700-001-014-A
Project Type: Cooperative Agreement
Start Date: Aug 15, 2026
End Date: Aug 14, 2029
Objective:
Early life nutrition is a critical contributor to life-long health. Improving maternal nutrition prior to conception has been shown to improve health outcomes for mother and child. This collaborative project aims to understand the associations between maternal nutritional and systemic inflammatory biomarkers during pregnancy on perinatal and neonatal health, which are primary drivers of chronic disease later in life.
Approach:
This agreement will facilitate collaborative research to address the links between maternal nutrition, precision biomarkers in pregnancy and the health of the child. The cooperator has conducted one of the few studies to date where preconception maternal nutrition was enhanced via food and supplement-based methods. There is an urgent need to discover precision, early-pregnancy predictors of fetal and childhood growth failure by establishing a comprehensive, multiomic framework. By evaluating how systemic maternal inflammation associated with factors like nutritional status, metabolic stress, and environmental insults, impact development, the project seeks to determine if a specific maternal inflammatory endotype can independently predict a child’s growth trajectory from birth through two years of age. To achieve this, the investigation will leverage maternal serum samples, clinical metadata, and birth outcome data from two existing maternal-child cohort studies. Methodologically, the team will combine elemental analysis, high-throughput proteomics, and both targeted and untargeted metabolomics to map maternal nutrient levels, cardiovascular markers, and inflammatory profiles at weeks 12 and 34 of pregnancy. Data will be integrated with existing clinical data and placental gene expression data using advanced biostatistics and machine learning. Specifically, the pipeline will use network analysis, regression modeling, and unsupervised clustering to distill these complex biological signals into a unified Maternal Inflammatory Burden Score and identify distinct maternal gestational endotypes. Finally, predictive machine learning classifiers (such as Random Forest or XGBoost) will be trained to identify infants at the highest risk for postnatal stunting. Identifying these precise, data-driven maternal endotypes will help uncover biological pathways linking maternal stress and perinatal outcomes.