A Study on Factor Analysis and Timing Optimization for NIPT in High-BMI Pregnant Women Based on Linear Mixed-Effects Models and Dynamic Programming

Abstract

Non-invasive prenatal testing (NIPT) is a pivotal prenatal screening tool, yet its accuracy for male fetuses depends on achieving a sufficient fetal Y chromosome concentration (fYCC), which is influenced by individual factors like maternal BMI and gestational age (GA). This study aims to resolve the clinical conflict between early detection and reliable results by optimizing the NIPT timing for high-BMI pregnant women. First, a linear mixed-effects model (LMM) was employed to identify key factors affecting fYCC, revealing significant positive correlations with GA, X chromosome concentration, and number of blood draws, and a negative correlation with BMI. Building on these findings, a personalized BMI grouping strategy was developed using a hierarchical risk mapping function optimized via a dynamic programming (DP) algorithm. This approach defined five optimal BMI intervals and determined the recommended testing time as the 90th percentile GA for fYCC attainment within each group. Error analysis demonstrated the scheme's high robustness with a detection error (σ) of ≤1%. The proposed model fusion framework provides a data-driven solution for personalized NIPT scheduling, effectively minimizing risk for high-BMI individuals.