Quantitative Study on the Correlation Between Fetal Y Chromosome Concentration in NIPT and Maternal Indicators Based on Multiple Regression and Quantile Regression, and the Optimal Detection Timing

Abstract

This study focuses on the quantitative association analysis between fetal Y chromosome concentration and maternal indicators, and the determination of the optimal timing for non-invasive prenatal testing (NIPT) based on BMI grouping. The aim is to optimize NIPT strategies and minimize potential risks to pregnant women. Regarding the association characteristics between Y chromosome concentration and maternal indicators such as gestational age and BMI, Pearson correlation analysis first revealed a negative correlation between Y concentration and BMI (-0.1683) and a positive correlation with gestational age (0.09842), though both linear relationships were weak. Building upon this, a multivariate polynomial regression model incorporating quadratic and interaction terms was constructed, with parameters fitted using least squares. F-test validation confirmed the model's overall statistical significance (p<0.001). To determine the optimal testing time point, considering that K-means clustering might cause sample size imbalance, this study employed quartile grouping to reasonably categorize the BMI of pregnant women carrying male fetuses, ensuring an even distribution of sample sizes across groups. Subsequently, a quantile regression model was established with the objective of minimizing the potential risk to pregnant women. The results revealed optimal NIPT testing timepoints of 12.59 weeks, 12.87 weeks, 13.06 weeks, and 13.31 weeks for the four BMI groups, respectively. Detection error increased with rising BMI.