Quantitative Study on Factors Influencing Male Fetal Y Chromosome Concentration Based on Linear Mixed-Effects Models

Abstract

This study aims to quantify the association between fetal Y chromosome concentration and key indicators such as maternal gestational age and body mass index to inform non-invasive prenatal testing strategies. First, fetal testing data underwent rigorous preprocessing, including exclusion of concentrations below 4%, removal of missing and outlier values, and conversion of gestational age formats to continuous weeks suitable for continuous numerical variables. Exploratory data analysis revealed a wide range of fetal Y chromosome concentration values at identical or similar gestational ages, indicating stratified clustering and nested structures—significant inter-individual variation among pregnant women. Consequently, a linear mixed-effects model with random intercepts was constructed, incorporating gestational age, BMI, and their interaction term as fixed effects, while treating individual variation as random intercepts. The model fitted 541 observations from 236 pregnant women. The random intercept variance was significantly non-zero, confirming substantial individual variation influencing concentration variability. Significance tests for fixed effects revealed a significant negative main effect of BMI on Y chromosome concentration, along with a significant interaction effect between gestational week and BMI, confirming that BMI amplifies the influence of gestational week on Y chromosome concentration. The model achieved a conditional coefficient of determination of 0.725, but the marginal coefficient of determination was only 0.117, suggesting the need to incorporate additional important fixed effect factors.