Association of ultrasound-based measures of fetal body composition with newborn adiposity

Abstract

Background: Newborns exhibit substantial variation in gestational age-adjusted and sex-adjusted fat mass proportion. The antecedent characteristics of fetal body composition that are associated with newborn fat mass proportion are poorly understood.

Objective: The aim of this study was to determine whether a composite measure of fetal fat mass is prospectively associated with newborn adiposity.

Methods: In a longitudinal study of 109 low-risk pregnancies, fetal ultrasonography was performed at approximately 12, 20 and 30 weeks gestation. Estimated fetal adiposity (EFA) was derived by integrating cross-sectional arm and thigh per cent fat area and anterior abdominal wall thickness. Newborn per cent body fat was quantified by Dual Energy X-Ray Absorptiometry. The association between EFA and newborn per cent body fat was determined by multiple linear regression.

Results: After controlling for confounding factors, EFA at 30 weeks was significantly associated with newborn per cent body fat (standardized β = 0.41, p < 0.001) and explained 24.0% of its variance, which was substantially higher than that explained by estimated fetal weight (8.1%). The observed effect was driven primarily by arm per cent fat area.

Conclusions: A composite measure of fetal adiposity at 30 weeks gestation may constitute a better predictor of newborn per cent body fat than estimated fetal weight by conventional fetal biometry. Fetal arm fat deposition may represent an early indicator of newborn adiposity. After replication, these findings may provide a basis for an improved understanding of the ontogeny of fetal fat deposition, thereby contributing to a better understanding of its intrauterine determinants and the development of potential interventions.

Keywords: fetal body composition; fetal fat mass; newborn adiposity; per cent body fat.

Figure 1

Scatterplot and the regression line with 95% CI depicting the association between EFA at 30 weeks gestation and newborn percent body fat (N = 109, r = 0.457, p < 0.001) EFA, estimated fetal adiposity.

Figure 2

Post hoc analysis of the relative contributions of fetal arm and thigh percent fat area and anterior abdominal wall thickness at 30 weeks gestation explaining variation in newborn percent body fat^† *p < 0.05, **p < 0.001. ^†Adjusted for maternal age, parity, interaction term of pre-pregnancy BMI and maternal gestational weight gain per week, gestational diabetes, gestational age at birth, and postnatal age at DXA scan.

Figure 3

Estimated ROC curves for best fitting predictive models of high newborn percent body fat, defined as ≥75^th sample percentile. The best fitting model included arm percent fat area and anterior abdominal wall thickness at 30 weeks gestation, maternal age, and pre-pregnancy BMI (red line). The comparison model excluded arm percent fat area and anterior abdominal wall thickness at 30 weeks gestation (blue line). Models were chosen using 5-fold cross-validation and reported AUC estimates and 95% confidence intervals are based upon the cross-validated samples. wk, week; EFA, estimated fetal adiposity.