Effects of in utero exposure to multiple metal(loid)s on neonatal birth size and hormones in the growth hormone/insulin-like growth factor axis: A mixture analysis

Abstract

Metal(loid)s are widespread in the environment, and closely linked to adverse birth outcomes. However, evidence linking exposure to metal(loid) mixtures to neonatal growth is limited and the effects of multiple metal(loid)s on hormones in the growth hormone (GH)/insulin-like growth factor (IGF) axis remain unclear. This study aimed to explore the relationships between in utero exposure to metal(loid) mixtures and neonatal birth size and GH, IGF-1, and IGF-binding protein 3 (IGFBP-3) within the GH/IGF axis. We assessed 16 metal(loid) levels in umbilical cord blood collected from 224 newborns in Guangzhou, China, and retrieved their anthropometric data from medical records. Multiple linear regression (MLR) models were employed to assess the effects of single-metal(loid) exposure. Bayesian kernel machine regression (BKMR) and Quantile g-computation (Qgcomp) models were utilized to examine the joint effects of metal(loid) mixtures. We observed that exposure to arsenic (As) and mercury (Hg) exhibited a significantly negative correlation with neonatal birth weight in MLR models. Each ln-unit increase in As and Hg reduced birth weight by 59.52 g (95 % CI: -101.71, -17.33) and 91.69 g (95 % CI: -146.21, -37.17), respectively. In addition, selenium (Se) was significantly positively correlated with IGFBP-3. With each ln-unit increase in Se, the levels of IGFBP-3 increased by 41.77 % (95 % CI: 24.76 %, 61.10 %). The associations of As and Hg with birth weight, and Se with IGFBP-3 levels remained statistically significant in the BKMR model, when other metal(loid)s were held at their 25th to 75th percentiles. Furthermore, both the BKMR and Qgcomp models demonstrated a significantly inverse combined effect of 16 metal(loid)s on neonatal birth weight and ponderal index, along with a positive association between metal(loid)s exposure and IGFBP-3 levels. These findings provide substantial evidence for both the individual and joint effects of various metal(loid)s on neonatal birth size and hormones in the GH/IGF axis.

Keywords: Bayesian kernel machine regression; GH/IGF axis; Joint effect; Metal(loid)s; Neonatal growth; Quantile g-computation.