Nutrient-toxic element mixtures and the early postnatal gut microbiome in a United States longitudinal birth cohort

• Published in: Environment international Vol. 138; p. 105613
• Main Authors: Laue, Hannah E., Moroishi, Yuka, Jackson, Brian P., Palys, Thomas J., Madan, Juliette C., Karagas, Margaret R.
• Format: Journal Article
• Language: English
• Published: Netherlands Elsevier Ltd 01.05.2020; Elsevier
• Subjects: 16S rRNA gene; Bayes Theorem; Bayesian kernel machine regression; Child; Cohort Studies; Elemental nutrients; Gastrointestinal Microbiome; Humans; Infant; Infant gut microbiome; Metals/metalloids; Mixtures; New Hampshire; Nutrients; RNA, Ribosomal, 16S - genetics; United States; United States; New Hampshire; Elemental nutrients; Infant gut microbiome; Bayesian kernel machine regression; Metals/metalloids; Mixtures; 16S rRNA gene
• ISSN: 0160-4120; 1873-6750
• DOI: 10.1016/j.envint.2020.105613

•Aim was to examine mixture of nutrient and toxic elements and the infant microbiome.•Arsenic and antibiotics were synergistically associated with lower diversity.•Found associations with Bifidobacterium and zinc (negative) and arsenic (quadratic). The infant microbiome contributes to health status across the lifespan, but environmental factors affecting microbial communities are poorly understood, particularly when toxic and essential elements interact. We aimed to identify the associations between a spectrum of other early-postnatal nutrient or toxic elemental exposures measured and the infant gut microbiome. Our analysis included 179 six-week-old infants from the New Hampshire Birth Cohort Study. Eleven elements were measured in infant toenail clippings. The gut microbiome was assessed using 16S rRNA V4-V5 hypervariable region targeted sequencing. Multivariable zero-inflated logistic normal regression (MZILN) was used to model the association between element concentrations and taxon relative abundance. To explore interactive and nonlinear associations between the exposures and specific taxa we employed Bayesian Kernel Machine Regression (BKMR). Effect modification by delivery mode, feeding mode, peripartum antibiotic exposure, and infant sex was assessed with stratified models. We found a negative association between arsenic and microbial diversity in the full population that was accentuated among infants exposed to peripartum antibiotics. Arsenic, cadmium, copper, iron, lead, manganese, nickel, selenium, tin, and zinc were each associated with differences in at least one taxon in the full study population, with most of the related taxa belonging to the Bacteroides and Lactobacillales. In stratified analyses, mercury, in addition to the other elements, was associated with specific taxa. Bifidobacterium, which associated negatively with zinc in MZILN and BKMR models, had a quadratic association with arsenic concentrations. These associations varied with the concentration of the other element. Early postnatal toxic and nutrient elemental exposures are associated with differences in the infant microbiome. Further research is needed to clarify the whether these alterations are a biomarker of exposure or if they have implications for child and lifelong health.