Multimodal Data Integration Reveals Mode of Delivery and Snack Consumption Outrank Salivary Microbiome in Association With Caries Outcome in Thai Children

Abstract

Early childhood caries (ECC) is not only the most common chronic childhood disease but also disproportionately affects underserved populations. Of those, children living in Thailand have been found to have high rates of ECC and severe ECC. Frequently, the cause of ECC is blamed on a handful of cariogenic organisms, such as Streptococcus mutans and Streptococcus sobrinus. However, ECC is a multifactorial disease that results from an ecological shift in the oral cavity from a neutral pH (~7.5) to an acidic pH (<5.5) environment influenced by the host individual's biological, socio-behavioral, and lifestyle factors. Currently, there is a lack of understanding of how risk factors at various levels influence the oral health of children at risk. We applied a statistical machine learning approach for multimodal data integration (parallel and hierarchical) to identify caries-related multiplatform factors in a large cohort of mother-child dyads living in Chiang Mai, Thailand (N=177). Whole saliva (1 mL) was collected from each individual for DNA extraction and 16S rRNA sequencing. A set of maternal and early childhood factors were included in the data analysis. Significantly, vaginal delivery, preterm birth, and frequent sugary snacking were found to increase the risk for ECC. The salivary microbial diversity was significantly different in children with ECC or without ECC. Results of linear discriminant analysis effect size (LEfSe) analysis of the microbial community demonstrated that S. mutans, Prevotella histicola, and Leptotrichia hongkongensis were significantly enriched in ECC children. Whereas Fusobacterium periodonticum was less abundant among caries-free children, suggesting its potential to be a candidate biomarker for good oral health. Based on the multimodal data integration and statistical machine learning models, the study revealed that the mode of delivery and snack consumption outrank salivary microbiome in predicting ECC in Thai children. The biological and behavioral factors may play significant roles in the microbial pathobiology of ECC and warrant further investigation.

Keywords: diet; early childhood caries; machine learning; multimodal analysis; oral microbiome; saliva.

Figure 1

Alpha Diversity of salivary microbiome among children. Microbial variation measured by alpha diversity index among children from different age group (A), mode of delivery (B), Bottle feeding (C), and Consumption of snacks (D). T-test was used for the statistical comparisons.

Figure 2

Diversity of salivary microbiome among children. Principle coordinate analysis (PCOA) plot is generated using OTU metrics based on beta diversity (Bray-Curtis index) for different sex groups (A), Age (B), Mode of delivery (C), Mother chew food for child (D), Child caries severity dmft (E), and ECC status (F). Permutational MANOVA (PERMANOVA) was used for these statistical comparisons between or among the categorical groups.

Figure 3

Core salivary microbiome of ECC and carries-free children. Taxa at the genus and species level with more than 20% prevalence and more than 0.01% relative abundance are depicted. in caries free children (A1, B1) and ECC children (A2, B2).

Figure 4

Heat trees of salivary microbiome abundance among ECC and caries-free children. The heat trees depict the OTU classifications and differential abundance comparison at the genus level (A) and at the species level (B) between salivary microbiome of children with ECC and without ECC. In the heat trees, size and color of nodes and edges are correlated with the abundance ratio of organisms in ECC children vs. caries-free children. Taxa colored in red are enriched in ECC Children, whereas taxa colored in blue are enriched in caries-free children. Taxa with labels indicate a significant abundance difference between ECC children vs. caries-free children measured by the Wilcoxon Rank Sum test (p<0.05).

Figure 5

Taxa at genus level differently enriched in ECC and caries-free children. (A) Linear discriminant analysis (LDA) effect size (LEfSe) method was performed to compare taxa between ECC and caries-free children. The bar plot lists the significantly differential taxa based on effect size (LDA score log₁₀ >2.0 and FDR <0.1). (B) Random forest identified important features at the species level that were differently enriched among ECC and caries-free children. Red indicates a higher abundance in ECC, whereas blue indicates a higher abundance in caries-free children.

Figure 6

Identified factors associated with child’s caries risk using factors via two-step model. LASSO penalized logistic regression modeling was used for caries predictor selection based on 413 variables, including children’s saliva samples. Specifically, variables from three separate platforms were identified shown in (A) Maternal socio-demographic-behavior-environmental factors, (B) Children’s socio-demographic-behavior-environmental factors, and (C) Children’s salivary microorganisms. The final two-step model using variables identified from (A–C) is shown in (D). The LASSO solution path above shows how the model is built sequentially by adding one variable at a time to the active set. The 2-step predictive model is the following (area under the curve: 0.85).

Figure 7

Conceptual causal inference from multiplatform variables assessed in the study.