Abstract:

Early childhood caries（ECC） is one of the most prevalent chronic pediatric diseases worldwide，posing significant challenges to children’s health. Recent advances in multi-omics technologies have provided novel insights into the structural and functional characteristics of ECC-associated oral microbial communities. Research evidence confirms that ECC development is closely linked to oral microbial imbalance，characterized by changes in dominant bacterial species，altered diversity，and shifted functional expression，revealing dynamic microbial changes and host-microbe interactions during disease progression. These findings demonstrate how oral microbiome dysbiosis drives ECC through structural and functional alterations in the microbial community. Machine learning has enhanced high-throughput data analysis，which further advances ECC prediction models with integrated models combining microbiome features and host factors demonstrating superior predictive accuracy. However，current research still faces limitations including insufficient sample sizes and limited model generalizability. Future directions should focus on expanding microbial community profiling to understudied members，optimizing multi-omics data integration through systems biology approaches，and developing ultrasensitive detection methods for low-abundance biomarkers，which are all critical for ECC precise prediction and personalized prevention.

Key words: early childhood caries, oral microbiome, prediction model, multi-omics studies, high-throughput sequencing