Construction of an early childhood caries risk prediction model based on the oral microbiome: a nested case‒control study.

backgroundResearch on dental caries risk prediction models that integrate socio-demographics, oral health behaviors, and oral microbiota remains insufficient. The present study aimed to compare differences in socio-demographics, oral health behaviors, and oral microbial diversity among preschoolers with different levels of ECC progression and to construct a caries risk prediction model to prevent the occurrence and development of ECC in young children.

methodsA nested case-control study was conducted in a prospective cohort study to follow up preschoolers from 36 months to 48 months of age. Dental caries status and supragingival plaque data were collected at each follow-up visit, and questionnaire surveys of their mothers were conducted to collect information on the socio-demographic backgrounds and oral health-related behaviors of the preschoolers. High-throughput 16 S rRNA gene sequencing technology was used to observe the oral microbial diversity in children, and an ECC risk prediction model was constructed using the random forest method.

resultsThere were 18 preschoolers each in the caries-free to caries-free group (H2H), caries-free to decayed group (H2C) and decayed to decayed group (C2C) according to ECC progression. The alpha diversity analysis showed statistically significant differences in the Simpson and Shannon indices between the H2H and H2C groups (P < 0.05), and the beta diversity analysis revealed statistically significant differences between the H2H group and the H2C and C2C groups. Therefore, an ECC risk prediction model was constructed based on the H2H and H2C groups. The results of risk prediction model showed that place of residence, maternal completion of oral hygiene, toothbrushing frequency and use of dental floss, and three oral microorganisms, including Selenomonas sp. oral taxon 137, Porphyromonas sp. oral taxon 278 str. W7784, and the human oral bacterium C73, were risk factors for ECC progression after one year, and the area under the curve of the model was 95.00% (95% CI, 86.50-100.00%).

conclusionThe oral microbial diversity of consistently caries-free children differed considerably from that of children who eventually developed decay, and oral health behaviors such as toothbrushing frequency and some oral microorganisms had good ECC predictive ability.