Factors Influencing Dental Caries and Oral Microbiota Among Children and Youth With Special Health Care Needs

Objectives To compare caries associated factors, salivary IgA and oral microbiota between children and youth with special health care needs (CYSHCN) and the typically developed group (TD). Methods A cross‑sectional study included 380 participants (184 CYSHCN, 196 TD) aged 6-18 years. Parents or guardians completed a questionnaire, and oral examinations were performed on the children. Salivary IgA was measured by ELISA in 250 participants. Dental plaque samples from 323 participants underwent 16S rRNA sequencing. Statistical analyses included logistic/quantile regression, α / β diversity, LEfSe and random forest. Results CYSHCN had higher caries prevalence (74.3% vs. 61.7%, P <.001) and mean DMFT/dmeft (3.83 vs. 2.42, P <.001) than TD. In CYSHCN, univariate analysis identified late-night habits, female sex, and poorer guardian perception as associated with high caries; only female sex and guardian perception remained significant in multivariate analysis. Quantile regression showed an inverse IgA–caries association only at the 0.1 quantile, where very low caries group had higher IgA than high caries group. CYSHCN had higher absolute microbial abundance but lower microbial health index. 33 differential taxa were identified, with Campylobacter concisus distinguishing very low caries subgroups. A 14-species random forest model achieved an area under the curve of 0.91 (0.78–0.92) for distinguishing CYSHCN from TD. Conclusion CYSHCN showed a higher caries burden than TD controls. Female sex and guardian perception were associated factors; late-night habits were significant only in univariate analysis. An inverse IgA-caries association was observed only within a low-IgA subgroup of CYSHCN. Oral microbiota differed between groups, and an exploratory microbiota-based model distinguished CYSHCN from TD. However, given CYSHCN heterogeneity and lack of individual diagnostic data, these findings are preliminary and require validation in diagnostically homogeneous cohorts.