Abstract
The oral microbiota plays a vital role in human health, yet most studies have focused on individual bacterial taxa. To provide a more comprehensive understanding, we analyzed tongue microbiota data from 729 Japanese individuals and classified samples into three types (orotypes): Neisseria-dominant (N), Prevotella-dominant (P), and Streptococcus-dominant (S) types. Each orotype exhibited distinct co-occurrence network structures and was associated with lifestyle factors such as oral care, diet, and smoking. The S type was associated with higher odds of abnormal oral health and metabolic syndrome-related outcomes compared to the N type. In addition, we developed a robust classification model (ROC-AUC > 0.95) to predict orotypes, which showed temporal stability in nearly half of individuals over a 6-year period. These findings highlight the value of orotype classification for monitoring tongue microbial communities and suggest its potential in health risk assessment.