This cross-sectional study investigated factors related to dental caries prevalence in a low caries group. The subjects consisted of 89 schoolchildren aged 12 years. The children had participated in a fluoride mouth rinse program since 5 years of age at nursery schools. Furthermore, pit and fissure sealants were applied to teeth with sticky fissures when detected at dental check-ups twice a year. Clinical examinations were conducted to obtain data on caries prevalence (dfs and DMFS). The salivary level of mutans streptococci, the salivary level of lactobacilli, and the acid production by bacteria in the plaque were measured using Dentocult SM Strip ® , Dentocult LB Dip Slide ® , and Cariostat ® , respectively. In addition, we obtained information on gender, sum of permanent tooth surfaces sealed, toothbrushing frequency, sweet snack frequency, sweet drink frequency, use of fluoridated toothpaste, and stimulated salivary flow rate. Stepwise backward logistic multiple regression analysis was used to estimate the relationship between caries prevalence and 11 total independent variables. Furthermore, we obtained information about some reasonable linear discriminators that could identify subjects with high caries experience using the final logistic multiple regression model. Thereafter, we calculated sensitivity, specificity, and Youden's index according to the classification table. As the results of stepwise backward logistic multiple regression analysis, three variables (number of sealants, the level of Cariostat R , and frequency of sweet snacks) were selected as independent variables in the final model. The level of CariostatR showed the highest odds ratio (Odds ratio : 6.87, p=0.02). The cutpoints at which Youden's index showed the highest value, sensitivity, and specificity were 76.9% and 75.7%, respectively. According to these findings, it is safe to say that we could propose that the plaque level of Cariostat ® might be the most important variable related to caries prevalence.
View full abstract