Family and school environmental predictors of sleep bruxism in children.

Aims: To identify potential predictors of self-reported sleep bruxism (SB) within children's family and school environments.

Methods: A total of 65 primary school children (55.4% males, mean age 9.3 ± 1.9 years) were administered a 10-item questionnaire investigating the prevalence of self-reported SB as well as nine family and school-related potential bruxism predictors. Regression analyses were performed to assess the correlation between the potential predictors and SB.

Results: A positive answer to the self-reported SB item was endorsed by 18.8% of subjects, with no sex differences. Multiple variable regression analysis identified a final model showing that having divorced parents and not falling asleep easily were the only two weak predictors of self-reported SB. The percentage of explained variance for SB by the final multiple regression model was 13.3% (Nagelkerke's R² = 0.133). While having a high specificity and a good negative predictive value, the model showed unacceptable sensitivity and positive predictive values. The resulting accuracy to predict the presence of self-reported SB was 73.8%.

Conclusion: The present investigation suggested that, among family and school-related matters, having divorced parents and not falling asleep easily were two predictors, even if weak, of a child's self-report of SB.