Statistical analyses

Descriptive statistics were calculated. The prevalence rates of BDD based on the full BDD questionnaire and its first 3 items were calculated separately. The sexes and BDD-positive and BDD-negative patients were compared in terms of age, marital status, history of previous orthodontic visit, IOTN, and BDD scores using unpaired t-test, chi-squared, Fisher exact, and Mann–Whitney U tests. Spearman correlation coefficient was used to estimate correlations among the variables. A multiple linear regression was used to assess the effects of IOTN, sex, age, marriage status, and education level on the overall BDD score (the sum of 12 items and that of the first 3 items). The effects of these variables on BDD-positive and BDD-negative diagnosis were also assessed using a multiple binary logistic regression. Spearman correlation coefficient was used to evaluate correlations between the 3- and 12-item BDD scores as well as between the dichotomized diagnoses based on the 3- and 12-item BDD questionnaires. The level of significance was set at 0.05 The statistical analyses were performed using SPSS version 25.0 (IBM, Armonk, NY, USA).