Subjects and study design

Approval from the Research Ethics Committee in Brazil was granted for the protocol and informed consent process of this study (no. 58301416.4.0000.5385).

This cross-sectional study included adolescents aged 12 years and 15 years from Salvador, Bahia, Brazil. The sample size was calculated, considering a 5% significance level, 90% test power, and 1.5 effect size, obtaining a minimum of 1054 patients. Probabilistic sampling was performed by conglomerates. The distribution of 12- and 15-year-old adolescents in each administrative district of Salvador was determined from information provided by the Municipal Secretary of Education. The sample was stratified according to the administrative district, and in the first phase, schools were selected using a simple randomization procedure. Similarly, in the second phase, adolescents were selected for the sample using a simple randomization procedure. The eligibility criteria were adolescents without current or previous orthodontic treatment and without physical or intellectual limitations that would prevent them from undergoing the examination. Adolescents with physical limitations, cleft lip, and palate, or any form of congenital craniofacial deformities were excluded from the study. Moreover, the sample was increased by 10% to compensate for possible dropouts, leading to a total of 1172 adolescents aged 12 years (n = 638) and 15 (n = 534) years.

A structured questionnaire was used to collect data on sex and age. The self-perception of orthodontic treatment need was determined by the IOTN-AC and by the OASIS. ^,

The IOTN-AC was used to assess the self-perception of orthodontic treatment need using a dental attractiveness scale illustrated by 10 colored photographs presenting a decreasing and continuous attractiveness level, in which image 1 represented the most attractive dental arrangement and image 10 was the least attractive. The adolescents themselves performed the assessments, identifying the level of esthetic compromise in the scale images considered similar to their own smile. Images 1 to 4 were related to patients with no or little orthodontic treatment need (grades 1 and 2); images 5, 6, and 7 showed moderate orthodontic treatment need (grade 3); and images 8, 9, and 10 showed severe orthodontic treatment need (grades 4 and 5).

To assess the perceived esthetic impact of malocclusion, the OASIS was used. The OASIS includes 5 questions with 5 answer options on a 7-point Likert scale, and a higher OASIS score indicates a higher esthetic concern. ^, The results of the OASIS were completed with the perceived need for orthodontic treatment scored by the adolescents using the IOTN-AC. ^,

The normative orthodontic treatment need was determined by the IOTN-DHC and by the DAI. The data were collected through clinical examination performed in the school environment, under natural light, using a World Health Organization probe. The researcher was calibrated, obtaining a weighted kappa higher than 0.97 for both normative indexes (IOTN-DHC and DAI).

The IOTN-DHC classifies malocclusions according to the severity of the occlusal characteristics in 5 grades: grades 1 and 2 represent no or little need, grade 3 shows moderate need, and grades 4 and 5 show a definite treatment need. The IOTN-DHC was dichotomized into no treatment need (IOTN grades 1 and 2) and treatment need (grades 3, 4, and 5). The DAI classifies malocclusion in 4 categories based on severity and treatment need: no malocclusion or mild change/no treatment need (DAI 1, 25), established malocclusion and/or optional treatment need (DAI 2, 26-30), severe malocclusion and/or highly desirable treatment need (DAI 3, 31-35), and debilitating malocclusion and/or mandatory treatment need (DAI 4, 36). The DAI was dichotomized into DAI 1 for no orthodontic treatment need and DAI higher or equal to 2 for orthodontic treatment need.

Statistical analysis

The self-perception of orthodontic treatment need was considered the outcome variable and dichotomized as without esthetic impact or with esthetic impact. To meet the analysis proposal, IOTN-AC was dichotomized as without esthetic impact (low self-perception of orthodontic treatment need; grades 1 and 2) or with esthetic impact (high self-perception of orthodontic treatment need; grades 3, 4 and 5). OASIS was dichotomized by the median of the scores as without esthetic impact (≤14) or with esthetic impact (>14). Dichotomization was used to discriminate adolescents who reported esthetic perception of malocclusion from adolescents who did not report such perceptions.

Descriptive analyses were performed, considering the outcome (IOTN-AC and OASIS) for the explanatory (independent) variables (IOTN-DHC and DAI, age, and gender). Multiple logistic regression was performed to detect possible associations between the independent variables and the self-perception of orthodontic treatment need (outcome). The crude odds ratios were estimated from the coefficients of the individual models of each variable, with their respective 95% confidence interval (CI). The variables with P ≤0.20 in the individual analyses were tested in multiple logistic regression models, maintained the variables with P ≤0.05. Then, the adjusted odds ratios were estimated in the multiple models. All analyses were performed using the SAS (SAS Institute Inc, Cary, NC, 2010) and R (R Foundation for Statistical Computing, Vienna, Austria) programs.