School academic climate and oral health (tooth loss) in adolescents

were lower for the oral examinations– 67.1% to 74.0%, depending on the outcome. The rea- sons for missing data in the oral examination were as follows: refusal to participate in the den- tal examination (dental caries + periodontal) (91), took part only in the periodontal examination (171), absent on the day data collection (160), too long a delay between the survey and oral examination (400), and not filled all variables used in this manuscript (441). The prevalence of tooth loss was 17.1% (95% CI: 15.0–19.3) and, the mean of tooth loss was 0.3 SD (0.6) per adolescent. The schools had a mean rate of participation in the National High School Examination -ENEM (2016 and, 2017) of 69.1 SD (17.2), ranging between 27.0 and 97.0. For data analysis, we used the z-score to standardize the variation in the participation between the two years (2016 and, 2017). The mean z-score for the rate of school participation in the National High School Examination among the adolescents with tooth loss was 0.0 (0.9) while it was 0.3(0.9) among adolescents with no tooth loss(p<0.001). More participants were girls (59.0%), at the older range of adolescence (16–19 vs. 14–15) (71.3%), not receiving a family allowance (53.9%), and had a mother with more than 9 years of schooling (63.7%) (Table 1). Considering the proportional change in variance (PCV), our findings showed that the reduction of heterogeneity among the schools was similar when contextual variables were added to the empty model (92.9%) versus when the individual variables were added to the

empty model (92.9%). When both contextual and individual variables were added, the hetero- geneity reached almost zero, i.e. the reduction was 99.9%. Hence, our findings suggested that the school academic climate contributed to explaining the tooth loss in adolescence and that when the compositional part was considered, all heterogeneity was accounted for. (Table 2) We also assessed the general contextual effect of school analyzing the MRR. In the empty model, the MRR was 1.42 suggesting that the difference in tooth loss among schools compar- ing similar students was 42%, i.e. a student with similar characteristics from a sorted school can be 42% more likely to have tooth loss than a student from another school. This variation in tooth loss among schools decreased in model 2 (adding contextual variables) and model 3 (adding individual variables), and in the final model 4 (both individual and contextual vari- ables) all remaining variation between schools was accounted for, i.e. MRR = 1. In model 3 (individual variables) and model 4 (individual and contextual variables), adoles- cents who presented with decayed teeth were almost twice as likely to have tooth loss, while periodontitis was not associated. Self-perception of poor oral health was also significantly cor- related with tooth loss, as well as receiving a family allowance and having a mother with low schooling. In the final model, adolescents in schools with lower rate of participation in the ENEM were more likely to have tooth loss (PR 1.42, 95%CI: 1.09,1.85) while adolescents from

neighborhoods with high/very high social vulnerability index and from regular vs. full-time/ partial full-time school were not associated. Discussion Our results corroborate our hypothesis that adolescents in a school academic climate are less likely to have tooth loss, independent of individual characteristics, such as income poverty, and low maternal educational attainment. Our findings showed a strong association between socioeconomic factors and tooth loss, including low maternal education and receiving a family allowance. The correlation between socioeconomic disadvantage and poor oral health status is firmly established [9–11]. The com- mon reasons that have been offered to explain this correlation include lack of financial resources to seek appropriate dental care, cultural norms that accept less-than-optimal denti- tion, as well as lack of health literacy. While these explanations place the responsibility on the PLOS ONE School academic climate and tooth loss in adolescents PLOS ONE | https://doi.org/10.1371/journal.pone.0233505 May 21, 2020 5 / 11