Introduction

Bullying in the school environment has been the central theme of studies in various countries, since this phenomenon can be extremely harmful to the individuals involved, leading to serious medium and long-term social consequences. There is evidence that the discrimination is associated with later onset of physiological stress, as well as psychosis ^1,2. Bullying can, therefore, be related to physical and emotional problems, such as eating disorders, depression, suicidal ideation or even suicide itself ^3,4.

Several factors have been associated with bullying victimization, including drug use and/or cigarette ^5,6; low motor skills; sexual orientation ^7,8; presence of a chronic disease ⁹; disabilities ¹⁰; deafness or hearing loss ¹¹; socioeconomic factors ¹²; income inequality ¹³, discrimination based on ethnicity ¹⁴, and educational background of the parents ¹⁵. The self-perception of being fat or thin, or the overweight/obesity condition often stands out amongst the factors associated with violence and bullying ^16,17.

A meta-analysis study conducted by van Geel et al. ¹⁸ identified that overweight/obese adolescents are more likely to be victims of bullying. In Brazil, according to the data extracted from the Brazilian Household Budget Survey conducted between 2008 and 2009, the prevalence of overweight/obesity among 10-19-year-old adolescents was 20.5% ¹⁹. However, no nationwide research was found when it comes to the association between bullying and nutritional status in this age group. A recent study, carried out in Southeastern Brazil, identified that 7.8% of the adolescents reported being bullied in the last 30 days, and most of them were black boys, children under 13 years, children of mothers with lower levels of education, and students of private schools ²⁰.

This study explored discrimination by social, cultural, economic, and anthropometric variables and analyzed the victimization of individuals who are discriminated against those who are not. The aim of this research was to estimate the prevalence of bullying and to verify its association with nutritional status, demographic and economic factors, as well as individual attributes among schoolchildren aged 11 to 14 years.

This study seeks to fill the gap and increase scientific knowledge of the factors associated with bullying by adopting intervention strategies to reduce discrimination against obesity, thinness, poverty, ethnicity, and other individual attributes.

Methods

Results

This study included 975 participants aged between 11 and 14 years; 635 (65%) of them were attending public schools and 340 (35%), private schools. Among the adolescents who participated in the research, 54.3% were girls aged 11 years (31%). The economic condition variable indicated that 27.2% of the adolescents have household monthly income below USD 650.00 and 24.5% of them have income greater than USD 1,861. Boys had higher prevalence of overweight/obesity (33.5%), and, only for this variable, we observed a difference between sexes (p = 0.005). The prevalence of bullying was 12.3% for boys and 14% for girls. Out of the total sample, 13.2% have been bullied (Table 1).

In the univariate analysis, the variables age, sex, school type, education of the mother and household income were not associated with bullying (victimization). An association between BMI and bullying was found among adolescents considered overweight/obese (OR = 1.44, 95%CI: 1.03; 2.02) (Table 2). After performing the univariate analysis, the variables anthropometric (overweight/obesity) and individual attributes remained with p-value < 0.20, and they were included in the hierarchical model of multivariate analysis. The variables that remained significant in the hierarchical model were the ones related to individual attributes.

In the adjusted logistic regression model for household income, the odds ratio of bullying was 18.60 higher in boys discriminated for being “fat” when compared with those who have not being bullied for this specific reason (OR = 18.60, 95%CI: 2.80; 123.45, p = 0.002). In addition, “good-looking” or “ugly”, “disabled”, and “tall” boys were more likely to report discrimination (p < 0.05) (Table 3).

As for the girls, we observed they had higher odds of being victims of bullying for being “fat” (OR = 5.80, 95%CI: 1.24; 27.20, p = 0.026). “Tall”, “short”, “ugly”, or “poor” girls were more likely to report having been bullied (p < 0.05) (Table 4).

Adolescents of both sexes discriminated for being “fat”, “ugly” or “tall” had higher odds of being considered victims of bullying at different levels. The adolescents who suffered discrimination due to “ethnicity” and “weight” showed no significant adjustment for both boys and girls. Discrimination for being “poor” and “other characteristics” remained associated with a higher prevalence of peer victimization only for girls (Tables 3 and 4).

Discussion

The main results of this study show the prevalence of victims of bullying and overweight/obesity was 13.2% and 29%, respectively. In the crude analysis model, the association between bullying and overweight/obesity and the variables called individual attributes (fat, thin, tall, short, good-looking, ugly, rich, poor, regarding ethnicity or disabilities and/or others) indicated that the adolescents discriminated for having such attributes had a greater chance of being victims of bullying. In the analysis model adjusted by household income and stratified by sex, we verified that the individual attributes associated with bullying were different amongst boys (fat, good-looking, ugly and with a disability) and girls (fat, tall, short, ugly, rich, poor, and others). This study found no association between bullying and the investigated variables such as age, sex, school type, education of the mother, household income, and overweight/obesity.

The prevalence of bullying in studies conducted in the United States, and in European, Asian and Latin American countries, including Brazil, varied from 7.4% ²⁹ to 72% ³⁰. In relation to this study, we found a higher prevalence of bullying in nine other studies ^{5,30,31,32,33,34,35,36,37}, a similar prevalence in three other studies ^38,39,40, and lower prevalence in two other studies ^29,41.

These variations may be explained by the different methodological procedures used in the studies and the criteria adopted for bullying classification, in addition to the explanatory model of the phenomenon, given that it is a multifaceted construction comprised of different domains. Distinct geographic and temporal contexts, differences in age range and size of the samples investigated can also be considered explanatory factors for the wide variation among studies. Therefore, we highlight the difficulty in discussing more broadly and accurately our results in relation to the literature, due to the different instruments and methodological procedures used to measure both the outcome variable (bullying) and the exposure variable (overweight/obesity).

Even though sex was not associated with prevalence of bullying, some specific behaviors were observed for boys and girls, such as the association of frequent discrimination for being “short”, “poor” and/or “ugly”, which only remained significant in girls after adjustment. These results presented a different behavior from those observed in other studies. One cross-sectional study conducted in Italy ⁴² with children and adolescents aged 6 to 14 years identified that obese individuals were more involved in bullying episodes (information collected through a questionnaire answered by the schoolchildren themselves). In addition, severely obese participants experienced verbal and physical victimization and exclusion from group activities ⁴². For the diagnosis of obesity, this Italian study used the cut-off points of BMI recommended by the Centers for Disease Control and Prevention (CDC). The BMI standard deviation score (BMI-SDS) was employed to define normal-weight (BMI-SDS between -1 and +1); overweight (BMI-SDS between 1.01 and 1.63), moderate obesity (SDS-BMI between 1.64 and 2.5); and severe obesity (BMI-SDS > 2.5) ⁴².

A cohort study conducted in England ⁴³ found that bullying episodes have become more frequent over the years when it comes to the English population. This study identified that, at the age of eighteen, individuals intimidated by others are more likely to be overweight than those who are not bullied. Such association was even stronger in chronically intimidated individuals. In addition to the longitudinal design (follow-up of children from birth to 18 years of age), this English study obtained the information on victims of bullying from questionnaires applied with mothers and with the schoolchildren themselves. For the diagnosis of overweight/obesity, they used the cut-off points of the BMI recommended by the CDC (BMI equal to or above the 85th percentile were classified as overweight), as well as other measures, such as waist-hip ratio and a visual assessment of weight performed subjectively by the examiners. Such evaluations were performed at different times, in particular, at the ages of 7, 10 and 12 years ⁴³, age groups that resemble this study.

Investigations have shown that prevalence of bullying has a specific behavior according to the sex and depends on other physical or mental characteristics. Also, the victims are usually individuals with low self-esteem, who are shy or emotionally fragile, especially girls who deviate from the aesthetic standards imposed by society ^{1,2,5,10,18,31,35,37,41,42,44}.

In this study, both sex and physical appearance (measured by individual variables) were a determinant of bullying in the adolescents investigated, which was also observed in other studies ^12,45,46,47. Furthermore, by revealing the association between being discriminated and being bullied, for both sexes, the findings show that this phenomenon is a risk factor, particularly in societies that value an ideal body based on weight, which may lead to discrimination in various contexts ^48,49.

Our results showed that adolescents of both sexes discriminated for being fat, thin, tall, short, good-looking, poor or other characteristics are more likely to being bullied. These findings confirm that, within the peer victimization process, individuals who are judged by their individual characteristics show a greater predominance of negative actions in relation to their colleagues, thus leading to conflicts related to culturally valued standards in which differences are not well-accepted ^{1,2,3,4,5,6,7,8,9,10,11,12,13,14,15}.

Regarding the factors associated with bullying victimization in adolescents, a few studies carried out in Brazil in the last ten years must be mentioned. Several studies using data from the Brazilian National Survey of School Health (PeNSE) have made it possible to monitor the pattern of bullying victimization and associated factors in adolescents in the 9th year of primary education in the Brazilian capitals. In these studies, the prevalence of bullying among the adolescents investigated was 5.4%, 7.2%, and 7.4% in the years 2009, 2012, and 2015, respectively, showing a pattern of victimization related to percentage increase. Regarding the factors investigated, such studies have pointed out that black boys younger than 13 years, as well as indigenous adolescents from public school, children of mothers with a lower level of schooling, lonely adolescents who suffer from insomnia, physical aggression from their relatives, who missed classes without warning their parents, as well as smokers ^50,51,52.

An interesting analysis using PeNSE 2012 data found that 51.2% among the causes of bullying could not be specified. The second highest frequency of victimization was related to physical appearance (18.6%), followed by good looks (16.2%), race/color (6.8%), sexual orientation (2.9%), religion (2.5%), and region of origin (1.7%) ⁵³. Two other Brazilian studies with localized populations showed a prevalence of 17.6% (in students from 1st to 8th grades) ³⁷ and 29.5% (in students from 6th to 9th grade) ⁵⁴, such as those found in PeNSE. They have also pointed out a higher prevalence in boys.

This study found no association between bullying and overweight/obesity, thus contrasting with the results verified by other studies. A study conducted in China ⁵⁵ with children and adolescents between 7 and 18 years showed that the risk of bullying increased among obese students; however, it was only considered statistically significant among women. Another study, conducted in Norway ⁵⁶, with children and adolescents between 10 and 20 years, also found an association between interpersonal violence and higher BMI. A meta-analysis on the relation between weight status and bullying, published in 2014, included 14 articles (N = 55,231) and has verified that both overweight and obese youths were more likely to being bullied. The results were not moderated by sex and weight. Therefore, obese boys and girls were equally likely to be discriminated. The study concluded that both overweight and obesity are risk factors of bullying victimization ¹⁸.

For this reason, to pay attention to the individual and to the contextual factors, such as household income and discriminatory behaviors based on individual attributes, is fundamental to explain the impacts of this phenomenon.

This study has several strengths and limitations that need to be pointed out. Regarding its strengths, it is noteworthy that (1) this is a pioneering study showing the prevalence of victims of bullying through a questionnaire based on IRT ^23,24,25,26; (2) the methods adopted for sample calculation and selection used a sample with representative data on the schoolchildren population from Florianópolis; (3) the study used methodological rigor and considered a standardization of anthropometric measurements and other variables; and (4) it also used double data entry to avoid biased compilation. There are some methodological advantages indicating rigor and reliability on the results in the way they are exposed.

On the other hand, this study also presents some weaknesses. First, the cross-sectional data collection did not make it possible to infer the temporality of relationships between bullying and the exposure variables investigated. Second, the missing data for some of the variables, such as household income and educational level of the mother, among others, can be pointed out as one of the limitations of this study, which may have implications in the evidences produced and, consequently, hinder their internal validity. Third, we emphasize that the methodological procedures used to categorize the overweight/obesity variable (BMI > +1SD) and other characteristics of the age group (11 to 14 years), such as those related to sexual maturation, may have influenced the findings of this study, showing no association between overweight/obesity and bullying. Therefore, such results must be cautiously understood and other methodological procedures and variables related to the diagnosis of overweight/obesity of schoolchildren aged 7 to 14 years must be added into the analysis models in future research.

Conclusion

In this study, the prevalence of bullying and overweight/obesity in schoolchildren aged 11 to 14 years was 13.2% and 29%, respectively. No association was found between the outcome and the exposure variables for the investigated sample - age, sex, school type, mother’s education, household income, and overweight/obesity. On the other hand, in the analysis model adjusted by household income and stratified by sex, we verified that fat, good-looking, ugly, and boys with disabilities, as well as fat, tall, short, ugly, rich, and poor girls had greater chance of becoming victims of bullying.

These results reinforce the need to adopt intervention programs in schools to reduce the prevalence of overweight/obesity, as well as the discrimination behaviors related to obesity and other individual attributes. Finally, we emphasize that attributes are individual characteristics, but bullying and obesity are social problems. Even though this study brings some evidence, further research is needed to provide a better understanding of the magnitude of such phenomena.