Peres MFT, Azeredo CM, de Rezende LFM, Zucchi EM, França-Junior I, Luiz ODC, Levy RB. Int. J. Public Health 2018; 63(8): 957-965.
Affiliation
Departamento de Medicina Preventiva, Faculdade de Medicina FMUSP, Universidade de Sao Paulo, Av. Dr. Arnaldo 455, 2º andar, 01246-903, São Paulo-SP, Brazil.
OBJECTIVES: To investigate the association between personal, relational and school factors with involvement in fights with weapon among Brazilian school-age youth.
METHODS: Using data from the Adolescent School-Based Health Survey 2015 (n = 102.072), we conducted multilevel logistic regression models.
RESULTS: IFW was associated with female sex (OR = 0.45), and with older age (OR = 1.15), previous involvement in physical violence (OR = 2.05), history of peer verbal (OR = 1.14) and domestic victimization (OR = 2.11), alcohol use (OR = 2.42) and drug use (OR = 3.23). The relational variables (e.g., parent's supervision) were mostly negatively associated with IFW. At the school level, attending public school and attending schools in violent surroundings were both positively associated with IFW. The intraclass correlation coefficient estimated in the empty model showed that 5.77% of the variance of IFW was at school level. When all individual- and school-level variables were included in the model, the proportional changes in variance were 61.7 and 71.55%, respectively.
CONCLUSIONS: IFW is associated with personal, relational and school factors. Part of the variance in IFW by school is explained by characteristics of the school context.