FACTORS ASSOCIATED WITH VIOLENCE IN SCHOOLS : EXTENDING KNOWLEDGES AND PRACTICES FOR NURSING 1

This study aimed to identify the factors associated with violence among students of public schools located in the central region of Rio Grande do Sul. It is an observational, transversal and analytical study. The sample was made up of 435 students aged from 10 to 19 years old. The data were collected through a structured questionnaire applied via interview. Associations between the dependent and independent variables were taken from the Poisson regression model. Once analyzed, the data evidenced greater prevalence between violence and the following variables: male sex, larger number of siblings, and a low income; while religion was associated as an indicator of protection for psychological violence. It is concluded that interdepartmental, interdisciplinary and multidimensional approaches constitute an important strategy for promoting health and reducing violence in schools, above all in communities with indicators of vulnerability. Specific strategies for the work of the nurse are indicated for preventing risks associated with violence. DESCRIPTORS: Students. Violence. Adolescent health. Community health nursing. Nursing research. FATORES ASSOCIADOS À VIOLÊNCIA EM ESCOLARES: AMPLIANDO SABERES E PRÁTICAS PARA A ENFERMAGEM RESUMO: Objetivou-se identificar os fatores associados à violência em alunos de escolas públicas localizadas na região central do Rio Grande do Sul. Estudo observacional, transversal e analítico. Amostra foi constituída por 435 alunos de 10 a 19 anos. Os dados foram coletados por meio de questionário estruturado aplicado em forma de entrevista. Associações entre as variáveis dependentes e independentes foram retiradas do modelo de regressão de Poisson. Os dados analisados evidenciaram maior prevalência entre violência e as variáveis sexo masculino, maior número de irmãos e baixa renda, enquanto que a religião foi associada como indicador de proteção para a violência psicológica. Conclui-se que as abordagens intersetoriais, interdisciplinares e multidimensionais se constituem em importante estratégia para promoção da saúde e redução da violência escolar, sobretudo em comunidades com indicadores de vulnerabilidade. Estratégias específicas de atuação do enfermeiro são indicadas para prevenção de riscos associados a violência. DESCRITORES: Estudantes. Violência. Saúde do adolescente. Enfermagem em saúde comunitária. Pesquisa em enfermagem. http://dx.doi.org/10.1590/0104-07072016004530014


INTRODUCTION
Violence is revealed to be an important component in discussions on health due to its impact on individual and collective health.It is a growing social problem, revealed through the rates of morbidity and mortality of victims of violence. 1In this perspective, violence goes beyond aggression per se, due to the consequences reflected in society, often causing fear and restrictions for the people involved in it and weaving various interrelationships which affect the whole and the parts which integrate the context characterized by violence. 23] When the violence takes place in schools, it can come to prejudice their potential for being shaped as a person and citizen, which can lead to a poor quality of social life, besides causing individual and collective suffering. 2Adolescents' exposure to a wide variety of forms of violence can lead to the development of illnesses such as depression, and other risks to their health. 4ne study 5 evidences the prevalence of violence at 18.9% in the age range from 12 to 19 years old among students at basic education * and senior high school level in public schools.The predominance was for individuals of the male sex, who used alcohol and other drugs, and whose family relationships were not satisfactory. 6Another study undertaken with 399 adolescents in an outpatient center in Pernambuco showed a rate of 41.4% of violence, within which the most frequent types were psychological and moral bullying (66.7%); within this context, 33% of the adolescents reported that the aggression was recurrent. 7The violence appears on a large scale in the school space, showing the need for broad and interdisciplinary approaches so as to address the multiple causes.This environment, which should be one of protection and strengthening of the student, can end up turning into a space of violence itself, harming physical, psychological and moral aspects. 6,8iolence, as a social event, has been revealed to be associated with various factors, 5,[7][8][9] such as: fragility of intrafamilial relationships, absence of warmth and of limits, complicated relationships with fathers or mothers, unpreparedness on the part of the school, poor educational performance, absence of things to do, drug use, conflicts in the community, use of weapons, social inequalities, and poverty, among others.
1] Vulnerable communities are understood as those which experience environmental, economic, political and cultural influences which weaken the individual, familial and social relationships and associations. 12The study of the indicators of violence and of the context of the student who lives in vulnerable communities is revealed to be a possibility for extending strategic actions for healthcare. 13n order to stimulate a new understanding of health in the light of these problems, which by their nature are complex, one requires a new way of thinking and acting -also complex.It is necessary to describe the complexity of the phenomena which emerge from the violence, for these to be analyzed under a new perspective, a perspective which weaves together the interactions, thus leading to a broadened and contextualized view of the emerging social realities. 14o this end, one requires qualified professionals, including the nurse, who are capable of inter-vening in an articulated and proactive way in the different social contexts. 15Considering the importance of identifying the factors related with violence in schools, as well as how to intervene proactively and responsibly, above all in regions with indicators of vulnerability, the question is raised: What factors are associated with violence in schools?With the objective of answering this question, the present study aimed to identify the factors associated with violence among students of public schools located in the Central Region of Rio Grande do Sul.

METHOD
This is an observational, transversal and analytical study.Its unit of analysis is the individual.It is part of the "Promotion and education for the health of children and adolescents of public schools of a vulnerable communities" project, financed by the Research Support Foundation of the State of Rio Grande do Sul (FAPERGS), approved by the FAPERGS 02/2011 -PqG bid for tender.
The study was undertaken with students aged from 10 to 19 years old who lived in a community with approximately 26,000 inhabitants, in the West Region of the municipality of Santa Maria Rio Grande do Sul (RS), Brazil.The community in question has indicators of social vulnerability, suffering unfavorable rates for its growth and strengthening such as, for example, social inequalities and difficulties in accessing the various services provided by the state.
The sample size was previously calculated through sample calculation in six state schools of small, medium and large size, located in the abovementioned community.This calculation was made based on the following formula: n school = (n total calculated X n total in the school) / 3,659 (total number in all the schools).In order to obtain the number of interviewees, the same proportional system was undertaken, considering the total number of students in each school and the total number in each classroom.
From a total of 3,659, a sample was calculated of 435 students for this study, based on an outcome of 50%, with a confidence interval of 95% and a non-response rate of 30%.The eligibility criteria included the students enrolled in the state schools selected for the study, who were residents of the above-mentioned community, and who were aged between 10 and 19 years old at the time of the research.The signing of the Terms of Free and Informed Consent, on the part of those responsible for the students, was the indispensable criteria for participating in the study.The research project was approved by the Research Ethics Committee of the Centro Universitário Franciscano (UNIFRA), under N. 285/2011.
The data were collected through a structured questionnaire, applied via interview, with the help of researchers from the health area, this data collection instrument being based on a questionnaire validated for research in a previous study on the issue of a violence among adolescents. 16Firstly, a pilot study was undertaken with 15 students from the above-mentioned schools.Based on the analysis of these initial data, some questions were changed for better adaptation of the questionnaire, due to the region's specific needs.The data were collected between October 2011 and December 2012, in six state schools of small, medium and large size, located in the above-mentioned community.
The data were presented through distribution of simple and relative frequency.Associations between the dependent and independent variables were firstly calculated using the Poisson regression model, as well as the prevalence ratios (PR) and the confidence interval of (CI) 95% being presented.The outcome variables considered were: "Have you ever suffered some type of violence?"(yes; no), "If you have suffered violence, what type was it?"(physical/sexual: psychological).The independent variables were: sex (female; male), age dichotomized by the median (≤4 years old; >14 years old), race (Caucasian; others), housing situation (living with parents; living with others), number of siblings (≤2; >2), family income dichotomized by minimum Brazilian monthly salary at the time of data collection (≤US$ 232.00; >US$ 232.00), educational level dichotomized by median (≤8 years; >8 years), time at which the student studies (daytime; night-time), whether the student has been held back a year in school (yes; no), occupation (only studies; studies and works; and religion (has a religion; does not have a religion).
The statistical level was 5% (p=0.05) and the software used was the Statistical Package for the Social Sciences (SPSS), version 18.0 for Windows.A multivariate analysis was run using the Poisson regression model, considering the following outcomes: suffering violence, and type of violence suffered.Adjusted and non-adjusted Poisson regression models with robust variance were calculated in order to ascertain the ratios of prevalence of the dependent and independent variables.An adjustment was undertaken for all the variables with a p-value <0.2 in the univariate analysis.In order to select the variables which were retained in the multivariate model, a stepwise backward procedure was adopted, eliminating, one by one, the variables with higher p values -until only variables with p<0.05 remained.For the outcome of 'type of violence', the variables of "sex" and "number of siblings" were kept in the final model, independently of the p-value.

RESULTS
The total sample was of 435 students, with a non-response rate below 10%.The predominant characteristics in the students interviewed were: female sex, family income higher than US$ 232.00, fewer than eight years of study, age equal to or less than 14 years old, and Caucasian.Among the adolescents studied, 18.9% reported having suffered some type of violence.For those who responded that they had suffered violence, 59.8% responded that they had suffered physical/sexual violence, and 40.2% had suffered psychological violence.The demographic characteristics, and characteristics related to school behavior are described in Table 1.Table 2 analyzes the indicators of demographic and school-related prevalence related to the outcome "Have you suffered any type of violence?".In the univariate analysis, male sex increased the probability of suffering violence by 1.48 times.Regarding the independent variable of "housing situation", the data collected showed that living with people who were not one's parents led to a prevalence 1.65 times greater of suffering violence.In relation to the statistical association between number of siblings and the outcome of suffering violence, it was possible to perceive that having a number of siblings greater than two increased by 114% the prevalence of the students suffering violence (p=0.00).In relation to religion, it was evidenced that not having a religion is an indicator, increasing the prevalence of suffering violence by 1.58 times.The independent variable of monthly family income showed a statistical association (p=0.00)with the outcome studied, and a monthly family income of less than or equal to US$ 232.00 showed a prevalence which was 2.11 times higher of suffering violence.After adjustment of the following variables: sex, housing situation, number of siblings, religion, monthly family income, occupation, time at which studies and having been held back a year in school for the multivariate model, housing situation and religion were retained in the final adjusted model.The statistical association with the outcome of "suffering violence" and the variables of sex (PR 1.47; CI 95% 1.02-2.12),"number of siblings" (PR 2.03; CI 95% 1.39-2.90)and "monthly family income" (PR 1.98; CI 95% 1.38-2.83)remained.The data related to the risk indicators of the outcome of type of violence suffered were described in Table 3.For the variable of "number of siblings", the students who had a number higher than two presented a probability 60% higher of suffering psychological violence than of suffering physical/ sexual violence.In relation to religion, having a religion reduced the prevalence of suffering psychological violence by 71%.For the multivariate model, the variables of sex, number of siblings, religion and being held back one year in school were adjusted.After the adjustment, the variable of "number of siblings" lost its statistical significance for the outcome of "type of violence", with only religion remaining as a prevalent indicator of protection against psychological violence (PR 0.32; CI95% 0.12-0.82)associated with the outcome of "type of violence".

DISCUSSION
Violence is, per se, a complex phenomenon, on the rise in the most differing of social contexts, as has been evidenced in various studies.One study undertaken with 699 students from the region of Barra do Garças, in the Brazilian State of Mato Grosso revealed that the prevalence of the behavior of violence among adolescents was 18.6% 6 , corroborating the present study, which showed a frequency of violence suffered of 18.9%.One study with 229 adolescents in outpatient services of a tertiary public pediatric hospital ascertained that only five adolescents reported not suffering psychological violence, showing the extent to which this issue is present in adolescents' lives and that, often, it is confounded with normality. 17n adolescent is considered to be an individual in transition, due to the biological development and transformations which are inherent to the age. 18n this stage of life, the individual is more vulnerable to the processes of complex social interactions, such as, for example, violence, in which he or she seeks reaffirmation of his or her values and beliefs.The social spaces in which the student is inserted, therefore, must be receptive to understanding her context and must offer support for promoting humane integration and strengthening of the interrelationships.The school, therefore, can be a healthy social space strategy for adolescents.
Being of male sex appears in research as a risk indicator for the phenomenon of violence. 6The association with violent behavior is related to prevalent factors such as alcohol use, drug use, being of the male sex, and having a relationship with parents which is not satisfactory. 6,19Living with parents, as protection for not suffering violence, may have lost statistical significance in the multivariate analysis, due to the family being understood as a system in which violence occurs, and not as a protective factor for the adolescents. 20However, when the family interacts satisfactorily, it contributes to the consolidation of atmospheres which are healthy for human and social development.As a result, it is fundamental for investment to be made in promoting family health, as a complex unit in the network of relationships and social associations.
In this study, it was evidenced that the economic situation of the student is associated with the complex phenomenon of violence.Unfavorable living conditions, and few opportunities, due to low purchasing power, may be factors for a greater tendency to become involved in situations of violence. 20his evidence is contrary to that of a separate study undertaken, 6 which indicated that adolescents from families with better socioeconomic situations present a greater probability of behaviors of violence.Within this perspective, it is revealed to be necessary to highlight that the interaction of the social determinants of health influences situations of violence of adolescents; including social, economic, cultural, and behavioral factors, among others. 21eligion, related to a set of rituals and beliefs, appears as an indicator of protection for not suffering psychological violence, coming to contradict that found in a previously undertaken study, which did not show religion as associated with the variables of violence. 22It is believed that this factor of protection against psychological violence emerges through the influence of religious thinking in the individuals' behaviors, taking into account that religions are associated with spaces for meeting -and that, consequently, they constitute social spaces.One study undertaken 23 showed the importance of religious groups for reducing violence among adolescents, bearing in mind that these are characterized as spaces for affective exchange and which have a disciplinary effect on human behavior.
Dysfunctional families, who experience a high frequency of disturbed family relationships, are associated with different forms of violence. 17he increase in the number of siblings indicates a greater tendency for the phenomenon of violence.Therefore, in this process, emphasis is placed on Brazil's National Health Promotion Policy, which aims to cover each part in its whole, as well as the whole -the family -in the interrelationship with each part -the individual.It is important that in this interdependence, there should be effective and affective dialogical exchanges for the encouragement of healthy behaviors.Through this Brazilian health policy, it is aimed to reduce vulnerabilities and improve quality of life, strengthening family relationships in a positive way. 24t was observed in this study, as also in other studies, that relationships in the family environment which are not satisfactory and are divisive generate violent behaviors. 6,17,25Intrafamily violence was mentioned by 9.5% of one sample of a conglomerate of 60,973 students from basic education in state and private schools of the capitals of the Brazilian states and of the Federal District. 5issatisfactory family relationships can emerge from a family dynamic which requires (re-) organization.Visualizing the individuals and their family system is revealed to be a relevant strategy for better coexistence between the family members, causing the family interactions to have the impor-tance of the whole which they are made up from.Public policies and actions in the schools, encouraging cooperative work between students and their family members, constitute strategies which are increasingly urgently needed for strengthening the relationships and healthy social interactions.
The area of health has a privileged space in the family group and, due to this, must act in preventing violence and in practices of health promotion. 26The nurse must be aware of her personal and professional commitment to the human being, in order to understand the student, the families and the living contexts, thus being able to perceive the interactions established between children and adolescents, acting as a mediator in a process of organization of a new family order, through bond and dialogue.
7] There was no difference in the frequency of the 59.8% who suffered physical/sexual violence and of the 40.2% who suffered psychological violence.It is therefore necessary for the nurse, as a professional who articulates knowledges, through a broad view, to be able to perceive the students in their context and not to separate them from it.Only in the interaction with the family, school and community would it be possible to perceive the complexity of the issue of violence and which strategies may be outlined through dialogue with a single context in its multiple interactions.
Through complex thought, the Dialogic Principle goes far beyond synthesis, integrating and understanding ideas considered to be antagonistic in sets. 28A dialogicity between student, family, school and context is configured as a strategy such that one may eliminate the existence of a single truth through the understanding of the local context and of the autonomy of each component of the system.
Violence is, therefore, a complex phenomenon which cannot be disassociated from its family and community context.The factors associated with violence show the need for an interdepartmental, transdisciplinary and multidimensional approach, articulating service networks for the elaboration of public policies for effective confrontation of the same. 8his being the case, interdepartmental approaches constitute an important strategy for promoting health and reducing school violence, above all in communities with indicators of vulnerability, so as to transcend the linearity of piecemeal and verticalized actions. 14nterdepartmental policies show the necessity for professional training and preparation for identifying, notifying and monitoring students in situations of violence, the support of the authorities being necessary if there is to be resolution of issues related to this question which is both social and related to public health. 17ith healthcare as the objective, professional nurses can integrate knowledges through transdisciplinary knowledge of the contexts proposed, 29 with an increasingly recognized social aspect to their work, capable of being articulators in the different social fields. 30ransdisciplinary knowledge is associated, in its turn, with the dynamic of the multiplicity of the dimensions of reality and of human beings, in which the fragmentation of knowledge does not hinder the social needs.It is necessary for there to be interconnection of knowledges through the articulation of antagonistic concepts, leading to a logical contradiction of health with situations of vulnerability, interlinked by the networks of associated factors such as family income, sex, number of siblings and specific cultural processes.Transdisciplinarity, as a possibility for the construction of knowledge in integrated and continuous movement, requires the comprehension of the human condition, through a crossing of disciplines, constructing a relevant knowledge together. 14n order for the multidimensionality of the students to be perceived, broadened and contextualized approaches by the nurse are required, such as, for example, the perspective of complexity.The individual needs to be perceived in her multiple dimensions, this being, at the same time, biological, psychic, social and spiritual. 14In the academic training, the nurse must be inserted and prepared to perceive the multiple dimensions which surround questions of students' health, not reducing the social conditions to the individual, and the individual to the social conditions.Through the understanding and the bond with the community, the nurse will be able to act in the issue of violence and in its repercussions in communities with indicators of vulnerability.Future studies must address specific strategies for the work of the nurse in the different contexts, with regard to mediating processes of reorganization related to the intervenient aspects of violence in the life of students, families and communities.

FINAL CONSIDERATIONS
The results evidenced that the factors associated with exposure to violence in students of the public educational network of a community with indicators of vulnerability in the South of Brazil are sex, number of siblings and family income, and that the type of violence suffered is related, in particular, to religion.
Violence is, per se, a complex phenomenon.Working on the multiple factors which surround this complexity, therefore, will enable as a broad understanding of this social and health issue.The interactions between the student, families and communities must be strengthened, so as to perceive violence in the economic, social, environmental and cultural contexts.Studies which address specific strategies for the work of the nurse must be undertaken for the prevention of risks associated with violence in vulnerable communities.
In summary, it is concluded that interdepartmental, interdisciplinary and multidimensional approaches constitute important strategies for the promotion of health and the reduction of school violence, above all in communities with indicators of vulnerability.As alternatives for change, the effective implementation is suggested of the Health Promotion Policy in communities with indicators of vulnerability; the active insertion of the professional nurse into the school space, and the institutionalization of healthy practices, both in the school space and in the family and community.
The study's limitations are those relating to the method chosen, as the reverse causality bias cannot be eliminated in transversal research.

Table 2 -Indicators of demographic and school-related prevalence associated with the outcome of suffering violence, Santa Maria-RS, 2014. (n=82)
Prevalence ratio adjusted for sex, number of siblings and monthly family income; † Variables which did not enter the adjusted analysis; ‡ Variables which were not retained in the final model. *