Adolescent violence is a global public health and social problem that affects physical and psychological health and can have negative social consequences for society [1, 2]. The perpetration of violence not only entails welfare costs and frays the social fabric [3] but may also increase health care costs, decrease property values and disrupt social services [4]. While studies have documented the prevalence of and risk factors for violence perpetration, we lack research focusing on variations in smaller geographical units and on the combined impact of socioecological factors on violence perpetration among adolescents. Knowledge of local variations in adolescent violence and associated risk factors may provide essential input for policymakers in designing area-specific policies and intervention strategies to reduce the prevalence of violence.
Although Norway is a low-violent society, violent crime among adolescents is increasing [5]. Norwegian municipalities also have a high degree of autonomy in implementing adolescent policy and may develop their own planning and intervention strategies to combat violent crime [6]. The current study examines municipal-level variations in prevalence of violence perpetration among Norwegian adolescents. Drawn from the socio-ecological model [7, 8], it investigates how individual level health-risk behaviours, family factors and municipal-level characteristics explain local variations in violent behaviour.
The Socio-Ecological Model
The prevention of violence among adolescents requires understanding its risk factors [9]. The socio-ecological model is commonly used to explore factors promoting or impeding violent behaviours among adolescents [10]. Bronfenbrenner’s ecological systems theory (1979) is the most influential contributor to socio-ecological thinking and has been extensively applied in many types of health research [11]. The model asserts that no single factor promotes or impedes violent behaviour on its own. Consequently, we may benefit from examining the overall effects and interactions of individual, relationship, community and societal factors that affect violent behaviour. By analytically dividing the social environment into different categories, ecological thinking provides a systemic focus on the different levels and types of social influence, such as the categorization of environmental influences into micro-, meso-, exo- and macrosystem levels [12].
The microsystem comprises interactions between individuals and their immediate environment [7]. The most direct influences on violent behaviour are those that exist within a microsystem, such as adolescents’ interactions with parents and peers [13]. The mesosystem consists of interactions between the adolescent’s microsystems, such as parents with school administrators or parents with religious institutions. The exosystem may have an indirect influence on adolescent behaviour that could result in the perpetration of violence [13, 14]; examples include poverty and unsafe neighbourhood environments. The macrosystem involves cultural patterns within the micro-, meso- and exosystems, such as political culture, religion, social values and norms [13, 15].
The socio-ecological model has been refined in a public health context by the World Health Organization (WHO), to better understand the causes of interpersonal violence [16]. The WHO socio-ecological model comprehends the multifaceted nature of violence as the interaction among individual, relationship, community and societal factors [8] .
Individual Factors
Studies have shown that violence perpetration is associated to a wide range of individual level health-risk behaviours, among which are depression, smoking, excessive use of alcohol and use of illicit drugs—all which are risk factors for non-communicable diseases [1]. Yu et al. [17] examined the longitudinal associations between youth depression and later violent outcomes in the Netherlands, the United Kingdom and Finland. The results indicated a pattern of subsequent increased risk of violence perpetration among respondents with high levels of depressive symptoms. Another study by Fazel et al. [18] reported a high prevalence of depressive symptoms among adolescents in juvenile detention and correctional facilities. This corresponds with a study by El-Slamoni and Hussien [19], which found a positive correlation between depressive symptoms and physical aggression. Due to age-related outbursts of anger and impatience, adolescents with more severe depressive symptoms may be more likely to commit violent acts [17].
According to the WHO (2015) [1], alcohol and substance use may be risk factors for violence perpetration. Studies have shown that increased exposure to alcohol, cigarette and illicit drugs among adolescents increased the likelihood of violent behaviour [20]. The National Survey on Drug Use and Health [21] reported that the probability of violence perpetration increased with the number of drugs used among youth. Adolescents who used an illicit drug were twice as likely to be involved in violent behaviour than those who did not use. Goldstein’s psychopharmacological model postulates that intoxication effects, such as disinhibition, cognitive distortion and attention deficit may increase adolescent’s involvement in violence perpetration when under intoxication [22]. When it comes to alcohol use, Swahn et al. [23] reported that alcohol intoxication may impair adolescents’ thoughts and behaviours and thereby increase their risk of engaging in violent behaviours. Studies have also reported that adolescents who consumed alcoholic beverages early in life and did so frequently was at higher risk of violence perpetration [9]. A longitudinal study on developmental trajectories of alcohol use and violent behaviour showed associations between the level of alcohol consumption among adolescents and violence perpetration, where those who drank frequently of large quantities was at higher risk for showing violent behaviour [24]. Similar to alcohol use, the level of cigarette use may be associated with adolescent violent behaviour. A study in Canada by Morris et al. [25] found that adolescents smokers were more likely to be bullies than non-smokers [25]. Correspondingly, Matuszka et al. [26] showed that addictive cigarette use was associated with elevated physical aggression and the prevalence of aggression was higher among addicted tobacco users than in single users or non-smokers. Although previous studies rarely have discussed why cigarette use is associated to violence perpetration, a study by Lewis et al. [27] has indicated that the association may have a neurobiological explanation. The components of tobacco, such as nicotine, may disrupt the neural circuitry of smokers and thereby increase aggressive responses.
Family Factors
A range of factors associated to the functioning of the family may increase the risk for adolescent’s involvement in violence perpetration. Parental conflict, poor attachment between parents and children, adolescents of single-parent households and low level of family cohesion have been associated with violence perpetration [3]. A study by DiClemente et al. [28] stated that observing risky behaviours among parents’ may increase the risk of such behaviours among adolescents. As such, conflicts existing within the family have been found to influence adolescent engagement in risky behaviours. Parents’ involvement in and monitoring of children’s development may also play a major role in influencing adolescents’ involvement in violence perpetration. Children who grow up with less than ideal parental supervision tend to be involved in violence perpetration more often than children who grow up under proper supervision [1]. Additionally, Nilsson and Nordentoft [29] reported that adolescents from low income families had the highest risk of violent offending, while those from high income families committed significantly less violence.
Community and Societal Factors
Community and societal factors, such as school environment and demographic and socioeconomic characteristics, may increase the risk of adolescent violent behaviour. Examples are factors in young people's social environments, including the community’s influence on their families, the composition of their peer group and their exposure to potentially violent situations. Exposure to an unhealthy social environment and deviant behaviours in the neighbourhood, such as criminal acts, drug dealing, or abuse, gangs and problems related to housing may be predictors of adolescent violence [3, 30]. For example, adolescents living in urban areas may be more likely to show violent behaviour than those living in rural areas. Children who grow up in neighbourhoods characterized by a high level of criminal behaviours and poverty have been identified as being at the risk of adolescent violence [3].
The Current Study
The social-ecological model suggests that violence perpetration does not occur in isolation; therefore while studying its risk factors, it is important to consider the effect of individual factors within the context of environmental input [30]. The current study used a combination of Bronfenbrenner's socio-ecological framework and the WHO [8] ecological model developed for public health research to frame the data analysis, including explanatory factors at the individual, relationship, community and societal levels. The first level, individual, correlated violence perpetration and individual health risk behaviour. The second level, relationship, associated family factors and violence perpetration. The third level, community and societal, correlated municipal-level factors and violence perpetration (Fig. 1).
The study had two aims. First, it sought to examine variations in the prevalence of violence perpetrated by adolescents at a municipal level. The expectation was that the prevalence of violence perpetration among Norwegian adolescents would vary between municipalities and that the differences would be explained by individual-, family- and municipal-level factors. Second, it aimed to examine the combined impact of individual health-risk behaviours, family predictors and municipal characteristics on violence perpetration among Norwegian adolescents. Consistent with previous studies, it was expected a positive correlation between violence perpetration and socio-ecological correlates.