Bullying in children and adolescents: A modifiable risk factor for mental illness

Introduction

It is a serious concern to health practitioners and policymakers that, in spite of substantial investment, there has been no meaningful decline in the prevalence of mental illness in Australia (Slade et al., 2009). It is now understood that a complex array of biopsychosocial factors confer varying degrees of risk of mental illness. Genetic predisposition, obstetric complications, environmental toxins, poverty, developmental delay, substance abuse, exposure to loss and trauma, chaotic family environments with accompanying abuse and neglect, chronic physical illness and maladaptive interpersonal interactions all contribute to an increased risk of developing mental disorders (Kieling et al., 2011). Bullying in childhood and adolescence is an identified risk factor for mental disorders, suicide attempts and drug and alcohol problems (Copeland et al., 2013; Moore et al., 2013).

Australia has been a world leader in several public health initiatives that have resulted in substantial changes in behaviour at a population level. The compulsory wearing of seat belts in cars, banning smoking from restaurants and public areas and educating parents about the sleeping position of babies are examples where epidemiology has informed policy, legislation and public awareness, resulting in reductions in associated mortality and morbidity. The mental health field now faces the challenge of translating findings from epidemiological research into effective, population-based programmes that result in a discernible reduction in mental illness.

Clinicians, researchers and educators are raising awareness of the adverse impact of bullying on mental health and the need to develop and implement effective interventions. These interventions reduce the prevalence of bullying in childhood and adolescence, potentially resulting in a reduction of mental illness that is attributable to bullying. This year (2013), the Australian and New Zealand Mental Health Association ran a national bullying conference that brought together individuals with a common goal of reducing bullying in the Australian community.

We suggest that bullying in schools is arguably the most important aetiological factor for mental illness that could be systemically targeted at a population level. We base this on five criteria :

1. Bullying is easily defined, enabling it to be reliably measured.

2. A substantial proportion of Australian children and adolescents are exposed to bullying in the school environment.

3. Unlike child maltreatment or abuse, bullying is a risk factor that is acceptable to measure in schools.

4. In addition, bullying is a modifiable risk factor for mental disorders across the life course.

5. Finally and most importantly, with concerted effort, the number of children and adolescents in the Australian community exposed to bullying could be substantially reduced.

The most widely used definition of bullying is that proposed by Olweus (1993), which states that a person is bullied when they are exposed, repeatedly and over time, to negative actions from one or more people with an imbalance of power being present. These negative actions occur in various forms, incorporating direct physical contact as well as indirect methods, such as verbal harassment, spreading rumours, intentional exclusion from a group, obscene gestures and cyberbullying (Olweus, 1993; Olweus, 1994; Smith et al., 2008).

In recent years, the increase in internet usage and mobile phones availability has seen a surge in cyberbullying (Williams and Guerra, 2007). Unlike traditional bullying, cyberbullying is not restricted to a specific environment, which means that the child or adolescent has limited escape from the victimisation, since cyberbullying has the ability to take place at any time without the perpetrator coming in direct contact with the victim (Smith et al., 2008). The escalation of cyberbullying over the last decade mirrors the increased use of technology among students in both primary and secondary education.

It is clear that at least one in 10 Australian school children or adolescents are the victims of recent bullying, both within Australia and globally. In Australian studies, the prevalence of bullying victimisation ranges between 10 and 35% (Bond et al., 2001; Cross et al., 2009; Forero et al., 1999; Hemphill et al., 2011; Rigby and Slee, 1991). These discrepancies in prevalence can be linked to the inconsistency in how bullying is measured (Griffin and Gross, 2004). Some studies capture lifetime prevalence while others examine bullying in the last 12 months or the last 30 days. Furthermore, without asking the child or adolescent about the frequency of bullying, those children or adolescents who have been bullied occasionally and those who are bullied several times a week are grouped together in some studies, thereby inflating prevalence. Many bullying measurements also do not capture cyberbullying, consequently leading to a possible underestimation of prevalence. However, regardless of these inconsistencies in the measurements, it is clear that at least one in 10 Australian schoolchildren or adolescents are the victims of bullying.

In contrast to other risk factors for mental health, bullying is acceptable to report. For example, it is enormously problematic to measure abuse and neglect in children or adolescents at the population level. Among other things, students will not provide the information even if asked because of legal ramifications to the family; additionally, there would be likely opposition by some parents to their children being asked such sensitive questions. Furthermore, a child or adolescent may not report exposure to abuse or neglect because often they live with the unresolvable paradox where the perpetrator of their harm is also their caregiver (Fallon et al., 2010). Bullying is more likely to be accurately reported by those children and adolescents who are victims of bullying and its measurement is more acceptable to parents.

There is now a wealth of studies strongly supporting bullying as a putatively causative risk factor for mental illness in children, adolescents and also later in adulthood. Longitudinal studies consistently show a strong, temporal relationship with bullying victimisation preceding later mental health problems including depression, anxiety, suicidal ideation and suicide attempts (Copeland et al., 2013; Moore et al., 2013). Furthermore, a dose–response exists with children and adolescents exposed to more frequent and persistent bullying victimisation at the highest risk of mental illness (Bond et al., 2001).

The Olweus Bullying Prevention Program is a widely used ‘whole-school-based’ intervention and has reported reductions of up to 70% in student’s reports of being bullied and bullying (Olweus, 1994). Following this and other early studies of bullying interventions, evidence has shown that bullying victimisation is modifiable at the population level. The initial implementation of the KiVa bullying intervention program in 888 schools in Finland resulted in a significant reduction of approximately 20% in both bullying perpetration and victimisation. It was estimated that in Finland, this would translate to a reduction of approximately 7500 in perpetrators of bullying and 12,500 students being victimised in the Finnish population of 500,000 school students (Kärnä et al., 2011). Interventions that use a whole-school approach involving multiple disciplines and the entire school community are more effective in addressing bullying compared to curriculum interventions and behavioural and social skill training interventions (Vreeman and Carroll, 2007).

Opportunities

We suggest that the most effective way to reduce bullying victimisation is through the following steps. First, we recommend that schools should measure the prevalence of bullying each semester using standardised questions (Appendix 1, available online) and share the results of this survey with the school community. This would ensure that the principals and administrators, teachers, parents and students at the school are aware of the magnitude of the problem and would enable the monitoring of trends of bullying within the school. Furthermore, it would potentially galvanise ‘bystander’ students to intervene and help to enable self-reflection in some students who may otherwise thoughtlessly perpetrate bullying without consideration for the potential harm of this behaviour. It would also increase the likelihood of families discussing school bullying as an issue at home, so that children or adolescents would be more likely to seek support if they started experiencing bullying. Importantly, it would provide the school community with objective data on the prevalence of bullying that could not be ignored. The suggested questions are brief so as to maximise compliance, however, consultation with stakeholders and piloting of questionnaires that capture other information of potential interest, such as the type of bullying behaviour and the subjective distress experienced, would determine if more information would assist in reducing the prevalence of bullying.

We further recommend that in time, this same bullying measure be used to estimate prevalence in every Australian school through the infrastructure currently available for the National Assessment Program – Literacy and Numeracy (NAPLAN). Like the NAPLAN results, bullying prevalence would be reported on the My School website (Australian Curriculum, Assessment and Reporting Authority, 2013). This would allow schools of similar socio-demographic status to be benchmarked against each other.

Challenges

It is unrealistic to expect that school bullying will ever be entirely eradicated. There is a risk that children or adolescents who bully others will be ostracised and perhaps even excluded from schools, therefore those who bully others will also need clinical and educational support (Kärnä et al., 2011). The effectiveness of any clinical intervention will be dependent on the cooperation of the child’s or adolescent’s family, which is an ongoing challenge in child and adolescent psychiatry (Sanders and Kirby, 2012).

It is unclear if a child’s or adolescent’s report of bullying perpetration or victimisation should be shared with parents and teachers as it happens currently with the NAPLAN results. There are two main reasons for not sharing the findings with teachers and parents; however, both have counterarguments. First, an absence of anonymity might be a barrier to some students reporting bullying behaviours, particularly perpetration. Opposing this argument is the duty of care to the participants and the obligation to disclose this information to a teacher or parent if the child or adolescent reports that they are at risk of harming themselves or others. Second, the publication of bullying prevalence rates may disadvantage those schools in lower socio-economic areas, since it has been shown that children and adolescents from lower income families are at increased risk of both bullying others and being victimised (Jansen et al., 2011). On the other hand, community awareness of higher rates of bullying may enable schools from low socio-demographic areas to obtain more support. Funding would also be required to ensure that schools could reliably measure bullying, however, it is likely that the costs would be offset by the benefits to the child’s or adolescent’s health and educational experience.

Finally, it might be argued that with the advent of cyberbullying, schools may no longer be the primary platform from which to intervene. However, cyberbullying is often linked to peer relationships within the school environment, and school-based interventions have been effective in reducing both traditional bullying and cyberbullying (Pearce et al., 2011). School-based interventions would be supported by public awareness campaigns, such as those delivered by the Australian Government (Bullying No Way! Safe and Supportive School Communities Working Group, 2013) and by organisations like headspace (National Youth Mental Health Foundation, 2013).

Conclusion

We described an opportunity to translate the findings from mental health epidemiology to the real world. Bullying is a modifiable risk factor for mental illness. We hypothesised that a substantial reduction in the prevalence of bullying would translate quickly and persistently into a non-trivial improvement in the mental health and well-being of young Australians. The regular measurement of bullying and the benchmarking of schools on bullying prevalence would increase the probability that schools would implement effective strategies to provide a safe environment for students.