The Role of Disability in the Relationship Between Mental Health and Bullying: A Focused, Systematic Review of Longitudinal Studies

Being bullied is harmful with negative consequences for the victim. It is one of the main reasons why adolescents feel unsafe in school [1] and a significant environmental trauma associated with mental health problems (MHP) [2]. Bullying constitutes a severe form of harassment that is systematic and repetitive, where the victim loses power over the aggressor [3, 4]. Children with disabilities, both with and without a diagnosis, are at a higher risk of being bullied [5‐7].

A transactional model can be used to theoretically explain the relationship between disability, bullying, and MHP. According to a transactional model [8], children’s development is shaped in a reciprocal interaction with the immediate environment. Further, children’s previous experiences affect their interpretations of new interactions, and certain traits will be strengthened or weakened in the transactional process [8]. While a disability can be thought of as an impairment related to the individual, a fuller account of disability can be viewed as the impact of a disability that is created within the transactions between the individual and their environment. Similarly, Bronfenbrenner’s bioecological model also draws attention to the importance of interaction between the person, the context, and the processes between them, i.e., the interaction will, over time, drive development [9]. Development can be seen as the product of a reciprocal interaction between a person and their environment, i.e., in the proximal process [10]. In such a model, bullying is a powerful contextual element that repeats over time, forming part of a proximal process. Bullying victimisation is likely to affect development substantially, given that repetition is a part of the definition of bullying. Moreover, the loss of power experienced by the victim is expected to impact or restrict a child’s interactions with the world, and affect how the child then interprets new social situations. Especially where bullying is chronic, maintained over several years or even school transitions [11], this will arguably affect school and social functioning and have longitudinal effects on health, especially psychological factors, such as mood, stress, and concentration. Previous research has shown that bullying victimisation increases the risk for MHP [12‐15], social problems [16] and physical health issues [17]. There is also a risk of these issues continuing into adulthood [15].

The WHO definition of health is complete physical, mental, and social wellbeing and not merely the absence of disease or infirmity [18]. This definition suggests that factors that affect wellbeing in the short or long term, such as bullying, are problematic and detrimental to health [2]. It also indicates that disability does not reduce wellbeing but can be a risk factor for developing adverse health due to environmental barriers or responses [19]. In the International Classification of Functioning, health, and disability (ICF) [20], a shift toward a biopsychosocial model of disability and health has emphasised the role of environmental factors in creating disability. In other words, how impairments impacts, through environmental factors, to shape disability. However, the ICF is a complement to the global standard for diagnostic health information International Classification of Diseases (ICD), which focuses on classifying functioning and disability via diagnoses. Hence, the ICF defines disability differently compared to the ICD.

The benefit of using the ICF’s definition of disability is its breadth of scope. For example, it includes children who have impairments relating to mental functioning, such as difficulties with attention or the regulation of emotions. It also includes children who have restricted physical activity that can affect their capacity to handle general tasks and demands. Within a transactional process, both these children will have visible symptoms within a context, independent of a medical diagnosis or not. Hence, this definition of disability can include those with a diagnosed disorder as well as those children who display symptoms but do not have a confirmed diagnosis.

When investigating MHP outcomes in studies of children with disabilities/symptoms of MHP and bullying, the measurement of early symptoms of MHP needs to occur before the measurement of bullying. Similarly, the measurement of bullying needs to occur before measuring later MPH outcomes. The measurement thus needs to be in the order of: (1). disability, either as a disorder or as early symptoms of MHP; (2). Bullying; and (3). the relation to later MHP in longitudinal studies. As bullying is more prevalent in early adolescence [26, 27], studies with the first point of measurement in adolescence might miss essential aspects of the relation between early, i.e., childhood, symptoms and later mental health outcomes concerning bullying, for example, in children exposed to chronic bullying. Therefore, the current review only includes studies with initial measurements before 10 years of age to isolate the effect of bullying on functioning or on later MHP.

This study is a limited systematic review. The literature searches for the review were conducted in multiple search engines in August 2019 and again in August 2020 to include any recently published studies. Databases used were CINAHL, ERIC, MEDLINE, Psych Info, Sociological abstracts, PubMed, Swepub, open grey, OECD library, and Scopus. Primarily thesaurus terms were used, though these differed somewhat depending on databases. All terms were related to MHP for children with disabilities, bullying or peer victimisation, and comparisons with other children [for specific search strings used, see supplement (1)]. The thesaurus terms used reflected our definitions of participants, exposure, comparison-group, and outcome (PECO). Table 1 shows the PECO criteria.

This review investigated the longitudinal associations between disability/early symptoms of MHP, bullying victimization, and later additional MHP. Most of the 12 included studies used a measure of symptoms of MHP at baseline (k = 8). These studies had a symptom-driven perspective, which is more in line with the ICF definition of disability that we follow, whereby psychological symptoms in early life were related, via subsequent bullying, to later mental health problems. All but one of these studies were rated as having medium or high quality and had long follow-up periods. Three of the studies used register data as outcomes. Five of the twelve included studies used a diagnosis of disability or disorder as its baseline measure.

In general, the combined results support our hypothesis that children with disabilities will have a higher risk of developing MHP as adolescents if they have been bullied compared to children who have not been bullied or only have a disability. However, two aspects need to be considered: the results are equivocal when only considering children with diagnosed disorders; only 3 out of 5 studies supported our hypothesis. In addition, regarding diagnosed disorders, only studies focusing on neurodevelopmental disorders were found. Second, the only relationships identified concerned externalising symptoms in childhood and internalising symptoms as outcomes, especially depressive symptoms. Regarding studies including children with diagnosed disorders, no study focused on depression specifically as an outcome. The outcomes used were other than for those used in studies that measured symptoms of MHP in childhood.

The studies that measured symptoms early in life primarily focused on outward, behavioural aspects, i.e., emotional dysregulation, norm-breaking behaviour, aggression, and behavioural problems. Difficulties with self-regulation are common in many neurodevelopmental disorders as well; the two population groups (children with symptoms or disorders) might overlap in this way. Children with ADHD are present in several studies [37, 40, 42], and one study focused on DCD [39]. Of these, all but one [40] found a relation between disorder and MHP as outcome mediated by bullying. The study that did not support this relation only had a 2 year follow-up and focused on 4- to 6-year-olds. This aberrant finding might be explained by the fact that MHP are more prevalent in adolescents [26].

Often children with symptoms of MHP or neurodevelopmental disorders have difficulties in the social context and have less access to protective factors, such as good social abilities. Children with higher levels of social skills had lower anxiety at follow-up [42]. Additionally, two other studies found that behaving aggressively in different forms toward peers might lead to peer rejection [31, 32]. Even though peer rejection might not always increase the likelihood of being bullied, it indicates that children with early difficulties may find themselves excluded from social situations. Such situations are essential for positive development as they provide environments for children to learn and improve their social abilities. Therefore, as these issues are visible early in life, it is vital to support these children’s management of their social context. While disorders are not always preventable, adaptations and support from the environment could lead to less problematic behaviour, which could reduce the risk of other negative experiences and thus also reduce the risk of later mental health problems.

Children’s behaviour will impact their surroundings through a reciprocal interaction [8]. Children who have a higher risk of developing depression tend to display early externalising behaviours, such as defiance [31], aggression [32], or behavioural problems [37]. The display of these externalising behaviours increases the likelihood of being disliked, rejected [32], or chosen as victims of bullying, perhaps for not having as many friends to protect them [43, 44]. Being rejected might increase negative behaviours, and over time, these children who externalise problems may also develop internalising issues, such as depression. The two studies on a Finnish sample [33, 38] included more than just internalising issues, such as depression, in their psychiatric disorder outcome. However, they supported the notion that early behavioural problems will increase the likelihood of less than optimal social and mental development. Children who poorly manage their contextual demands will risk peer victimisation and later psychiatric disorders.

Based on a biopsychosocial model [20], MHP in childhood produce a more conclusive finding than focusing on the diagnosed disorder as a measure of disability. This finding can be supported by the view that the role of environmental, contextual factors interacting with outward displays of symptoms serves to create disability [20]. The results indicate that having externalising behaviours in childhood is a risk for unnecessary, additional MHP in adolescence due to being bullied. The studies that measured symptoms early in life primarily focused on behavioural aspects, i.e., emotional dysregulation, norm-breaking behaviour, aggression, and behavioural problems. Many children with neurodevelopmental disorders also display these behavioural aspects, but they do not necessarily need to. The diagnosis does not give information regarding the contextual adaptations made due to the disorder nor the in-group variance of symptoms. Therefore, a categorical definition of diagnosed disorders does not give us information about barriers and facilitators the way symptoms might. This lack of contextual information could explain the inconclusive results for those with diagnosed disorders. However, symptoms of MHP, arguably, indicate a mismatch between a person’s ability and contextual demands. Adding risk factors for MHP in adolescence, in this case, bullying, the risk for additional internalising symptoms increases [44]. Having a neurodevelopmental disorder increases the likelihood of environmental barriers, raising awareness of the need for adaptations. It could be argued that behavioural problems, rather than a diagnosis, are of more relevance in the developmental relation between bullying and MHP. What weakens this argument is that we do not have any studies that look at children who have similar diagnosed disorders of other sensory or physical impairments.

Measurement of internalising outcomes was not the same between those with childhood symptoms of MHP and those with a diagnosed disorder. No identified study focused on depressive symptoms alone for children with neurodevelopmental disorders. At the same time, this was the most common outcome in studies of children with childhood symptoms of MHP. Previous research has shown a strong link between internalising symptoms and bullying, and the most common measure of this is depressive symptoms [21].

From a systems theory perspective, bullying is a significant environmental trauma [2] and is a barrier to participation in meaningful life situations. Many children with disabilities have reduced participation in everyday activities, either in the number of activities [45, 46] or level of intensity [46]. The degree of participation outside of school seems to vary in relation to different aspects of the environment rather than the type and severity of impairment [47]. Many children with disabilities have sedentary activities in the home, such as watching TV [48]. Children who are bullied watch much more TV than children who are not bullied, and children with disorders who are bullied even more so. Also, girls with ADHD have a high risk of peer rejection, predicting poor social adjustment [49]. Sedentary home activities can be one way of handling social isolation, especially concerning social skills or self-regulation difficulties. Over time, this could explain low participation in peer activities, less social adjustment, and increased risk of continued internalising problems [19].

Many school-aged children experience bullying, and those with disabilities are more likely to be victimized [5‐7, 32, 52]. Being bullied increases the likelihood of later internalising symptoms, such as depression. Whilst there are many studies on bullying and MHP, these are primarily cross-sectional studies [21]. The current review only identified 12 studies with a longitudinal design that examined the developmental paths for children with a diagnosis/early symptoms of MHP exposed to bullying and their later MHP. These 12 studies primarily focused on childhood psychological symptoms or neurodevelopmental disorders. We found no longitudinal studies that included children with learning disabilities, intellectual disabilities, or physical disabilities, even though these groups also have an increased risk of being bullied [7]. Cross-sectional or retrospective studies usually find higher correlations between MHP and bullying for typically developing children and children with disabilities [47, 48, 54]. Arsenault [6] argued that children with externalising behaviour are more likely to be bullied and have poorer mental health outcomes. The current review’s results support the finding that early externalising problems were more closely related to later internalising issues via exposure to bullying. However, the sub-set of studies that used a confirmed diagnosis at baseline did not support a clear relationship between bullying and later MHP. This inconclusiveness might indicate that the externalising behaviours have a more significant impact on the transactional process than a diagnosis does. A diagnosis does not automatically indicate an issue with social interaction or lack of appropriate adaptations,externalising behaviours, visible in the social context, indicate a mismatch between expectations and adaptations. Alternatively, measuring symptoms rather than diagnoses results in broader inclusion criteria in studies and reduces bias due to design, which captures more significant amounts of variance, i.e., changes in MHP following bullying.