Abstract
Non-suicidal self-injury (NSSI) has attracted increasing attention due to its high detection rate, high risk and high repeatability. There is a need for the early identification of preventable occurrence factors, which is necessary to facilitate screening and intervention, especially to facilitate the early detection of high-risk individuals. This research aims to investigate the relationship between bullying behaviour and non-suicidal self-injury among children and adolescents by means of meta-analysis. The PubMed, Embase, Web of Science, SCOPUS, PsycINFO, CKNI and WAN FANG databases were searched from inception to 14 December 2021 for studies that explored the relationship between bullying behaviour and NSSI among children and adolescents. A total of 29 articles met the inclusion criteria of the meta-analysis, and 54 independent effect sizes were obtained, including 53,501 subjects. Victims [OR 2.46 (95% CI 2.14–2.83); p < 0.001], bullies [OR 2.12 (95% CI 1.37–3.27); p < 0.001], and bully-victims [OR: 2.98 (95% CI 1.85–4.82); p < 0.001] were more likely to have NSSI than uninvolved children or adolescents. In addition, analyses showed the absence of publication bias. In the victim group, the older the age was, the lower the risk of NSSI (z = − 3.74, p = 0.00). Gender does not play a moderating effect on the association between bullying behaviour and non-suicidal self-injury. The relationship between involvement in bullying and NSSI was demonstrated. By taking measures to prevent bullying, the incidence of NSSI in children and adolescents can be potentially reduced.
Similar content being viewed by others
Introduction
NSSI involves to intentional direct self-harm to one's own body without suicidal thoughts, may include scratches, bumps, cuts, burns and other forms, and jt is not socially and culturally recognized1. The most recent version of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) introduces non-suicidal self-injury as a disorder for further exploration2. In a meta-analysis, the incidence of non-suicidal self-injury in adolescents was as high as 17.2%3. Liu et al.4 surveyed 2716 Chinese adolescents and found that 26.9% of the subjects had NSSI in the past year suggesting that paying attention to NSSI should be one of the important tasks for maintaining adolescent mental health. Although most people who repeat non-suicidal self-injury stop this behaviour within a few years, they usually follow a more chronic process, with approximately 20% of them continuing for more than 5 years5. NSSI is often strongly associated with depression or bipolar disorder in adolescents6,7. The reason may be that brain changes in early-onset individuals (especially adolescents) with mood disorders have a significant impact on negative experiences8. Studies have shown that adolescents with bipolar depression have more abnormalities in their brains than those with unipolar depression8. In addition, many empirical studies have indicated that NSSI will not only increases the risk of mental illness, but also significantly increases the risk of suicidal behavior9,10. The incidence of suicidal behaviour is frequently underreported. For example, studies have confirmed that more than 2% of traffic accidents are considered suicide behaviours11. However, car accident suicides are often reported as accidental events in national statistics, so this phenomenon may be underestimated11. Therefore, given the negative outcomes of NSSI, timely detection of risk factors for non-suicidal self-injury may be helpful for the formulation of prevention programs.
Within this context, several studies have shown that bullying behaviour is an important predictor of NSSI among children and adolescents12,13,14 There are cultural differences in the definition of bullying, however, bullying generally refers to repeated physical or verbal harassment and involves the imbalance of power between the bully and the bullied15,16. Participants in bullying are generally divided into three types: bully (bullying others), victims (suffering from bullying), and bullying-victims (both bullying and being bullied by others)17. Although the effects of participating in bullying are about the same, the incidence of bullying may vary according to the role played by individuals in bullying. According to related research, approximately 15–50% of children participate in bullying behaviour as a bully, victim or bullying-victim18. Adolescents who are bullied may use self-harm as a form of calling for help16. Self-harm may be a form of relief from bullying-related stress and is a kind of self-punishment of the individuals19. In addition, due to a lack of self-regulation and impulsiveness, bullies may also develop self-injurious behaviours20. Bully-victims are also highly likely to have risky behaviours, as they have the risk factors for both bullies and be bullied21.
Meta-analysis is a scientific research method that can summarizes the results of several studies to evaluate the overall effects. In this study, the aim was investigate the association between victims, bullies, bully-victims and NSSI in adolescents by conducting 3 meta-analyses. As publications are more likely to publish studies with positive results and reject papers without positive results, this may lead to potential bias. Therefore, publication bias of the results were also verified through several methods22. In addition, the moderating effects of sampling strategy, gender and age were tested by subgroup analysis or meta-regression analysis to determine whether the relationship between variables was moderated by the above factors.
Methods
This study was conducted strictly according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) lists. See Supplementary file 1 for specific PRISMA Checklist.
Study retrieval and selection
The PubMed, Embase, SCOPUS, Web of Science, PsycINFO, CKNI and WANG FANG databases were searched. These databases have the characteristics of being widely used, extensive and authoritative. The search words were ("bully*" OR" bullie*" OR "peer victim*" OR "peer harassment" OR "school violence" OR "mobbing "OR" ragging") and “self-harm*" OR "self-injur*" OR "self-cut*" OR "self-destruct*" OR "parasuicid*" OR "automutilat*" OR "auto-destruct*" OR "non-suicidal" OR "self-mutilation" OR "NSSI") (14 December 2021). See Supplementary file 2 for specific retrieval strategies. References to retrieved studies were also scanned to further identify studies that met the inclusion criteria. This search strategy yielded 1391 studies. Two researchers independently reviewed all references to find as many studies as possible that met the inclusion criteria. After the duplicate studies were removed with Endnote, 781 studies remained. The specific literature screening process is shown in Fig. 1. Only peer-reviewed studies published in journals were considered for inclusion. Only studies that provide an effect size for the relationship between bullying and NSSI were considered for inclusion; Therefore, studies that met the inclusion criteria either compared the NSSI of adolescents who participated in bullying to the NSSI of adolescents who did not participate, or treated bullying behaviour and NSSI as continuous variables, providing the degree of correlation between the two variables. Only studies that provide an effect size of the current link between bullying behaviour and NSSI were considered for inclusion. If only prospective or retrospective data were presented, the study was not considered for the time being. In previous meta-analyses, retrospective studies and prospective studies are usually tested separately from cross-sectional studies because the first two pay attention to how bullying affects victims' later lives, which is different from how bullying is directly related to NSSI. However, two studies23,24 provided both prospective and current linking data, so we chose to include the Time 1 section containing the current links between bullying behaviour and NSSI in the meta-analysis. Studies with participants over the age of 21 were excluded. In addition, only literature that clearly focuses on bullying was considered. The forms of bullying mainly included physical bullying (e.g., hitting, kicking), verbal bullying (e.g., name-calling, insulting), and psychological or social bullying (e.g., lying, spreading rumors). Studies focusing on victimization or aggression against non-peers were excluded. Research was only included if it focuses on actual NSSI behavior rather than thoughts or ideation of self-injury. Papers were coded as NSSI only when the interviewees of the study were explicitly required to report self-injury with no intention to die.
Only studies with a sample of children and adolescents from the community were considered for inclusion in the meta-analysis; papers focusing on clinical samples were excluded as conclusions drawn from clinical samples are not suitable for generalization in the general population. To prevent each researcher from being repeatedly included in our study, the articles using more respondents for the studies using the same dataset were included. Studies written in English and Chinese were eligible for inclusion. Two authors independently evaluated the literature according to the inclusion and exclusion criteria. Finally, 29 studies met our inclusion criteria. Table 1 shows the basic characteristics of the included literature.
Coding
Twenty-nine studies with 54 independent effect sizes were included in the meta-analysis. If the article included multiple independent samples (e.g., different countries, men and women), these samples were included separately in this study. The odds ratios (ORs) were coded as the effect size measure in the study. Three studies that did not report the OR of all samples but provided enough data for us to calculate the OR29,32,46. Several studies13,14,23,24,25,27,30,31,37,38,39,42,43,44,45,48 provided correlations instead of ORs. For these studies, Comprehensive Meta-analysis (Biostat; http://www.metaanalysis.com/index.php) was applied to convert correlations into ORs. If an article included both adjusted and unadjusted effect sizes, then including an effect size adjusted for most confounders would be our choice. Several articles offered different ORs for different levels of bullying behaviour (e.g., occasional, repetitive), or divided bullying behaviour into different forms (e.g., physical, verbal, or relational). For these studies, children or adolescents who had experienced any form of bullying behaviour into a group were combined and compared them with children or adolescents who did not engage in bullying. All studies provide data on the NSSI risk of victims, ten studies12,13,14,32,34,39,40,42,47,48 provide data on bullies, and six studies provide data on bully-victims 12,32,34,39,47,48. Two researchers coded the included studies, including study location, sample size, age, sampling method, data collection method, and effect sizes. In case of disagreement, a decision was discussed with a third researcher. Before the discussion, the agreement rate was 92.1%.
Statistical analysis
Analyses were performed with the Comprehensive Meta-analysis. OR was used as the effect size of this meta-analysis. Other effect sizes, such as Pearson correlation coefficients, were converted to ORs prior to analysis. A Random effect model was used for data analysis. The random effect model is more suitable for the current meta-analysis than the fixed effect model because it does not assume the common potential effect size of all studies included in the meta-analysis. We performed meta-regression using the average age of the participants as a moderating factor. The average age of the subjects was not described in four studies27,30,41,45, but the age range of these four studies was relatively small, so we used the average of the age range instead of these. We compared the effects of studies involving only girls with those involving only boys to analyse the moderating role of gender on the association between bullying behaviour and NSSI. In addition, subgroup analysis was used to compare whether the effects of articles with random sampling designs were different from those with non-random sampling designs.
To test potential publication bias, Orwin's fail-safe N49, Duval and Tweedie’s trim-and-fill analysis50, and Egger’s regression intercept51 were calculated. The first estimates how many studies with insignificant results would be required to offset the effect size obtained in a meta-analysis. If many studies were needed, we might assume that although the effect size obtained through meta-analysis may be slightly overestimated, there is no obvious publication bias in its significant effect. The recommended rule is that the number of articles estimated by Orwin's fail-safe N procedure has to be more than 5K + 10,52, where K represents the number of effects included in the study. The Duvall and Tweedie trim and fill method estimates the effect size until the error distribution is very close to normal to provide a more unbiased estimate of the effect size than observed estimates. Egger's regression intercept used the linear regression method to evaluate the study effect size relative to its standard error to measure potential publication bias; Q-test and I2-test were used to detect the heterogeneity of the results53.
Quality assessment
The quality assessment of the included studies was conducted using the JBI Checklist54. The options for each evaluation item were yes, no and unclear. A yes answer counted for one point, and the rest counted for zero. The higher the score, the higher the quality of the included literature. Two researchers evaluated the quality of the included literature and communicated with the third researcher if there was any disagreement. The evaluation results showed that the 29 articles included in this study were of medium to high quality (total score ≥ 6). For specific evaluation, see Supplementary file 3.
Results
Association between being bullied and NSSI
Twenty-nine articles that met the inclusion criteria included 35 effect sizes with a total of 53,501 youths. The results showed that victims were significantly associated with NSSI [OR: 2.46 (95% CI 2.14–2.83); p < 0.001]. Table 2 shows the details. Figure 2 shows the forest plot. Egger regression analysis showed that t = 0.44, p = 0.66. The Duval and Tweedle trim-and-fill procedure indicated that two studies needed to be imputed, but the adjusted effect size (OR 2.34; 95% CI 2.03–2.70) was consistent with the size observed in existing studies (Fig. 3). Orwin Fail safe N shows that an additional 290 studies of invalid effect sizes are needed to reduce this combined effect size to an insignificant value (5k + 10 benchmark = 185). The Egger regression, the Duvall and Tweedle trim-and-fill procedure, and the Orwin fail-safe N indicated that the impact of publication bias was negligible for the meta-analyses. The methodological characteristics of the study were analysed through the moderating effect analysis. Studies using non-random sampling designs [OR = 2.73 (95% CI 2.28–3.26)] were not significantly different from those using random sampling designs [OR = 2.09 (95% CI 1.84–2.37)], [Q = 1.51, p = 0.22]. Studies using only boy samples [OR = 2.34 (95% CI 1.71–3.21)] were not significantly different from studies using only girl samples in effect sizes [OR = 1.71 (95% CI 1.33–2.20)], (Q = 2.31, p = 0.13). In addition, a meta-regression analysis of age shows that the older the age, the lower the risk of NSSI for the bullied (z = − 3.74, p = 0.00).
Association between active bullying and NSSI
Eight studies provided data on bullying youths. Meta-analyses indicate that bullying youths are more likely to develop NSSI than youth who are not involved in bullying [OR: 2.12 (95% CI 1.37–3.27); p < 0.001]. See Table 2 for details. Figure 4 shows the forest plot. Egger regression analysis shows that t = 0.25, p = 0.80. The Duval and Tweedle trim-and-fill procedure suggested that no additional research is needed (Fig. 5). Orwin Failsafe N showed that an additional 89 studies of invalid effect sizes are needed to reduce this combined effect size to an insignificant value (“5k + 10” benchmark = 70). The Egger regression, the Duvall and Tweedle trim-and-fill procedure, and the Orwin fail-safe N show no publication bias. Studies using non-random sampling designs [OR = 2.91 (95% CI 1.39–6.12)] were not significantly different from those using random sampling designs [OR = 1.79 (95% CI 1.12–2.86)], [Q = 1.19, p = 0.28]. The effect sizes of samples with only boys [OR = 2.11 (95% CI 1.00–4.45)] did not differ from the effect sizes of samples with only girls [OR = 1.11 (95% CI 0.62–2.02)], [Q = 1.72, p = 0.18]. The meta regression showed that the moderating effect of age is not statistically significant (z = 0.69, p = 0.49).
Association between bully-victims and NSSI
Finally, six studies examining the risk of bullying-victims developing NSSI compared with non-participating peers were analysed Meta-analysis shows that bullying-victims are more likely to engage in NSSI behaviour than non-involved peers [OR: 2.98 (95% CI 1.85–4.82); p < 0.001]. Table 1 shows the details. Figure 6 shows the forest plot. There was no evidence of publication bias. Egger regression analysis shows that t = 0.13, p = 0.90. The Duval and Tweedle trim-and-fill procedure indicated that 1 study needed to be imputed, but the adjusted effect size (OR 2.57; 95% CI 1.59–4.19) was consistent with the size observed in existing studies (Fig. 7). An additional 72 studies with null effect sizes would be needed to attenuate this omnibus effect size to a nonsignificant value (“5k + 10” benchmark = 45). Among the studies that included the bullying-victim group, only one study8 had independent male and female data, and only one study8 used a random sampling design, so meta-analysis was temporarily impossible. Meta-regression shows that age is not a significant moderating variable (z = 0.84, p = 0.40).
Discussion
Meta-analysis is an invaluable tool to integrate previous research, illuminate research gaps, and define priorities for future research. Twenty-nine articles were finally included in this study, and the quality evaluation results showed that they were of moderate to high quality. Meta-analysis results indicated that victims, bullies, and bully-victims were at higher risk of developing NSSI than their uninvolved peers. This is consistent with previous research results55. The reason may be that tensions can cause people to produce depression, anxiety, anger and other negative emotions. To release these negative emotions, dangerous behaviors are adopted such as self-injury and aggression to deal with them56. According to Agnew's general strain theory (GST), bullying incidents have physical and psychological effects on teenagers. It can cause lasting harm and the feeling of social injustice. In the face of such negative life events, deviant behaviours (such as violence and self-injury) become the most effective way to manage negative emotions57. Being bullied is an important source of stress58. Teenagers who have experienced this behaviour often have difficulty adapting to school or society, which may increase the risk of implementing dangerous behaviours (such as NSSI). This view has been confirmed by some studies30. The interpersonal model can also be used to explain the significant relationship between bullying and NSSI. This model considers NSSI as a negative coping strategy to reduce the pressure or tension caused by adverse interpersonal events59. For adolescents, bullying and being bullied are relatively common negative interpersonal events60. Bad interpersonal events may cause serious emotional distress. If adolescents cannot adjust in time, NSSI behaviours may occur, which will make the adolescent’s emotional and interpersonal communication tendencies desperate. Once NSSI promotes personal emotional comfort, it is reused this way to release emotions61.
The methodological characteristic that was studied as a moderator, namely, sampling strategy, was found not to significantly moderate the relationship between bullying and NSSI. This suggests that the effect size is not inflated by the inclusion of studies with relatively less rigorous methodologies. To analyse the moderating role of age in the relationship between bullying behavior and non-suicidal self-harm, a meta-regression was performed. The regression results found that in the bullied group, the older the age was, the lower the risk of NSSI. The results do not necessarily mean that younger children are more susceptible to bullying, however, peer victimization has been found to be more frequent in younger people than in older people62. In the current meta-analysis, it is possible that younger victims simply experienced more bullying incidents and thus experienced more NSSI than older victims. In addition, younger children may lack coping strategies to appropriately deal with peer victimization and are therefore more likely to develop problems as a result of peer victimization. Finally, no difference was found in effect size between boys and girls. One study found that being female was protective against bullying63, and another found that it was a risk factor for later bullying64. The literature for children indicates that bullies appear to most often be boys, but both males and females tend to be victims65. In their study of the relationship between bullying and suicidal ideation in adolescents, Brunstein Klomek et al.66 found that for girls, being bullied was associated with an immediate increased risk of suicidal ideation, while for boys, only chronic bullying was associated with suicidal ideation. However, similar gender differences were not verified in this meta-analysis, which may be related to the greater similarity in bullying behaviours experienced by boys and girls. In the bullying-victim group, gender differences in the relationship between bullying-victimization and NSSI could not be verified because the included articles did not provide the effect size needed to address this relationship. In conclusion, the current study suggests that NSSI in both boys and girls may be associated with bullying in equally strong ways.
Limitations and future implications
Although there are some highlights, several limitations must be considered. First, although the included studies were of relatively high quality, most of the data were collected in the form of self-reports by the subjects, which may cause recall bias. One possible limitation of this method is that it requires the interviewee to have a good level of self-awareness. In addition, some children who are bullied may tend to deny their status, while active bullies may be reluctant to admit that they are real bullies. Finally, the correlation between data from the same source (that is, when both the bullying experience and NSSI are self-reported by the child) may be exaggerated by the variance of the commonly used methods. One possibility for future research is that information should be collected through multiple independent information providers, such as children themselves, their peers within the classroom, and their teachers or parents. Additionally, an assessment of adolescents’ physical health may also be useful. Second, our meta-analysis is based on a cross-sectional study, and can only conclude that bullying is related to NSSI and cannot make any claims about the causal relationship between bullying and NSSI. Therefore, more longitudinal studies are needed in the future to test whether involvement in bullying increases NSSI or whether NSSI increases bullying behaviour. Third, since this paper only considers peer-reviewed articles, some grey literature may be missed, which may affect the results to some extent. Therefore, future research may further consider the inclusion of high-quality grey literature.
Fourth, in this meta-analysis, very few studies have measured different forms of bullying (verbal bullying, physical bullying, etc.) or did not report separate analyses for different forms of bullying. Although these two forms of bullying are not independent experiences types in the lives of children and adolescents. More precisely, they are two partially overlapping forms of harassment. However, recent research shows the importance of distinguishing between these two forms of bullying because they may have different relationships with personal adaptation67,68. Future research should further analyse the negative effects of physical and relationship or indirect bullying experiences on NSSI. The current analysis excludes research that only focuses on cyber bullying to focus on the risks of traditional forms of bullying; however, people know little about cyber bullying and cyber victimization, but they have become increasingly important in modern culture69. Future research should aim to determine whether the risk of cyberbullying and cyber victimization from NSSI is different from the traditional forms of these behaviours. Finally, only six studies were included in the bully-victims analysis. It is best to update the results on bully-victims when more research is available. Due to the small number of studies included and lack of information, we were unable to analyse the moderating effect of sampling strategy gender among bullying—victims.
Conclusions
Adolescent NSSI is a significant public health concern at the international level. Bullying behaviour is also an increasing concern particularly due to its associations with adverse health and behavioural outcomes. The negative outcomes of bullying behaviour, coupled with the negative outcomes of NSSI make both issues incredibly deleterious and pervasive. This study examined associations between bullying behaviour and NSSI. The findings support the fact that NSSI behaviours are related to bullying behaviors, and teenagers who participate in bullying may report more NSSI behaviours than those who do not. Considering the significant relationship between bullying and adolescent NSSI, NSSI problems in children and adolescents can be minimized by reducing bullying among children.
Clinical implications
The research reviewed supports the fact that bullies, victims, and bullying-victims are all associated with the risk of developing NSSI in adolescents. This result has significant implications for paediatricians, psychologists, and other health care professionals. It is very important that these professionals be able to identify adolescents who are at risk of being involved in NSSI because the potential negative health, psychological, and educational consequences are far-reaching. Based on significant association between bullying and psychosomatic problems, it has been suggested that when encountering psychosomatic complaints in children or adolescents, practitioners should regularly ask children or their parents about bullying to assess emotional function and peer experience70. Based on the results of the current study, we would like to extend this advice to NSSI. Therapy can also be provided for youth involved in bullying. Research has shown that when people receive the intervention of Treatment for Self-Injurious Behaviors (T-SIB) intervention, the frequency of NSSI decreases significantly71. At the same time, teenagers may choose not to disclose their involvement in bullying given the significant relation. NSSI could be used as a warning sign of children's involvement in bullying. Finally, interventions to reduce bullying should be considered. The need to support adolescents reporting bullying is a key priority in both school- and community-based settings in order to prevent further perpetration and/or victimization. Furthermore, understanding the factors that influence bullying perpetration and victimization is critical in reducing their incidence and persistence. Applying prevention science approaches that integrate population intervention studies and developmental research to examine modifiable risk and protective factors will assist in identifying and understanding these factors. Such understandings can then be used to inform evidence-based prevention and early intervention approaches targeted at bullying and its adverse consequences for both perpetrators and victims.
Data availability
The datasets used and/or analysed during the current study available from the corresponding author on reasonable request.
References
Zetterqvist, M. The DSM-5 diagnosis of nonsuicidal self-injury disorder: A review of the empirical literature. Child Adolesc Psychiatry Ment Health 9, 31 (2015).
Association, A. P. Diagnostic and Statistical Manual of Mental Disorders 5th edn. (American Psychiatric Association, 2013).
Swannell, S. V. et al. Prevalence of nonsuicidal self-injury in nonclinical samples: Systematic review, meta-analysis and meta-regression. Suicide Life Threat Behav. 44(3), 273–303 (2014).
Liu, S. et al. Emotion reactivity, nonsuicidal self-injury, and regulatory emotional self-efficacy: A moderated mediation model of suicide ideation. J. Affect. Disord. 266, 82–89 (2020).
Whitlock, J., Eckenrode, J. & Silverman, D. Self-injurious behaviors in a college population. Pediatrics 117(6), 1939–1948 (2006).
Zhu, J., Chen, Y. & Su, B. Non-suicidal self-injury in adolescence: Longitudinal evidence of recursive associations with adolescent depression and parental rejection. J. Adolesc. 84, 36–44 (2020).
Esposito-Smythers, C. et al. Clinical and psychosocial correlates of non-suicidal self-injury within a sample of children and adolescents with bipolar disorder. J. Affect. Disord. 125(1–3), 89–97 (2010).
Serafini, G. et al. Brain changes in early-onset bipolar and unipolar depressive disorders: A systematic review in children and adolescents. Eur. Child Adolesc. Psychiatry 23(11), 1023–1041 (2014).
Aboussouan, A. et al. Non-suicidal self-injury, suicide ideation, and past suicide attempts: Comparison between transgender and gender diverse veterans and non-veterans. J. Affect Disord. 259, 186–194 (2019).
Wilkinson, P. et al. Clinical and psychosocial predictors of suicide attempts and nonsuicidal self-injury in the Adolescent Depression Antidepressants and Psychotherapy Trial (ADAPT). Am. J. Psychiatry 168(5), 495–501 (2011).
Pompili, M. et al. Car accidents as a method of suicide: A comprehensive overview. Forensic. Sci. Int. 223(1–3), 1–9 (2012).
Hamada, S. et al. Association between bullying behavior, perceived school safety, and self-cutting: A Japanese population-based school survey. Child Adolesc. Ment. Health 23(3), 141–147 (2018).
Claes, L. et al. Bullying and victimization, depressive mood, and non-suicidal self-injury in adolescents: The moderating role of parental support. J. Child Family Stud. 24(11), 3363–3371 (2015).
Wright, M. F. The roles of bullying involvement and parental warmth in non-suicidal self-harm and suicidal ideation among adolescents from residential program. J. Crim. Psychol. 6(4), 202–213 (2016).
Olweus, D. Bully/victim problems among schoolchildren: Basic facts and effects of a school based intervention program. Dev. Treat. Childhood Aggress. 17(17), 411–448 (1991).
Heerde, J. A. & Hemphill, S. A. Are bullying perpetration and victimization associated with adolescent deliberate self-harm? A meta-analysis. Arch. Suicide Res. 23(3), 353–381 (2019).
Sourander, A. et al. Association of bullying behavior at 8 years of age and use of specialized services for psychiatric disorders by 29 years of age. JAMA Psychiat. 73(2), 159–165 (2016).
Nansel, T. R. et al. Cross-national consistency in the relationship between bullying behaviors and psychosocial adjustment. Arch. Pediatr. Adolesc. Med. 158(8), 730–736 (2004).
Hawton, K., Saunders, K. E. & O’Connor, R. C. Self-harm and suicide in adolescents. Lancet 379(9834), 2373–2382 (2012).
Meszaros, G., Horvath, L. O. & Balazs, J. Self-injury and externalizing pathology: A systematic literature review. BMC Psychiatry 17(1), 160 (2017).
Liang, H., Flisher, A. J. & Lombard, C. J. Bullying, violence, and risk behavior in South African school students. Child Abuse Negl. 31(2), 161–171 (2007).
Borenstein, M. et al. Introduction to Meta-analysis (Wiley, 2021).
Garisch, J. A. & Wilson, M. S. Prevalence, correlates, and prospective predictors of non-suicidal self-injury among New Zealand adolescents: Cross-sectional and longitudinal survey data. Child Adolesc. Psychiatry Ment. Health 9, 28 (2015).
Jutengren, G., Kerr, M. & Stattin, H. Adolescents’ deliberate self-harm, interpersonal stress, and the moderating effects of self-regulation: A two-wave longitudinal analysis. J. Sch. Psychol. 49(2), 249–264 (2011).
Hay, C. & Meldrum, R. Bullying victimization and adolescent self-harm: Testing hypotheses from general strain theory. J. Youth Adolesc. 39(5), 446–459 (2010).
Noble, R. et al. Safety first: The role of trust and school safety in non-suicidal self-injury. J. Education/Revue des sciences de l’éducation de McGill 46(3), 423–441 (2011).
Bakken, N. W. & Gunter, W. D. Self-cutting and suicidal ideation among adolescents: Gender differences in the causes and correlates of self-injury. Deviant Behav. 33(5), 339–356 (2012).
Giletta, M. et al. Adolescent non-suicidal self-injury: A cross-national study of community samples from Italy, the Netherlands and the United States. Psychiatry Res. 197(1–2), 66–72 (2012).
Jantzer, V. et al. Does parental monitoring moderate the relationship between bullying and adolescent nonsuicidal self-injury and suicidal behavior? A community-based self-report study of adolescents in Germany. BMC Public Health 15, 583 (2015).
Jiang, Y. et al. Buffering the effects of peer victimization on adolescent non-suicidal self-injury: The role of self-compassion and family cohesion. J. Adolesc. 53, 107–115 (2016).
Xavier, A. et al. Self-criticism and depressive symptoms mediate the relationship between emotional experiences with family and peers and self-injury in adolescence. J. Psychol. 150(8), 1046–1061 (2016).
Ji,. Attention bias in adolescents with being bullied and self-injury. Chin. J. Clin. Psychol. 25(04), 618–621 (2017).
Baiden, P., Stewart, S. L. & Fallon, B. The mediating effect of depressive symptoms on the relationship between bullying victimization and non-suicidal self-injury among adolescents: Findings from community and inpatient mental health settings in Ontario, Canada. Psychiatry Res. 255, 238–247 (2017).
Thomas, H. J. et al. Prevalence and correlates of bullying victimisation and perpetration in a nationally representative sample of Australian youth. Aust. N Z J. Psychiatry 51(9), 909–920 (2017).
Jiang, X. et al. Risk factors and cumulative effect of non-suicidal self injury behavior of rural senior high school students in Wuyuan. Chin. J. School Health 39(12), 1876–1878 (2018).
Cao, X. et al. Study on the prevalence and risk factors of non-suicidal self-injury among middle school students in Shenzhen. Sichuan Mental Health 32(05), 449–452 (2019).
Chen, X. et al. Relationships of bullying victimization and emotional behaviours to non-suicadal self-injury in middle school school student. J. Nanchang Univ. 59(01), 71–75 (2019).
Zhang, S. & Zhang, Y. The relationship between self-harm and Non-suicidal-self-injurious Behaviour among lelf-behind middle school students: Mediting effects of bullying and shame. Chin. J. Health Educ. 35(12), 1099–1103 (2019).
Esposito, C., Bacchini, D. & Affuso, G. Adolescent non-suicidal self-injury and its relationships with school bullying and peer rejection. Psychiatry Res. 274, 1–6 (2019).
Gaspar, S. et al. Non-suicidal self-injuries and adolescents high risk behaviours: Highlights from the Portuguese HBSC study. Child Indic. Res. 12(6), 2137–2149 (2019).
Yang, C. et al. Status and influencing factors of non-suicidal self -injury in junior students, Guizhou. Modern Prev. Med. 47(13), 2359–2363 (2020).
Zhang, S. & Zhang, Y. The impact of parental psychological control on non-suicidal self-injury: The role of campus bullying as a mediator. Chin. J. Dis. Control Prev. 23(04), 459–463 (2020).
Zhou, F. et al. Mediating role of help-seeking behavior in relationship between being bullied and non-suicidal self-injury in middle school students. J. Nanchang Univ. 60(06), 60–65 (2020).
Wang, Q. & Liu, X. Peer victimization and nonsuicidal self-injury among chinese left-behind children: The moderating roles of subjective socioeconomic status and social support. J. Interpers. Violence 36(23–24), 11165–11187 (2021).
Tong, W., Lei, Q. & Jiang, Q. Relationship between bullying victimization and non-suicidal self-injury in junior high school students: The mediating role of shame. China J. Health Psychol. 29(12), 1791–1796 (2021).
Mossige, S. et al. Suicidal ideation and self-harm among youths in N orway: Associations with verbal, physical and sexual abuse. Child Family Social Work 21(2), 166–175 (2016).
Lee, J. Y. et al. Non-suicidal self-injury is associated with psychotic like experiences, depression, and bullying in Korean adolescents. Early Interv. Psychiatry 15(6), 1696–1704 (2021).
Wu, N. et al. Bullying experiences and nonsuicidal self-injury among chinese adolescents: A longitudinal moderated mediation model. J. Youth Adolesc. 50(4), 753–766 (2021).
Orwin, R. G. A fail-safe N for effect size in meta-analysis. J. Educ. Stat. 8(2), 157–159 (1983).
Duval, S. & Tweedie, R. Trim and fill: A simple funnel-plot-based method of testing and adjusting for publication bias in meta-analysis. Biometrics 56(2), 455–463 (2000).
Egger, M. et al. Bias in meta-analysis detected by a simple, graphical test. BMJ 315(7109), 629–634 (1997).
Muellerleile, P. & Mullen, B. Sufficiency and stability of evidence for public health interventions using cumulative meta-analysis. Am. J. Public Health 96(3), 515–522 (2006).
Huedo-Medina, T. B. et al. Assessing heterogeneity in meta-analysis: Q statistic or I2 index?. Psychol. Methods 11(2), 193 (2006).
Munn, Z. et al. Methodological guidance for systematic reviews of observational epidemiological studies reporting prevalence and cumulative incidence data. Int. J. Evid. Based Healthc. 13(3), 147–153 (2015).
van Geel, M., Goemans, A. & Vedder, P. A meta-analysis on the relation between peer victimization and adolescent non-suicidal self-injury. Psychiatry Res. 230(2), 364–368 (2015).
Agnew, R. Foundation for a general strain theory of crime and delinquency. Criminology 30(1), 47–88 (1992).
Cullen, F. T. et al. Gender, bullying victimization, and juvenile delinquency: A test of general strain theory. Victims Offenders 3(4), 346–364 (2008).
Agnew, R. Building on the foundation of general strain theory: Specifying the types of strain most likely to lead to crime and delinquency. J. Res. Crime Delinq. 38(4), 319–361 (2001).
Hilt, L. M. et al. Longitudinal study of nonsuicidal self-injury among young adolescents: Rates, correlates, and preliminary test of an interpersonal model. J, Early Adolesc. 28(3), 455–469 (2008).
Scholte, R. H., Burk, W. J. & Overbeek, G. Divergence in self- and peer-reported victimization and its association to concurrent and prospective adjustment. J. Youth Adolesc. 42(12), 1789–1800 (2013).
Adrian, M. et al. Emotional dysregulation and interpersonal difficulties as risk factors for nonsuicidal self-injury in adolescent girls. J. Abnorm. Child Psychol. 39(3), 389–400 (2011).
Scheithauer, H. et al. Physical, verbal, and relational forms of bullying among German students: Age trends, gender differences, and correlates. Aggress. Behav. 32(3), 261–275 (2006).
Lam, S. F. et al. A latent class growth analysis of school bullying and its social context: The self-determination theory perspective. Sch. Psychol. Q. 30(1), 75–90 (2015).
Low, S. & Espelage, D. Differentiating cyber bullying perpetration from non-physical bullying: Commonalities across race, individual, and family predictors. Psychol. Violence 3(1), 39 (2013).
Rodkin, P. C. & Berger, C. Who bullies whom? Social status asymmetries by victim gender. Int. J. Behav. Dev. 32(6), 473–485 (2008).
Brunstein Klomek, A. et al. Bullying, depression, and suicidality in adolescents. J. Am. Acad. Child. Adolesc. Psychiatry 46(1), 40–49 (2007).
Storch, E. A. & Ledley, D. R. Peer victimization and psychosocial adjustment in children: Current knowledge and future directions. Clin. Pediatr. 44(1), 29–38 (2005).
Crick, N. R. & Bigbee, M. A. Relational and overt forms of peer victimization: A multiinformant approach. J. Consult. Clin. Psychol. 66(2), 337–347 (1998).
Livingstone, S. & Brake, D. R. On the rapid rise of social networking sites: New findings and policy implications. Child. Soc. 24(1), 75–83 (2010).
Gini, G. & Pozzoli, T. Association between bullying and psychosomatic problems: A meta-analysis. Pediatrics 123(3), 1059–1065 (2009).
Andover, M. S. et al. An intervention for nonsuicidal self-injury in young adults: A pilot randomized controlled trial. J. Consult. Clin. Psychol. 85(6), 620 (2017).
Acknowledgements
Yueming Ding is the co-first author
Funding
This research was sponsored by Graduate Education Reform and Quality Improvement Project of Henan Province (Grant Number: YJS2021AL074), Graduate Education Innovation and Quality Improvement Project of Henan University (Grant Number: SYL19060141), Planning and Decision Consultation Project of Henan Province (Grant Number: 2018JC38), Key Program of Research and Practice on Undergraduate Teaching Reform of Henan University (Grant Number: HDXJJG2020-25) and The survey subject of Henan Federation of Social Sciences Circles: "Research on the Status Quo and Cultivation Mechanism of Social and Emotional Abilities of Youth in Henan Province" (Grant Number: SKL-2022-55).
Author information
Authors and Affiliations
Contributions
H. H. and Y. D. analyzed the data, and wrote the main manuscript text. X. W., Y. L. participated in writing of the manuscript. H. H., Y. Z. and G. L. conceived and designed this study. G. L. and C. C. extracted the data and modifed the paper. All authors reviewed the manuscript.
Corresponding authors
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher's note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary Information
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Huang, H., Ding, Y., Wan, X. et al. A meta-analysis of the relationship between bullying and non-suicidal self-injury among children and adolescents. Sci Rep 12, 17285 (2022). https://doi.org/10.1038/s41598-022-22122-2
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41598-022-22122-2
- Springer Nature Limited
This article is cited by
-
The mediating effect of bullying on parental–peer support matching and NSSI behaviour among adolescents
BMC Public Health (2024)
-
Peer victimization and non-suicidal self-injury among high school students: the mediating role of social anxiety, mobile phone addiction, and sex differences
BMC Psychiatry (2024)
-
Does Bullying Victimization Accelerate Adolescents’ Non-suicidal Self-injury? The Mediating Role of Negation Emotions and The Moderating Role of Submissive Behavior
Child Psychiatry & Human Development (2024)
-
Online and school bullying roles: are bully-victims more vulnerable in nonsuicidal self-injury and in psychological symptoms than bullies and victims?
BMC Psychiatry (2023)
-
Challenges for psychiatric nurses working with non-suicidal self-injury adolescents: a qualitative study
BMC Nursing (2023)