Suicide is the second leading cause of death for young people between the ages of 15 and 29 years globally (WHO, 2014). In an effort to bring about meaningful reductions in suicide rates internationally, much work has focused on the identification of so-called ‘high risk’ groups for which suicide is particularly prevalent compared with the general population. These groups include (amongst others): males; people diagnosed with a mental illness; people who engage in frequent non-fatal self-harming behaviour; lesbian, gay, bisexual, transgender and intersex (LGBTI) persons; and indigenous persons.
Increasingly, persons exposed to bullying or interpersonal violence (IPV) (either as a perpetrator, victim, or both) are also receiving recognition as another group at high risk for suicide (WHO, 2014). Experiences of bullying at school (Holt MK et al., 2015) and in the workplace (Leach LS et al., 2017), for example, have both been associated with suicidal ideation in young people. Experiences of violence in childhood, including sexual abuse, physical abuse, and emotional neglect, have also been associated with suicidal ideation and behaviour in young people (Miller AB et al., 2013). There are even some suggestions that this association may be independent of whether or not the young person has been diagnosed with a psychiatric disorder (Sachs-Ericsson NJ et al., 2017), suggesting that experiences of IPV may represent an independent risk factor for suicidal ideation and behaviour in this age group.
To date, however, few studies have investigated associations between bullying and childhood IPV on completed suicide in young people. Additionally, the majority of evidence for the association between bullying or IPV comes from cross-sectional studies, meaning that the relative timing of the experience of bullying or IPV vis-à-vis the onset of suicidal ideation or behaviour cannot be untangled. Finally, no study has attempted to estimate the Population Attributable Risk (PAR) of exposure to bullying and/or IPV on suicide risk in this age group. This is an important limitation as the PAR indicates the proportion of attempted and completed suicide cases, occurring at the population level, that can be directly attributed to exposure to IPV and therefore provides valuable information on the likely public health burden of IPV.
In an attempt to provide some evidence for these remaining knowledge gaps, the authors of a recent review systematically searched the Cochrane Library, EMBASE, Medline, PsycINFO and Web of Science electronic databases to identify all population-based studies, including either cohort or case-control designed studies, that report the association between any form of IPV and attempted suicide or suicide death in young people between the ages of 12 and 26 (Castellví P et al., 2017). Two reviewers independently selected studies for inclusion in the review and extracted study information and data for inclusion in the meta-analyses.
Data on the association between each form of IPV and suicide attempts and suicide were pooled between studies using the Odds Ratio (OR) and the 95% Confidence Interval (CI). Data on the population attributable risk of exposure to each form of IPV for suicide attempts and suicide at the population level was calculated using the PAR. Finally, the authors assessed the quality of the studies included in their review using the Newcastle-Ottawa Scale.
A total of 29 independent studies were included in the quantitative meta-analysis. Whilst most studies were of good quality, six received a lower quality rating, mostly because unreliable methods had been used to determine the outcome of interest, suicide attempts and suicide. Additionally, a number did not report the method used to determine whether a young person had been exposed to IPV.
Associations with Suicide Attempt
A total of 21 studies investigated associations between any form of IPV and attempted suicide alone, mostly using a cohort design (72.7%). One further study investigated associations with both suicide attempt and completed suicide as a combined outcome. Given that this study also included suicide attempt as an outcome, the review authors combined this study with the remaining studies, giving a total of 22 studies reporting data on the association with suicide attempt.
The majority of these studies reported associations with childhood physical abuse (45.4%), followed by childhood sexual abuse (40.9%), neglect (18.2%), dating violence (18.2%), bullying (18.2%), and childhood emotional abuse (4.5%). Seven studies (31.8%) also report associations with community violence which the review authors define as “[i]nclud[ing] violence between individuals who are unrelated, and who may or may not know each other, generally taking place outside the home.”
- Physical abuse: 10 studies contributed to the meta-analysis, with the overall finding that young people reporting exposure to physical abuse in childhood were significantly more likely to attempt suicide than those not exposed to this form of IPV: Odds Ratio (OR) 2.53, 95% Confidence Interval (CI) 1.66 to 3.87. However, heterogeneity for this meta-analysis was considerable (86.5%) on the basis of the guidance contained in the Cochrane Handbook (Deeks JJ et al., 2011)
- Sexual abuse: 9 studies contributed to the meta-analysis, with the overall finding that young people reporting exposure to sexual abuse in childhood were almost four times more likely to attempt suicide than those not exposed to this form of IPV: OR 3.87, 95% CI 2.31 to 6.49. Again, heterogeneity was considerable (76.2%)
- Neglect: 4 studies contributed to the meta-analysis, with the overall finding that young people reporting neglect were not significantly more likely to attempt suicide than those not reporting exposure to this form of IPV: OR 1.76, 95% CI 0.71 to 4.36. Considerable heterogeneity was observed once again for this meta-analysis (75.2%)
- Dating violence: 4 studies contributed to the meta-analysis, with the overall finding that young people reporting dating violence were significantly more likely to attempt suicide than those not reporting exposure to this form of IPV: OR 1.65, 95% CI 1.40 to 1.94
- Bullying: 4 studies contributed to the meta-analysis, with the overall finding that young people reporting bullying were significantly more likely to attempt suicide than those not reporting exposure to this form of IPV: OR 2.39, 95% CI 1.89 to 3.01
- Emotional abuse: only 1 study contributed to this analysis, finding that young people reporting exposure to emotional abuse as a child reported a higher mean number of suicide attempts compared to those not reporting exposure to this form of IPV. However, the authors did not provide the raw data to support this claim. Nor was it clear from the information reported in the original study whether this difference was statistically significant
- Community violence: 7 studies contributed to the meta-analysis, with the overall finding that young people reporting exposure to violence in their communities were significantly more likely to attempt suicide than those not reporting exposure to this form of IPV: OR 1.48, 95% CI 1.16 to 1.87. Considerable heterogeneity was observed for this risk in the meta-analysis (76.5%).
Associations with Suicide
The authors were able to find only 3 case-control studies which examined associations between any form of IPV and suicide death. Of these, the majority (66.7%) reported associations with childhood maltreatment (i.e., physical abuse, sexual abuse, emotional abuse, and/or neglect). One further study reported associations with community violence.
- Maltreatment: 2 studies contributed to the meta-analysis, with the overall finding that young people reporting exposure to childhood maltreatment were over 10 times as likely to die by suicide compared to those not reporting exposure to this form of IPV: OR 10.05, 95% CI 3.22 to 31.136. Only a low level of heterogeneity was observed for this risk factor (26.9%)
- Community violence: 1 study of two independent samples contributed to this analysis, with the overall finding that young people exposed to violence in their communities were significantly more likely to die by suicide compared to those not reporting exposure to this form of IPV: OR 23.30, 95% CI 3.16 to 171.54.
Population Attributable Risk of IPV for Suicide Attempt and Suicide
The review authors next calculated the population attributable risk of each form of IPV for suicide attempt and suicide death. This statistic provides an estimate of the proportion of suicide attempts and suicide that could be reduced if exposure to each of these forms of IPV could be eliminated from the population.
- Population Attributable Risk for Suicide Attempt: Bullying was responsible for the greatest population attributable risk for attempted suicide according to this review (22.2%), followed by sexual abuse (14.3%), physical abuse (8.6%), community violence (6.8%), dating violence (4.9%), and neglect (3.5%). Overall, the reviewers determined that the elimination of any form of IPV would reduce the suicide attempt rate, at the population level, by 9.0%.
- Population Attributable Risk for Suicide: The reviewers were unable to undertake an analysis of the population attributable risk of IPV for suicide.
The reviewers conclude that their review:
…provides strong evidence that early victims, from childhood to young adulthood, of any [form of] interpersonal violence (IPV), regardless of the type, are consistently more likely than non-exposed peers to present [with] suicide attempts…
They also suggest there is strong evidence, albeit on the basis of a more limited number of studies, that exposure to IPV during childhood may also be associated with an increased risk of suicide death.
Additionally, given that each form of IPV investigated in this review was associated with a relatively strong PAR, the authors conclude that efforts to improve the identification of children and young people at risk of exposure to IPV at the population level could bring about meaningful reductions in the rates of both attempted and completed suicide in this age group.
Strengths and Limitations
One limitation of using PARs to estimate the effect of reducing exposure to any putative risk factor, such as IPV, on a multifactorial process such as suicide is that the PAR assumes a causal relationship exists between the exposure of interest, in this case IPV, and the outcome of interest, in this case suicide attempt or suicide. This is because the interpretation of the PAR requires that eliminating exposure to the exposure of interest will leave the association with all other risk factors along the causal pathway unchanged (Li Z et al., 2011). Given reciprocal relationships between the experience of childhood IPV and mental illness, alcohol and other drug use, and other risk factors that have also been associated with an increased risk of attempted or completed suicide, this assumption is unlikely to be tenable. Nevertheless, as the reviewers rightly state, their review provides valuable information to guide the development of universal prevention programs and adds weight to recent calls highlighting those with a history of IPV as group at particularly high risk for suicide (WHO, 2014).
Considerably heterogeneity between studies was observed for a number of the meta-analyses. Whilst, in part, this likely reflects the variety of different studies, in terms of design, methodology, setting and so on, there remains the suggestion that there may be considerable variability in the association between IPV and suicide attempt and suicide. Further work, such as gene-environment studies, will be necessary to unravel which young people are particularly vulnerable to suicide following an experience of IPV.
Lastly, publication bias was observed for a number of the meta-analyses included in this review; particularly for exposure to childhood physical abuse and childhood sexual abuse. The reviewers suggest this is most likely due to the fact that studies reporting a significant positive association between IPV and suicide attempt and suicide are significantly more likely to be published than those reporting a null, or even negative, effect. As the reviewers acknowledge, this may mean the effect size estimates for some of their meta-analyses may be inflated. Nevertheless, when the authors imputed effect size estimates in an attempt to mimic what might happen if further studies reporting a null or negative effect were to come to light, the association between childhood sexual abuse and risk of attempted suicide remained significant.
A number of school-based interventions have recently been developed to equip young people themselves with the knowledge and confidence to protect themselves from IPV (Topping KJ & Barron IG, 2009). However, these programs have typically been evaluated in terms of their ability to improve recognition and knowledge around IPV in a vignette. The extent to which these programs will prove to be effective in reducing the incidence of IPV against young people in the community remains to be determined (Rudolph J & Zimmer-Gembeck MJ, 2016).
Nevertheless, this review helpfully synthesises the current state of knowledge regarding the likely strength of association between exposure to various forms of IPV and the risk of attempted or completed suicide in young people, finding that the elimination of any form of IPV may have the potential to reduce the suicide attempt rate, at the population level, by as much as 9.0%. This, in turn, is likely to have significant effects in terms of reducing morbidity and mortality in this age group.
Castellví P, Miranda-Mendizábal A, Parés-Badell O, Almenara J, Alonso I, Blasco MJ, Cebrià A, Gabilondo A, Gili M, Lagares C, Piqueras JA, Roca M, Rodríguez-Marín J, Rodríguez-Jimenez T, Soto-Sanz V, Alonso J. (2017). Exposure to violence, a risk for suicide in youths and young adults. A meta-analysis of longitudinal studies. Acta Psychiatria Scandinavica, 135: 195-211.
Deeks JJ, Higgins JPT, Altman DG. 2011. Analyzing data and undertaking meta-analyses [Chapter 9]. In JPT Higgins and S Green (eds). Cochrane Handbook for Systematic Reviews of Interventions, version 5.1.0.
Holt MK, Vivolo-Kantor AM, Polanin JR, et al. (2015). Bullying and suicidal ideation and behaviors: A meta-analysis. Pediatrics, 135: e496-e509.
Leach LS, Poyser C, Butterworth P. (2017). Workplace bullying and the association with suicidal ideation/thoughts and behaviour: A systematic review. Occupational and Environmental Medicine, 74: 72-9.
Li Z, Page A, Martin G, Taylor R. (2011). Attributable risk of psychiatric and socio-economic factors for suicide from individual-level, population-based studies: A systematic review. Social Science and Medicine, 72: 608-16.
Miller AB, Esposito-Smythers C, Weismoore JT, Renshaw KD. (2013). The relation between child maltreatment and adolescent suicidal behavior: A systematic review and critical examination of the literature. Clinical Child and Family Psychology Review, 16: 146-72.
Rudolph J, Zimmer-Gembeck MJ. (2016). Reviewing the focus: A summary and critique of child-focused sexual abuse prevention. Trauma, Violence, & Abuse. ePub ahead of print. DOI: 10.1177/1524838016675478.
Sachs-Ericsson NJ, Stanley IH, Sheffler JL, Selby E, Joiner TE. (2017). Non-violent and violent forms of childhood abuse in the prediction of suicide attempts: Direct or indirect effects through psychiatric disorders? Journal of Affective Disorders, 215: 15-22.
Topping KJ, Barron IG. (2009). School-based child sexual abuse prevention programs: A review of effectiveness. Review of Educational Research, 79: 431-64.
World Health Organization. (2014). Preventing Suicide: A Global Imperative. Geneva, Switzerland: World Health Organization.