Non-suicidal self-injury (NSSI), which refers to direct and deliberate harm of bodily tissue in the absence of suicidal intent, is a major public health problem in the U.S. and around the world. Data suggest that approximately 4% of adults in the U.S. population exhibit NSSI (Briere & Gil, 1998
; Klonsky, et. al., 2003
), and that adolescents are at even higher risk, with approximately 12–21% reporting a lifetime history of NSSI (Ross & Heath, 2002
; Whitlock et. al., 2006
; Zoroglu et al., 2003
). Recent research has begun to systematically describe the form and function of NSSI (Brown, Comtois & Linehan, 2002
; Nock & Prinstein, 2004
); however, the potential pathways to this behavior are not well understood. One consistently reported relation is that between a history of child maltreatment, defined here as neglect and abuse during childhood, and the development of NSSI. For instance, sexual abuse has shown a strong association with different forms of self-injury, including NSSI (Bergen et. al 2003
; Nock & Kessler, 2006
; Peters and Range, 1995
; Yates, 2004
), while physical abuse has been associated with such outcomes in some studies (e.g., Joiner et al., 2007
) but not others (e.g., Nock & Kessler, 2006
There are at least two important limitations of prior research in this area. First, the relation between different forms of child maltreatment and NSSI remains unclear, as most studies examine only one type of neglect or abuse. This makes it difficult to compare results across studies. Examining the relations between specific forms of child maltreatment and NSSI within a single study provides more detailed information about the relative magnitude of these associations and may facilitate greater understanding of how and why these constructs are related. Second, beyond knowing that child maltreatment is related to NSSI, it is important scientifically and clinically to better understand the factors that mediate or explain this relation. The potential mechanisms involved may vary depending on the type of neglect or abuse in question and variability among individual cases also is likely. Careful study and delineation of the mediators of these relations is an important and necessary step in understanding how NSSI might develop, and in informing prevention and intervention programs in the future.
Prior work suggests that people most often engage in NSSI for the purposes of emotion regulation or social communication (Brown et al., 2002
; Nock & Prinstein, 2004
). However, what has not been explained is why some individuals choose NSSI to achieve these ends rather than other behaviors that might serve similar functions, such as alcohol/drug use or bingeing/purging. One possibility is that some people select NSSI due to the directly self-injurious nature of this behavior, and that they learn to do so via modeling of earlier abuse by others. In other words, individuals who are excessively criticized and verbally or emotionally abused may, over time, learn to engage in excessive self-criticism and use NSSI as a form of direct “self abuse.”
For instance, people who experience maltreatment during childhood in the form of repeated insults, excessive criticism, or some form of physical abuse may come to adopt a similarly critical view of themselves over time through modeling the behavior of those who criticized and abused them. This could lead to the development of a self-critical cognitive style, and may ultimately manifest in the engagement in NSSI as an extreme form of self-punishment or self-abuse whenever they disapprove of their own behavior. NSSI is undoubtedly a multi-determined behavior that cannot be explained through one simple pathway such as this, but the examination of multiple pathways, such as the one proposed here, is necessary to begin to understand this dangerous behavior.
According to this model, the presence of child maltreatment is related to later engagement in NSSI during adolescence, and this relation is mediated by the presence of a self-critical cognitive style. We are proposing that it is self-criticism specifically that would mediate this relation given the directly self-abusive nature of NSSI. Beyond demonstrating mediation, the case for self-criticism as a mechanism through which maltreatment is associated with NSSI would be strengthened considerably if we were able to demonstrate specificity of this proposed mechanism by ruling out plausible alternatives (see Kazdin & Nock, 2003
). For instance, we would first want to show that the alternative model in which maltreatment mediates the relation between self-criticism and NSSI is not supported. In addition, we would want to rule out other plausible alternative mediators. For instance, it is possible that it is not self-criticism per se, but the more general perception of criticism from others could also explain (i.e., mediate) this relation. The demonstration of such a relation would weaken the case for proposing self-criticism as a specific mediator. It is also possible that the relation between child maltreatment, self-criticism, and NSSI could be explained by the more general presence of major depression. Confidence in the importance of self-criticism as a specific mediator of the relation between maltreatment and NSSI would be strengthened if such relations existed even after taking the presence of major depression into account. We sought to test this theoretical model as well as these alternative explanations of the data in the current study.
The first goal of this study was to document the extent to which different types of childhood abuse and neglect, including: sexual, physical, and emotional abuse, as well as physical and emotional neglect, are associated with the presence of NSSI. The second goal was to test a mediation model in which adolescent self-criticism mediates the relation between childhood maltreatment and the presence of NSSI during adolescence. Prior work by Hooley and colleagues (2002)
has shown that self-harming adolescents are more likely to tolerate physical pain in an experimental study if they believe that they are bad, flawed, and defective. The current study builds on this earlier work by (a) examining such self-criticism as a potential mediator in the relation between childhood maltreatment and NSSI, and (b) by testing alternative mediators, including perceived criticism from others and the presence of major depressive disorder, in order to test the specificity of self-criticism as a mediator. Through this investigation we hope to further illuminate the determinants of NSSI and to delineate a potential pathway through which NSSI can occur.