by Michael D. Anestis, M.S.
Recently, we have received multiple requests for additional articles on the impact of childhood abuse on its victims. As always, we are happy to oblige reader preference when possible and encourage anyone to contact us with suggestions for future articles. With respect to this particular topic, the requests have come, in large part, in response to two specific articles. In the June 2009 featured article, Sarah Fischer, Ph.D., of the University of Georgia, described data that demonstrated that a specific personality variable, negative urgency, mediates the relationship between sexual trauma and future eating disorder symptoms. In other words, the relationship between sexual assault status and the development of eating disorder symptoms was explained by high levels of negative urgency. Soon after the publication of that article, we posted a review of the relationship between childhood sexual abuse (CSA) and the development of psychopathology. This review, which was based on a highly controversial article, provided evidence that the relationship between CSA and psychopathology was, in large part, explained by familial environment, meaning that in a positive overall environment, childhood victims of sexual assault, on average, showed surprising resilience. The point in these two studies was not that sexual assault has no impact on negative outcomes, but rather that simply thinking about the relationship in that manner is too simplistic and that other variables play a pivotal role in determining the impact of abuse on later outcomes.
Building off of those articles, today I would like to discuss some data published in 2007 by several of my colleagues and myself. In this study, published in Behaviour Research and Therapy, Joiner et al. (2007) examined the relationship between specific forms of childhood abuse and lifetime number of suicide attempts. The forms of abuse examined in this study were childhood physical abuse, rape, molestation, and verbal abuse, with childhood defined as prior to age 15. The idea behind this study was that, while several studies had examined abuse as a whole in its relationship with suicide (Glowinski et al., 2001; Roy, 2003), resulting in mixed findings (Brent el al., 1994; Kaplan, Pelcovitz, Salzinger, Mandel, & Weiner, 1997), no data existed that examined how particular forms of abuse related to suicidal outcomes when controlling for other forms of abuse.
This study utilized data from the National Comorbidity Study, which administered a rather amazing amount of assessment measures to a sample of thousands of individuals, examining a number of variables related to mental health. For these analyses, we used data only from individuals who had completed all of the measures we intended to examine, leaving us with a sample of 5,828 people. Building off of Joiner's (2005) interpersonal-psychological theory of suicidal behavior, we anticipated that forms of childhood abuse that involved more intense levels of physical pain (rape and physical abuse) would be better predictors of lifetime suicide attempts than would less painful forms of childhood abuse (molestation and verbal abuse). As you might remember from our prior descriptions of Joiner's (2005) theory, individuals develop an acquired capability for suicide through repeated exposure to painful and provocative life events. Repeated exposure to such experiences results in habituation to physical pain and to the fear of death, thus increasing fearlessness and the chances that an individual will be able to overcome the evolutionary instinct towards self-preservation. As such, we anticipated that more painful forms of childhood abuse would be better predictors of later suicide attempts because they were more likely to contribute to the acquired capability for suicide.
Importantly, in testing this hypothesis, we were careful to control for a large number of variables, including each form of abuse. In other words, we were not simply interested in whether each form of abuse correlated with lifetime suicide attempts. Instead, we were interested in determining whether such a relationship existed above and beyond the effects of other variables on suicide attempts. Considering two variables in a vacuum prohibits researchers from determining whether other variables at least partially explain that relationship. When a significant relationship exists when controlling for other variables, however, confidence can increase dramatically. In this case, in addition to each form of abuse, we controlled for all of the following variables:
Participant age
Participant sex
Participant alcohol dependence
Participant substance dependence
Participant major depression
Participant dysthymia
Participant bipolar disorder
Participant generalized anxiety disorder
Participant specific phobia
Participant panic disorder
Participant agoraphobia without panic disorder
Participant post-traumatic stress disorder
Participant adult antisocial behavior
Participant number of depressive episodes
Maternal and paternal depression
Maternal and paternal anxiety
Maternal and paternal substance use
Maternal and paternal adult antisocial behavior
Whether the participants' parents were divorced prior to the participant's 15th birthday
Death of a parent
Paternal conflict
Income
Maternal and paternal history of suicide attempts
So, a significant relationship between a form of abuse and suicide attempts when controlling for all of these variables would mean that, even when you consider the impact of this enormous list of powerful variables, that particular form of abuse still explains a significant, unique amount of variance. This is a very stringent test of our hypotheses, as it could potentially understate the power of the link of each form of abuse and suicide attempts.
As anticipated, our results showed that childhood physical abuse and rape exhibited a stronger relationship to lifetime suicide attempts than did molestation or verbal abuse. This is not to say that verbal abuse and molestation are unimportant to consider, but rather, that painful forms of abuse are more likely to contribute to this particular outcome once we consider the impact of all the other variables controlled for in this study.
The important question to consider when looking at these results is why more physically painful forms of abuse would exhibit such a strong relationship to suicide attempts, particularly relative to other forms of abuse. This data does not allow for a definitive answer to that question, but as hinted at above and explained in greater detail in prior articles, one plausible explanation with a strong level of empirical support is the acquired capability for suicide. Simply put, suicide attempts are typically very painful and the fear of death is a powerful deterant to such behaviors. When an individual is repeatedly exposed to physical pain - either his or her own pain or the pain of others - that individual gradually habituates to physical pain, such that it takes more for them to initially sense pain and they can tolerate a greater amount of pain once they feel it. Additionally, exposure to pain appears to have a similar impact on an individual's fear of death, particularly when the pain is self-inflicted. Although the pain in childhood abuse is not self-inflicted, the data indicate that they still have an impact similar to self-inflicted pain, increasing the likelihood that an individual will be able to overcome the pain and fear associated with a suicide attempt.
So how can we use this data clinically to decrease vulnerability to suicide attempts in victims of painful forms of abuse? Unfortunately, thus far there is no data supportive of the idea that we can decrease an individual's acquired capability for suicide, so an alternative path must be taken. As such, the importance of the other two components of Joiner's (2005) theory - thwarted belongingness and perceived burdensomeness - can not be overstated. These two variables - one of which describes a sense on the part of the individual that he or she does not have meaningful connections to others or that others can not understand his or her experiences and the other of which describes a sense on the part of the individual that he does not make any meaningful contributions to the world - comprise what Joiner terms the desire for suicide. Unlike the acquired capability for suicide, the desire for suicide is malleable and can be addressed quickly in treatment. Clinicians treating victims of painful forms of childhood abuse should thus prioritize helping their client develop meaningful relationships with others (either finding new relationships or nurturing and re-examining older ones) and find ways to engage in behaviors that they feel would be meaningful contributions. Doing this will not eliminate suicide risk, but will serve as a form of a buffer against other risk factors, namely the acquired capability.
There were some methodological weaknesses to this study that should be acknowledged. First and foremost, the study was cross-sectional, meaning that no longitudinal data was collected. Because of this, the data can not explicitly tell us whether one action (abuse) caused another (suicide attempts). Additionally, although the relationships between forms of abuse and suicide attempts were statistically significant, the effect sizes were small. With a large sample like ours, finding a statistically significant relationship is easier, as statistical significance simply indicates the likelihood that the results would be found due to chance. The greater the sample size, the greater the confidence that these results reflect the general population and, as such, the greater the likelihood that a correlation will be significant. So, in a small sample, a correlation of .15 would likely not be significant. In a large sample, on the other hand, that same correlation would be. As such, our findings appear to represent a real phenomenon, but it is important to note that the relationship between each form of abuse and suicide attempts was relatively small.
On the whole, the take home message from these findings is that, while abuse is problematic in general, specifying the type of abuse is important, as different forms of abuse may have different impacts on particular outcomes. In particular, it appears important to consider how painful an individuals abuse experience was, as this has importance relative to the acquired capability for suicide.
If you would like to learn more about Joiner's interpersonal-psychological theory of suicidal behavior or about the treatment of suicidal clients, we recommend the following products, all of which are available through our online store and are good reads both for professionals and the general public:
- Why People Die by Suicide
- Treating Suicidal Behavior: An Effective, Time-Limited Approach (Treatment Manuals For Practitioners)
- Suicide Science: Expanding the Boundaries
Mike Anestis is a doctoral candidate in the clinical psychology department at Florida State University.
Subscribe in a reader
