The other week I came across an interesting study by Cathy Spatz Widom (whose work has helped debunk the myth that being a victim of sexual abuse causes individuals to then sexually abuse others), Sally Czaja, and Joel Paris published in the current issue of the Journal of Personality Disorders that touches upon a topic we have discussed several times on PBB: the link between childhood maltreatment and the development of psychopathology. The authors of this particular study were interested in the link between specific forms of childhood maltreatment and borderline personality disorder (BPD) and, in order to investigate this topic, they examined a sample of individuals who had documented cases of maltreatment as children as well as a matched sample of individuals with no such history to see whether maltreatment prospectively predicted later onset of BPD.
In past articles on this topic, we have approached the issue from several angles. Sarah Fischer of the University of Georgia featured article looked at whether negative urgency – the tendency to act rashly in an effort to reduce negative emotions – explained the link between sexual trauma and the later development of bulimia nervosa (BN). We also examined the degree to which other factors, such as experiential avoidance and family environment, explain the link between maltreatment and psychopathology. In each of these articles, the point was not to diminish the danger of maltreatment, but rather to understand the mechanisms through which maltreatment influences later outcomes and to identify protective factors that allow some maltreatment victims to avoid highly aversive outcomes.
In the study I would like to discuss today, the authors had a similar goal and accomplished this end through impressive methods. Widom and colleagues opened their article with a summary of the prevalence of BPD as well as research that has linked BPD with childhood maltreatment, particularly childhood sexual abuse (CSA; e.g., Herman, Perry, & van der Kolk, 1989; Ogata, Silk, Goodrich, Lohr, Western, & Hill, 1990). The authors followed up this discussion with an important point though: all of this prior work has relied upon retrospective recollections of maltreatment. In other words, adults already diagnosed with BPD have been asked to discuss their past. Such research is useful and has provided important insights, however, it is no replacement for prospective research that uses documented cases of childhood maltreatment to see if such experiences directly influence the development of later psychopathology.
In an effort to address this, the authors first drew from the court records of a metropolitan area in the Midwest between the years of 1967 and 1971 in order to identify documented cases of childhood abuse and neglect. The authors only included cases of abuse and neglect that were documented prior to age 11. From there, the authors also selected a comparison group comprised of individuals who matched the maltreated sample on the basis of age, sex, race/ethnicity, and the socioeconomic status during the time period under investigation in the study. Matches were found for 74% of the abused and neglected sample. The total sample of participants with complete data for all time points included 892 individuals (396 controls, 79 cases of physical abuse, 68 cases of sexual abuse, and 406 cases of neglect – some participants experienced multiple forms of maltreatment). Participants were interviewed on two occasions, once between 1989 and 1995 and once between 2000 and 2002. All diagnoses were established using structured diagnostic interviews based on DSM-III-R criteria (due to the year in which the study began).
So…what did they find? 14.9% of the maltreated group met criteria for BPD compared to 9.6% of the control group. Additionally, individuals in the maltreatment group exhibited a greater number of BPD symptoms than did individuals in the control group, particularly in the case of childhood neglect. Men with a history of childhood physical abuse or neglect were more likely than men with no such history to develop BPD. Women with abuse and neglect histories, however, were no more likely to develop BPD than were women with no such history. CSA was not linked to BPD in either sex. The lack of relationship between CSA and BPD in general and any maltreatment and BPD for women were both surprising findings and I will discuss them in more detail shortly.
Widom and her colleagues were not simply interested in determining whether childhood maltreatment predicted adult BPD. This is, of course, an interesting question, but it is also overly simplistic. The better question is whether maltreatment directly influences the development of BPD or if any such link is better accounted for by other variables. By addressing this question, we can develop a better understanding of why maltreatment victims develop BPD at a higher rate and how clinicians might be able to better address specific vulnerabilities. Along these lines, the authors found that parental drug or alcohol problems, as well as the victim not being employed full-time, not graduating from high school, and developing drug abuse or dependence, depression, or post-traumatic stress disorder (PTSD) all mediated the link between childhood maltreatment and BPD. In other words, the link between those two variables was actually explained by the degree to which participants fell into any of the categories listed in the previous sentence.
What does this mean, though? Put simply, it means that there are many paths from maltreatment to BPD – that maltreatment does not mean an individual is fated to develop BPD, but that the risk of such an outcome is determined by whether or not certain other variables are or become present. As such, childhood maltreatment might represent a marker for family dysfunction and the impact of parental psychopathology on the development of later psychopathology in the maltreatment victim.
Overall, these findings tell us several things. First, being a victim of maltreatment did, for particular groups, predict the later onset of BPD. Secondly, the link between maltreatment victimization and adult BPD was explained by a host of other variables, indicating that there might not be a direct effect of maltreatment on BPD, but rather that other variables potentially impacted by maltreatment actually determine BPD vulnerability. But what about the surprising lack of a link between CSA and BPD in general and any maltreatment and BPD for women? One issue might be that the comparison sample included highly elevated rates of BPD relative to most community samples (7.9% for men and 11.4% for women). The comparison group in this study, because it was matched on such demographics to the maltreatment victims, had an extremely low average level of socioeconomic status, which has been linked to BPD in prior studies (Taub, 1996). Additionally, it could simply be that, when documented cases of maltreatment are used to predict later onset of BPD rather than current BPD clients being quizzed about past episodes of maltreatment, different results emerge. This might reflect the possibility that maltreatment victims with BPD are more likely to seek help and show up in clinical samples, thus skewing the data spuriously or it might reflect an issue of inaccurate or poorly coded recollections. Alternatively, it might simply reflect an odd sample and, of course, these results would need to be replicated independently before we can look at them as representing an empirical fact.
This study represents yet another in a growing list of investigations that have yielded results indicating that maltreatment itself does not cause many of the poor outcomes associated with victimization. This means that there are many other vulnerabilities associated with maltreatment that better explain the emergence of psychopathology and that clinicians thus have an opportunity to address a host of vulnerabilities and potentially help clients avoid falling into the grip of mental illness. At times, some people have read findings like this and come to the misguided conclusion that the authors are diminishing the impact of childhood maltreatment or, even worse, indicating that such treatment is acceptable. Please know, however, that this is not the case. The authors are merely explaining that maltreatment exerts its influences through alternative variables that, in and of themselves, might serve as particularly important targets in therapy. Additionally, they are helping us to better understand why some individuals are so resilient and manage to make it through horrific experiences of maltreatment without developing any form of mental illness.
This post is by Mike Anestis who is a doctoral candidate in the clinical psychology department at Florida State University