Borderline personality disorder is characteristically associated with a broad variety of psychiatric symptoms and aberrant behaviors. In this edition of The Interface, we discuss the infrequently examined association between borderline personality disorder and criminality. According to our review of the literature, in comparison with the rates of borderline personality disorder encountered in the general population, borderline personality disorder is over-represented in most studies of inmates. At the same time, there is considerable variation in the reported rates of this Axis II disorder in prison populations, which may be attributed to the methodologies of and populations in the various studies. Overall, female criminals appear to exhibit higher rates of borderline personality disorder, and it is oftentimes associated with a history of childhood sexual abuse, perpetration of impulsive and violent crimes, comorbid antisocial traits, and incarceration for domestic violence.
This ongoing column is dedicated to the challenging clinical interface between psychiatry and primary care—two fields that are inexorably linked.
The year is 1992. The movie is, “Single White Female.” The storyline entails protagonist, Hedra Carlson (Jennifer Jason Leigh), who appears to suffer from borderline personality disorder (BPD). She attempts to copy the appearance and characteristics of sequential roommates in a guilty attempt to virtually recreate her dead twin sister, who drowned during a family picnic. As the movie unfolds, the viewer is introduced to the fact that a previous roommate did not “work out,” so Hedy killed her.
Clinicians have long been aware of associations between mental illness and criminality. Indeed, according to a recent report by 57 independent monitoring boards of prisons in the United Kingdom (UK), 90 percent of inmates have at least one diagnosable mental disorder.1 While a number of Axis I disorders, such as bipolar disorder, are clearly represented in prison populations, there are also various Axis II disorders among the incarcerated. Importantly, these Axis II disorders extend beyond antisocial personality disorder—the traditional prison “personality.” In this edition of The Interface, we focus on one specific personality disorder, BPD, and its associations with criminality.
The Prevalence of BPD in Prison Populations
Studies indicating an over-representation of BPD in prison populations. The prevalence of BPD in the general US population ranges between two percent with rates reportedly greater in women (i.e., the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, Text Revision)2 and six percent with rates approximately equal between the sexes (the recent findings of Grant et al3). In comparison, we found that there is substantial empirical evidence that BPD is over-represented in prison populations. Although not intended as a full review of the literature, the following studies provide a general sense of the high prevalence of BPD in various prison populations.
In an interview-based study, Jordan et al4 examined female felons (N=805) who were newly admitted to a North Carolina prison. They found that 28 percent of these inmates met the criteria for BPD. In a Spanish study by Riesco et al,5 researchers examined 56 male prisoners with a structured personality disorder assessment and determined that 41 percent suffered from BPD. In a study comparing the diagnostic efficacy of two measures of personality disorders, Davison, Leese, and Taylor6 examined male prisoners in two UK prisons. On one measure, the prevalence of BPD in these populations was 45.7 percent, and on the second, 47.4 percent. Among imprisoned men convicted of sexual offenses, Dunsieth et al7 found that 28.3 percent of 113 participants met the criteria for BPD based on structured clinical interviews. Using a structured interview for diagnosis, Black et al8 examined the rate of BPD in newly admitted prisoners to the Iowa Department of Corrections. In this sample, 29.5 percent met the criteria for BPD; the rate among women was twice that encountered in men. Collectively, these studies— all using specific measures for personality disorder assessment— suggest that approximately 25 to 50 percent of prisoners suffer from BPD.
There are two studies of offenders in which researchers examined rates of BPD in a specific population— those in prison substance abuse treatment programs. In the first, Zlotnick et al9 systematically examined 272 offenders in these treatment programs and found that 8.3 percent of male and 20.7 percent of female participants suffered from BPD. In the second study of 280 participants, using a structured interview, Grella et al10 found that 13 percent of offenders in prison substance abuse treatment programs evidenced BPD.
Studies with differing rates depending on gender. Some studies have found differing rates of BPD in prison populations based on gender, with rates among men approximating the general population and rates among women being far higher. For example, using psychological testing, Burke examined 8,574 male and 894 female prisoners and determined that the prevalence of BPD traits was 5.3 percent and 11.5 percent, respectively.11 In a German study, using structured clinical interviews, von Schonfeld et al12 examined the rates of BPD among both male (n=76) and female (n=63) prisoners; the overall rate of BPD in this population was 22.3 percent. However, men demonstrated rates comparable to the general population (5.3%), whereas women had exceedingly higher rates (42.9%).
Studies refuting an over-representation of BPD in prison populations. We were able to locate two studies that found rates of BPD in prison populations that were comparable to community samples. In the first, a study of Cluster B disorders, researchers initially screened 802 female inmates.13 In the second phase of the study, all Cluster-B-positive patients (261) underwent a structured interview. Of the initial 802 subjects, ultimately only 5.2 percent evidenced BPD; both antisocial and paranoid personality disorders exceeded this percentage. While these data appear to refute the preceding findings, note that the initial sample of 802 participants did not undergo structured interviews for personality disorder assessment. It is possible that the screening measure in this study underdetected the rates of BPD in this sample.
In the second study, researchers retrospectively examined diagnoses of patients admitted to forensic psychiatry units in the UK between 1988 and 1994.14 Of these admissions, only 5.7 percent met the criteria for BPD. Again, the methodology may have influenced the percentage of patients with BPD, based upon the following: (1) the historical nature of the data, which is potentially limited by the individuals who unsystematically recorded these data; and (2) the possibility that patients with personality-disorders-only are under-represented in bonafide forensic facilities (i.e., in these settings, patients with Axis I disorders, such as bipolar, schizophrenic, and dissociative identity disorders, may be predominant).
Conclusions. We may draw several general conclusions from the preceding data. First, a substantial majority of studies in this area support the impression of higher rates of BPD in prison populations than in community samples, with rates generally ranging between 25 and 50 percent. Second, in studies that simultaneously compared men and women, rates among women appear to be consistently higher than in men. Finally, while two studies found rates of BPD similar to rates in community samples, the respective methodologies may explain these uncharacteristically low rates.
Factors Related to a Heightened Likelihood of BPD in Inmates
A number of factors appear to be associated with a greater likelihood of the diagnosis of BPD in a given inmate. In this section, we will review these factors.
Being female. It appears that incarcerated women harbor higher rates of BPD than incarcerated men.9,11,12,14 Whether this is genuine or related to a subtle bias in the measures for this disorder is unknown.
Childhood sexual abuse. Sexual abuse in childhood has long been known to be a general but nonspecific contributory factor to the development of BPD in adulthood. The data in prison populations appear to mirror this literature. For example, Christopher, Lutz-Zois, and Reinhardt15 examined 142 female inmates to examine contributory variables to BPD. In this sample, 61 participants were sex offenders and 81 were not. As expected, sexual abuse in childhood was associated with the diagnosis of BPD. In addition, participants in the sex-offender subsample were significantly more likely to report such histories.
Violent offenses. Borderline personality is associated with longstanding impulsivity and affective instability, including rage reactions. Therefore, one would suspect that more impulsive and violent offenders might be diagnosed with this disorder. Data seem to support these impressions.16,17 For example, Logan and Blackburn18 examined 95 women who had been incarcerated for violent offenses. Compared with women who had perpetrated minor violence, those with incarcerations related to major violence were four times more likely to be diagnosed with BPD. In keeping with these findings, Hernandez-Avila et al19 examined 370 alcohol/substance-dependent patients for criminal behavior and found that the diagnosis of BPD was associated with a greater number of pretreatment violent crimes.
While few studies have systematically examined the prevalence of BPD in those who commit homicide, Yarvis20 reported that BPD was one of the more common psychiatric diagnoses in a series of 100 murderers. In a British study of 90 men who were incarcerated for the murder of their female partner, Dixon et al21 found that 49 percent had borderline personality characteristics. In a German study, Hill et al22 examined individuals convicted of one-time sexual homicides and found that BPD was well represented. In contrast to these studies, in a French study, Pera and Dailliet23 found that only eight percent of 99 murderers suffered from BPD.
A number of authors have speculated about associations between variations of BPD (i.e., subtypes) and murderous acts. For example, Ansevics and Doweiko24 present the perspective that serial murderers represent a subtype of BPD, highlighted by manipulativeness. Cartwright argues that rage-based murders are related to a particular subtype of BPD characterized by elements of over-control.25 Finally, Papazian26 discusses the role of BPD in the serial killer. In summary, while not definitive at this juncture, the majority of current data and impressions indicate an association between BPD and the impulsive, rage-fueled murder.
Antisocial personality traits/disorder. Comorbid antisocial personality features may be associated with a heightened risk of criminality in individuals with BPD. In this regard, Howard et al27 compared those with mixed antisocial and borderline personality traits to individuals with various types of other personality dysfunction, including antisocial or borderline personality only. They found that the mixed cohort demonstrated higher trait anger, trait impulsivity, and aggression scores, resulting in an overall higher score on psychopathy.
Researchers have also compared criminals with antisocial versus borderline personality disorders and found some differences in the nature of their crimes. For example, de Barros et al28 found that while antisocial individuals tend to engage in more property crimes, borderline individuals tend to exhibit more episodes of aggression and physical violence. The authors concluded that criminals with pure antisocial personality are more calculating and exhibit more detailed planning, whereas those with BPD experience more impulsive and explosive episodes of violence. Again, one would assume that combining the two disorders would result in a very criminally combustible outcome.
Domestic violence. An association between BPD and partner violence has peppered the empirical literature over the past decade or so. For example, compared to nonbatterers, Else et al29 described a small sample of male batterers as evidencing comparatively higher scores on borderline and antisocial measures. Tweed and Dutton30 examined subtypes of male batterers and described a Type 2 profile with borderline personality characteristics. In a sample of 94 male batterers, Meyer31 found evidence of chronic personality dysfunction, again primarily with antisocial and borderline elements. Lawson et al32 examined 91 male batterers and identified three subgroups: (1) nonpathological, (2) borderline/dysphoric, and (3) antisocial. In developing a psychometric typology of male batterers in the UK, Johnson et al33 identified four types of offenders, one of which was BPD. In a similar vein, Chambers and Wilson34 examined 93 male batterers and found evidence for two clusters, one of which was BPD. Dutton35 summarized this literature in 2007 by indicating that many male batterers suffer from borderline personality as well as chronic symptoms of trauma.
Interestingly, the continual diagnostic appearance of BPD in male batterers is echoed in the literature on female batterers, as well. In this regard, Chavez36 examined both male and female batterers and found a higher prevalence of borderline personality characteristics in both when compared to nonbatterers. Stuart et al37 examined female batterers who were arrested for domestic violence and found that 27 percent met the criteria for BPD.
According to the findings of the majority of studies in this area, compared to rates expected in the community, BPD is over-represented in prison populations. This finding may be particularly evident among female prisoners. Rates vary, depending on the methodology, but generally appear to be in the range of 25 to 50 percent. Factors that may be associated with the presence of BPD among criminals include being female, having a history of childhood sexual abuse, committing an impulsive and violent crime (e.g., murder), having antisocial personality disorder traits, and perpetrating domestic violence. Given this association, clinicians in both mental health and primary care settings need to be aware of the possibilities of such histories in their patients with BPD.
Randy A. Sansone, Dr. R. Sansone is a professor in the Departments of Psychiatry and Internal Medicine at Wright State University School of Medicine in Dayton, Ohio, and Director of Psychiatry Education at Kettering Medical Center in Kettering, Ohio.
Lori A. Sansone, Dr. L. Sansone is a family medicine physician (government service) and Medical Director of the Primary Care Clinic at Wright-Patterson Air Force Base. The views and opinions expressed in this column are those of the authors and do not reflect the official policy or the position of the United States Air Force, Department of Defense, or US government.