Personality disorders
The assessment of PDs, based on the DSM-IV classification and structured diagnostic instruments, is conflicting. It is possible individuals who meet the criteria for a particular Personality disorder meet as well the criteria for other Personality disorders. The new diagnostic approach in DSM-5 describes the Personality disorders as qualitatively distinct clinical syndromes. Nevertheless, in this study, diagnoses were based on MINI Interview and PDQ-4 Questionnaire because the use of the same instruments with other similar studies makes the results comparable and helps the scientific discussion on the association between violent crimes and psychiatric disorders.
PDs were diagnosed in the vast majority (89%) of the prisoners’ sample. The most common PD was Antisocial (42.5%). The likelihood of committing a violent crime or homicide–attempt homicide,was significantly greater among those with Schizoid or Schizotypal PD.
The high prevalence of psychopathology in the population of incarcerated offenders is well documented in the literature and also reflected in this study. PDs were the most common disorders among prison inmates in Italy [
10]. At least one type of PD was diagnosed in 61% of a prison population sample according to Coolidge et al. [
11]. Langeveld and Melhus reported that PDs were found in 80% of the prisoners. In the same study, antisocial PD was present in more than 60% of the study population [
24]. Results from a systematic review of 62 studies with a total sample of 23,000 prisoners reported that 65% of the population had PDs, and 47% had Antisocial PD [
7]. Similar results were reported in Greek populations by Fountoulakis et al. [
17] and Fotiadou et al. [
16].
In this study Antisocial PD was diagnosed in 42.5% of the participants, Borderline PD in 15.9% and Narcissistic PD in 7.8%. Histrionic PD was diagnosed only in 0.6% of the individuals. A definite dominance of “Cluster B” PDs was obvious which is in concordance with the literature. Köhler et al. found that the prevalence of “Cluster B” PDs in a sample of male incarcerated juvenile offenders in Germany was up to 62%. Findings from this study were very similar to our results: the proportion of “Cluster B” PDs was 66.9%, whilst “Cluster A” was 16.2% and “Cluster C” 2.9%.
Personality disorders in relation to violent crimes
Of those diagnosed with Antisocial, Borderline and Narcissistic disorder in this study, 77.1%, 65.3% and 58.3%, respectively, had been imprisoned for non-violent crimes. Nevertheless, in absolute numbers, most of the violent offenses had been committed by inmates presenting Antisocial Disorder. However, a significant association of violent crimes and “Cluster B” PDs has not been established. According to Palmstierna, Antisocial PD and antisocial personality traits are connected with violence [
25]. Similarly Pondé et al. and González et al. suggest a strong association of Antisocial and Borderline PDs with violent crime [
26,
27].
High incidence of “Cluster B” PDs is often seen in the literature, although higher rates of “Cluster A” disorders have also been reported in prison’s populations, usually associated with a high prevalence of Paranoid PD [
9]. In this study, 16.2% of prisoners had “Cluster A” PDs; 1.9% Schizotypal; 7.1% Schizoid, and 7.1% Paranoid PD. Diagnosis with “Cluster A” disorders had an association with the commitment of violent crimes. Of those diagnosed with Schizoid, Schizotypal and Paranoid disorder 50.0%, 66.7% and 45.5%, respectively, have been incarcerated for violent crimes. These results are in concordance with Esbec and Echeburúa who reported that increased symptoms of DSM-IV “Cluster A” or “Cluster B” PDs, such as paranoid, narcissistic and antisocial symptoms are significantly associated with violence [
28]. Accordingly, Mouilso and Calhoun reported a strong association of Narcissistic PD with sexual assault [
29]. On the other hand, increased borderline personality tendencies have been reported in female sexual abusers [
30]. Serial offenders as well are more likely to have Narcissistic, Schizoid and/or Obsessive–Compulsive traits. They are also more likely to engage in sexual masochism, partialism, homosexual paedophilia, exhibitionism and/or voyeurism, according to Chan et al. [
31]. In addition, Pulay et al. reported an association between Schizoid, Paranoid and Obsessive–Compulsive PDs with violent behavior [
32]. Another study by Haller also suggested a significant association of paranoid disturbance with violent crimes [
33]. Schizoid PD is related as well with features of psychopathy and Antisocial personality according to Kosson et al. [
34]. Loza and Hanna argue that an association exists between Schizoid PD and violent acts [
35]. This study found a significant relation only of “Cluster A” disorders with violent offenses.
“Cluster C” disorders accounted for a minority of cases; individuals with Obsessive–Compulsive PD were only 2.3% of the study sample, whereas Dependent PD and Avoidant PD were few (~ 0.3%). This is in concordance with findings from Finland; “Cluster C” disorders comprised 3.5% of the entire sample of 593 offenders [
19]. However, in contrast with the aforementioned studies, there were no violent crimes committed by offenders with “Cluster C” PD in this study, possibly due to the very small proportion of this PD “Cluster” in the studied population.
As presented in logistic regressions’ results (Table
3), prisoners with any type of PD have greater likelihood (OR) of committing a violent crime. Nevertheless, violent crimes were associated significantly (
p ≤ 0.05) only with Schizotypal and Schizoid PD likely because the comparison group was composed by another type of criminals, instead of being composed by the general population.
Diagnosis with Personality disorders and association with homicides–attempted homicides
PDs have been found as principal or secondary diagnosis between homicides and attempted homicides offenders [
37]. According to Pera and Dailliet, in a sample of 32 Belgian offenders 17 had an Antisocial PD, 8 a Borderline PD, 4 a Paranoid, and 2 a Schizoid PD [
38]. Also, in a sample of 36 convicted Jamaican murderers 66% had an Antisocial PD [
39]. Antisocial PD and substance use disorders were the most prevalent psychiatric diagnoses among prisoners that had committed or attempted homicide, as suggested by Kugu et al. [
13].
Concerning sexual murderers, they are often diagnosed with a PD, especially with Schizoid PD [
40]. Myers and Monaco and others also found an association of Sadistic PD (as described in DSM-IV) with sexual homicide [
41,
42]. Concerning serial homicide offenders, they are more likely to have Narcissistic, Schizoid and/or Obsessive–Compulsive traits according to Chan et al. [
31]. Loza and Hanna reported, as well, an association between Schizoid PD and violent homicidal behavior [
35]. Analysis of case reports by Jeffrey Dahmer and Dennis Nilsen underlined an association between schizoid personality traits with violent antisocial behavior [
43].
In children, schizotypal features elicit victimization from other children, which in turn predisposes to reactive retaliatory aggression [
44]. Lam et al. found that schizotypal personality traits (schizotypy) are associated with antisocial behavior [
45]. This relation is replicated in the literature linking Schizotypal Disorder with antisocial behavior and violent crime [
45].
Regarding homicide–attempted homicide in this study, the majority was committed by individuals suffering from Antisocial PD. Subjects of “Cluster A” PDs, in total, had committed 17 of the 46 crimes of this type. These results are in contrast to Keue and Borchard [
36] and Laajalo et al. [
19] studies that found no association between disorders of “Cluster A” and homicides. Prisoners of this study, with disorders of “Cluster A”, were 4.25 times more likely to commit murder, while individuals with “Cluster B” disorders were 1.41 times more likely to commit the particular offense, compared with subjects without PD. Specifically for Antisocial PD odds ratio was 1.35, for Borderline PD was 3.34, for Narcissistic PD was 0.36, for Paranoid PD was 2.96, for Schizoid PD was 5.26 and for Schizotypal PD was 8.80, compared with subjects without PD. However, the committed homicide–attempted homicide was significantly associated with only Schizotypal and Schizoid PD.
Possibly, there is a neurobiological contribution to the association between Schizoid and Schizotypal PD and commitment of homicides or violence crimes. In literature, Schizotypal traits are associated with high hostility levels [
46]. According to Raine et al. schizotypy was associated with total and reactive aggression but not with proactive aggression [
44]. Sexual murderers are often diagnosed with a Schizoid PD [
40]. Lam et al. [
45] suggested that orbitofrontal cortex gray matter mediated the effect of schizotypy on antisocial behavior by 53.5%. On the other hand, this association was not significant for prefrontal cortex sub-regions. These findings highlight the specificity of the orbitofrontal cortex in understanding the schizotypy–antisocial behavior relationship. A link between Schizoid PD and Schizotypal PD was suggested by Via et al. [
47]. According to them, persons with Schizoid PD–Schizotypal PD have greater bilateral white matter volume in the superior part of the corona radiata, close to motor/premotor regions, compared to healthy controls.
Schug et al. reported that reduced skin conductance orienting to neutral tones may reflect a neurocognitive risk factor, for both Antisocial and Schizotypal PDs that indirectly reflects a common neural substrate to these disorders [
48]. Other researchers reported that individuals with Schizotypal PD display heightened activation in the neural circuitry, involved in reward and decision making when viewing biological motion stimuli in addition to a positive correlation between increased blood oxygenation level-dependent, signal responses related to biological motions and clinical symptoms [
49]. These findings suggest that enhanced responses arise within the reward network for individuals with Schizotypal PD and are possibly related to the “peculiar” ways that individuals with Schizotypal disorder behave in social contexts. It might be the “unemotional and cold part” of individuals with Schizoid and Schizotypal PD that contributes to the increased occurrence of “lethal violence”.
This study addresses certain limitations. Although the number of the participants is quite large, a bigger sample of individuals would have enhanced our results. For example, association of Paranoid (“Cluster A”) and Narcissistic (“Cluster B”) PDs with violent crimes was slightly not statistically significant (p values were 0.09 and 0.07, respectively). Another limitation is that issues of “free” psychopathology and counter-transference were not addressed in the initial protocol. It is also likely that Personality disorders have been overestimated in this study as well as in studies using structured diagnostic instruments. Much higher prevalence possibly has been reported compared to clinically based studies.