1
MASTERTHESIS
Name: Rogier Fenger
Student number: 10698280
Date: 12-02-2016
Master: Clinical Forensic Psychology
Mentor: Lieke Nentjes
External supervisors: Annemieke ter Harmsel and Lise Swinkels
Trauma, psychopathy and aggression in an offender
population: the role of psychopathy in the relation
2
INDEX
ABSTRACT ... 3
INTRODUCTION ... 4
Trauma and aggression ... 4
Psychopathy and aggression ... 5
Trauma and psychopathy ... 7
Trauma, psychopathy and aggression ... 7
Practical relevance ... 9 METHOD ... 9 Participants ... 9 Procedure ... 10 Instruments ... 10 Data-analysis ... 11 RESULTS ... 12
Trauma and aggression ... 12
Psychopathy and aggression ... 12
Trauma and psychopathy ... 14
Psychopathy mediation on trauma and aggression ... 14
Psychopathy moderation on trauma and aggression ... 15
Trauma subtypes ... 16
Trauma subtypes mediation ... 18
Trauma subtypes moderation ... 19
DISCUSSION ... 20
Trauma and aggression ... 20
Psychopathy and aggression ... 21
Trauma and psychopathy ... 21
Trauma, psychopathy and aggression ... 23
Primary vs secondary psychopathy ... 23
Limitations ... 24
Recommendations ... 25
3
Abstract
This study examined the relation between psychopathic traits, trauma and aggression in an offender population. These three constructs often occur together, but their interrelationships seem unclear. The primary goals of this study were twofold. Do psychopathic traits explain the presumed relationship between trauma and
aggression? And do certain psychopathic traits have a protective effect on the relation between trauma and aggression? Also, relations between the three constructs were explored to improve further understanding. A distinction was made in trauma
(emotional neglect, emotional abuse, physical neglect, physical abuse, sexual abuse), psychopathy (F1: interpersonal, F2: affective, and F3: lifestyle) and aggression (reactive and proactive).
Concerning the relations between the three constructs, results showed traumatic experiences to be related to reactive aggression, but not to proactive aggression. Also, trauma was related to psychopathy. This relationship seemed to be explained by the lifestyle factor of psychopathy (F3), which highly correlated with trauma, whereas the interpersonal factor (F1) and affective factor (F2) were not significantly correlated with trauma. As hypothesized, F3 was a significant predictor for reactive aggression, while F1 and F2 showed no links with reactive aggression. In addition, nor F1 or F2 were related to proactive aggression, which was contrary to the expectations.
Concerning the primary goals, strikingly, F3 fully mediated the relation between trauma and aggression. Support for a possible protective effect of psychopathy factors (F1 and F2) on the relation between trauma and reactive
aggression was not found, although F2 had a borderline significant effect. Results and directions for future research are discussed, as well as limitations in this study.
4
INTRODUCTION
Psychopathy is a pervasive personality disorder characterized by deviant interpersonal, affective and lifestyle traits. Traits of psychopathy are frequently found in forensic populations and include, for example, manipulation, egocentrism,
callousness and a problematic lifestyle. While the general publics’ perception and representations by the media equal psychopaths to cold blooded murderers, a more nuanced image seems to be true (Cleckley, 1976, as cited in Patrick, & Zempolich, 1998). While psychopathic traits may manifest themselves in all sorts of ways, it is believed that people with psychopathic traits are responsible for a very large amount of aggression and violence in society. “Whilst criminal activity may not be central to the psychopathic construct there is a large body of evidence displaying a clear link between psychopathy and crime” (Hare, 1993; as cited in Gowlett, 2014, p. 1). When looking at possible precursors for aggression, trauma may play a relevant role. A majority of research on traumatic experiences and aggression, finds a connection between one another (Cima, Smeets, & Jellicic, 2008). Also, Craparo, Schimmenti and Caretti (2013) found that experiencing trauma can play a role in the development of more severe psychopathic traits. It seems that trauma, psychopathy and aggression might be associated with each other. This study will look at the associations between trauma, psychopathy, and aggression. It’s important however to examine the different subtypes between these constructs, since they seem to relate in unique ways.
Researching the relations between these constructs and their subfactors could assist in predicting future aggressive behaviour, identifying risky combinations, obtaining indications for specific interventions and clarifying the psychopathy construct.
Trauma and aggression
Studies on trauma and aggression are numerous and results show high correlations between the two. Generally, trauma scores are positively related to
violence in prison samples and criminals show significantly higher trauma scores than control groups (Cima, Smeets, & Jellicic, 2008; Sarchiapone, Carli, Cuomo,
Marchetti, & Roy, 2008). Since trauma and aggression are broad concepts, it’s meaningful to look at different subtypes when studying these factors.
Trauma and aggression subtypes
Sufficient data supports the distinction between two forms of aggression: (1) proactive or instrumental aggression represents a deliberate, cold-blooded form of aggression with the aim of achieving a certain goal, while (2) reactive aggression is a more emotional and defensive reaction on a threat, frustration or attack (Raine et al., 2006; Cima, Smeets, & Jelicic, 2008; Reidy, Shelley-Tremblay, & Lilienfeld, 2011). A link between trauma and reactive aggression seems clear. Ford, Chapman, Connor, and Cruise (2012) found that trauma increases the risk for distortions in biopsychosocial development and may lead to numerous adverse outcomes (e.g. psychopathological disorders), which are related to displaying reactive aggression. Early trauma can lead to a heightened stress-reactivity and hyperarousal (modulated by de HPA-axis) and is associated with heightened reactive aggression in both youngsters and adults (Oosterlaan, Geurts, Knol, & Sergeant, 2005).
Literature shows strong evidence for an association between trauma and reactive aggression, but not for proactive aggression (Ford, Chapman, Connor, & Cruise, 2012). Indeed, research shows mixed results, but there are several exceptions. For example, Kerig, Bennet, Thompson and Becker (2012) introduced ‘acquired
5 callousness’. Suggestions are made that children who have suffered from trauma could express callous-unemotional traits (via blunting) which are predictors for proactive aggression (Farrington, 2005; Kerig et al., 2012). Although callousness seems to be an inherited trait, manifesting and measurable at a very young age, it’s not impossible that it may be reinforced or acquired by chronic trauma. Emotional
detachment, a characteristic of callousness, could serve as a coping method for severe trauma (Hawes, & Dadds, 2007; Ford et al., 2012; Kerig et al., 2012). In addition, maladaptive parenting is often found to have moderate to high correlations with proactive aggression in children (Raine et al., 2006). A connection between trauma and proactive aggression might be due to blunting and callousness, which are predictors for proactive aggression.
Based on the studies described above, the current study expected a positive correlation between trauma and reactive aggression (Hypothesis 1a). In addition, a positive correlation between trauma and proactive aggression was hypothesized (Hypothesis 1b).
Psychopathy and aggression
A great deal of research supports a connection between psychopathy and aggression (Reidy et al., 2011). Psychopathic traits and aggressive behaviour have high correlations and these results are found in both offender and nonoffender groups (Porter & Woodworth, 2006). For example, Cima et al. (2008) conclude that
psychopathic traits were positively associated with aggression. Psychopathy subfactors and aggression subtypes
It seems that people with psychopathic traits display a proactive kind of aggression, whereas aggression by non-psychopathic individuals tends to be more impulsive and reactive (Porter & Woodworth, 2006; Reidy et al., 2011). These finding are not surprising since psychopathy and proactive aggression share a callous and instrumental foundation. You would expect that people with high psychopathic traits are also more inclined to engage in reactive aggression, since it’s a more common form of aggression which almost everyone is prone to when provoked sufficiently (Berkowitz, 1998; Reidy et al., 2011). However, a considerable part of the
psychopathy research does not find a connection with reactive aggression. In fact, several studies conclude that psychopathy might be a protective characteristic against displaying reactive aggression (e.g. Patrick et al., 1993; Patrick, 1994; Dempster et al., 1996; Anderson, 2003; Benning et al., 2005; Dolan, & Holi et al., 2006; Kiehl, 2006; O'Leary et al., 2007, 2010; Cima et al., 2008; Veit et al., 2010; as cited from Reidy et al., 2011).
Reidy et al. (2011) reviewed the relation between psychopathy and aggression in different fields of study (i.e., behavioural, cognitive, and biological) and found much support for an association between psychopathy and proactive aggression. However, research on the link between psychopathy and reactive aggression is inconclusive. Possible explanations could be found in the emotionally detached characteristic of psychopathy. Psychopaths’ hypo-activity in emotional situations might make it unlikely that these individuals engage in reactive aggression (Raine et al., 2006; Dadds et al., 2009; Reidy et al., 2011; Ford et al., 2012). Overall, there is overwhelming support that psychopathy is related to proactive aggression.
Psychopathy seems to have an associations with reactive aggression as well, although the associations appears less characteristic. In general, research on psychopathy and
6 aggression show some inconsistent results. The multifactorial structure of
psychopathy may be responsible for these inconsistencies.
The term ‘psychopathy’ remains subject to much disagreement. Cleckley (1976), Hare (1991), and Cooke and Michie (2001) all laid foundations for the
psychopathy construct, each with their own vision. While the construct is still heavily debated, there seems to be sufficient support to make a distinction between subfactors. This study follows the notion of Cooke and Michie (2001), who proposed a three-factor model of psychopathy. It differentiates an ‘arrogant, grandiose and deceitful interpersonal style’ (factor 1), a ‘deficient, callous and unemotional affective style’ (factor 2), and an ‘impulsive and irresponsible behavioural style’ (factor 3). More specific, the interpersonal style (F1) includes glibness, superficial charm, a grandiose sense of self-worth, pathological lying and manipulative traits. The affective style (F2) cover a shallow affect, callousness or diminished empathy, a lack of
remorse/guilt, and not accepting responsibility. F3, lifestyle, includes a need for stimulation, proneness to boredom, impulsivity, irresponsibility, a parasitic lifestyle and lack of realistic, long-term goals (Cooke, & Michie, 2001).
The few psychopathy studies that have been conducted on a subfactorial level, showed Factor 1 (interpersonal) and Factor 2 (affective) to have unique positive correlations with proactive aggression, whereas Factor 3 (lifestyle) was only linked to reactive aggression. These results were found in offender- and non-offender
populations (Dempster et al., 1996, as cited from Reidy et al., 2011; Woodworth & Porter, 2002; Reidy et al., 2011). Cornell et al. (1996) showed similar results, by linking F1 and F2 to proactive aggression, whereas F3 only had a connection with reactive aggression. This is not surprising, since the interpersonal (F1) and affective (F2) factors of psychopathy both share characteristics with proactive aggression. On the other hand, the lifestyle (F3) factor shares characteristics with reactive aggression (e.g. impulsivity, irresponsibility). Based on studies and findings as described above, the current study expected a positive relation between F1, F2 and proactive aggression (Hypothesis 2a) and a positive relation between F3 and reactive aggression
(Hypothesis 2b).
Interestingly, characteristics of psychopathy and aggression appear to actually be at odds with each other. There is some support for a possible protective role of psychopathy subfactors on reactive aggression. F1 and F2 have been “linked empirically to diminished threat reactivity, autonomic response, and sensitivity to fear, as well as heightened serotonergic function suggestive of low impulsivity” (Reidy et al., 2011; p. 250). While F3 has a strong connection with anger activation, a correlation for F1 and F2 with anger activation seems to be absent. Reidy et al. (2011) showed that F1 and F2 had negative relations with reactive aggression. Although the psychopathy construct as a whole has been linked to reactive aggression, its
subfactors (F1 and/or F2) may be responsible for a protective effect. Traits of F1 and F2 contain callous features, this could make it more unlikely to display reactive aggression which is emotionally charged. F1 and F2 include factors of a callous and calculating mind which can be considered at odds with the short tempered core of reactive aggression. Therefore this study hypothesized that no significant relation will be found between F1, F2 and reactive aggression (Hypothesis 2c). Overall, a positive relation between total psychopathy and aggression is hypothesized in the current study (Hypothesis 2).
7
Trauma and psychopathy
Craparo, Schimmenti and Caretti (2013) found that early exposure to relational trauma can play a relevant role in the development of more severe psychopathic traits. These results are in line with the understanding that parenting styles have an influence on the gradient of psychopathic traits (Lykken, 1995; Farrington, 2005). Also, higher scores on trauma questionnaires were found for criminals with high psychopathy traits (Craparo et al., 2013). Severe trauma seems to have connections with diminished empathy and antisocial behaviour (Craparo, Schimmenti, & Caretti, 2013). This study predicted a positive correlation between trauma and total psychopathy (Hypothesis 3) Trauma subtypes and psychopathy subfactors
Much support is given for a relationship between trauma and the lifestyle (F3) factor of psychopathy (Poythress, Skeem, & Lilienfeld, 2006). Trauma is correlated with several problems which are also included in the lifestyle factor of psychopathy, such as impulse control difficulties (Valentine, 2001). A connection between trauma and F3 is supported by many studies (Valentine, 2001; Poythress, Skeem, &
Lilienfeld, 2006; Cima, et al., 2008; Daversa, 2010; Craparo et al., 2013). Therefore, this study predicted a positive correlation between trauma and F3 (Hypothesis 3b).
In addition, some studies show that early traumatization may actually have an influence in the development of affective deficits as well. Krischer and Sevecke (2008) show that traumatic experiences can have adverse effects like a lack of empathy. Trauma may be a precursor to similar affective deficits to such an extent that they are measured by diagnostics instruments for psychopathy (Krischer & Sevecke, 2008). Daversa (2010) provides a biological model in which early childhood maltreatment (i.e., physical, sexual, emotional abuse) alters the neurobiology of the brain (the amygdala in particular). These alterations may lead to core features which are observed in the affective subfactor (F2) of psychopathy (Daversa, 2010). As previously mentioned, ‘acquired callousness’ may be a factor linking trauma and psychopathy F2 (Bennet, Thompson and Becker, 2012). Although trauma is
commonly associated with the lifestyle (F3) factor of psychopathy, it may also have an effect on the development of the affective factor (F2) and to a lesser extent the interpersonal factor (F1). Therefore, a positive correlation between trauma and F1, F2 was expected (Hypothesis 3a).
Trauma, psychopathy subfactors and aggression subtypes
A clearer understanding of trauma, psychopathy, aggression and their correlations can be obtained by including all three factors and their subtypes. The primary goal of this study is to examine the role of psychopathic traits in the presumed relationship between trauma and aggression. Several studies show support for a connection
between trauma and F1/F2 and proactive aggression. The relation between trauma and proactive aggression might be due to the interpersonal (F1) and affective (F2)
psychopathic factors. Therefore it was predicted for F1/F2 to mediate the relationship between trauma and proactive aggression (Hypothesis 4a).
Many studies support the relation between trauma and reactive aggression, but this relation might be due to the lifestyle (F3) psychopathic factor. This study
expected F3 to mediate the relationship between trauma and reactive aggression (Hypothesis 4b).
The main goal of this study was to examine if psychopathic traits have a protective effect on the relation between trauma and aggression. Characteristics of
8 F1/F2 and reactive aggression are opposing, making it less likely to coexist. A
diminished stress/threat reactivity and low impulsivity associated with F1/F2, counters the features for displaying reactive aggression (Reidy, et al., 2011). Also, F1/F2 had no connections with anger activation (Reidy, et al., 2011). Features of F1/F2 may suppress the tendency to engage in reactive aggression, even in individuals with considerable trauma scores. This study predicted F1/F2 to moderate the relation between trauma and reactive aggression and possibly providing a protective effect for a significant connection (Hypothesis 4c, main research question).
Figure 1. The hypothesized relationships and mediation between trauma, psychopathy interpersonal (F1) / affective (F2) and proactive aggression
Figure 2. The hypothesized relationships and mediation between trauma, psychopathy lifestyle (F3) and reactive aggression.
Figure 3. The hypothesized relationships and moderation between trauma, psychopathy interpersonal (F1) / affective (F2) and reactive aggression.
F1/F2
TRAUMA
AGGRESSION
PROACTIVE
F3
TRAUMA
AGGRESSION
REACTIVE
F1/F2
9
Practical relevance
The criminal population often displays considerable elevations on trauma, psychopathy and aggression. Extensive knowledge of possible associations may provide great progress in predicting future aggressive behaviour, identifying risky combinations, obtaining indications for specific interventions and clarification of the psychopathy construct. This study may provide a modest start for understanding the interactions between these trauma, psychopathy, aggression and their subfactors.
METHOD
Participants
Participants in this study were clients (Dutch speaking only) from ‘Inforsa Forensisch Ambulante Zorg’. Inforsa provides ambulant treatment for forensic clients with a great variety of psychiatric, addiction and/or personality problems. A joint factor between clients is involvement with the police or criminal justice system via violence, property, or sexual offenses. Participants showed different index offenses, including violent offenses (n = 32), non-violent offenses (n = 29) and
different/unknown offenses (n = 8).
The recruitment of clients took place from December 2014 till March 2015. The majority was male and had a legal treatment obligation. After the standard intake meeting, clients were approached for voluntarily participation in academic research. Clients were informed about the study and given the opportunity to ask questions. Anonymity was emphasized and an informed consent was handed out. A new appointment for administering the questionnaires was made when clients agreed to participate. Approximately 140 clients were approached, 75 clients completed all questionnaires. Participants with an IQ < 70 were excluded from this study, for more reliable results. Due to insufficient resources, intelligence tests could not be
administered to all participants. In the absence of recent IQ scores, clinical judgement from a professional health-care psychologist was used, where possible, to estimate IQ. Five participants were excluded based on their intelligence test scores, one participant was excluded based on clinical judgement. This resulted in a total of 69 participants (65 men; age M = 39.30, SD = 9.52), with IQ classifications: borderline impaired (N = 13), below average (N= 17), average (N = 27), above average (N = 3), unknown (N = 9).
Because of the great diversity in age, ranging from 23 to 60 years old, age was used as a covariate. Specific and recent IQ measurements were not available for all participants. This study considered the IQ classification estimations to be sufficiently reliable for participants’ inclusion, though not reliable enough for incorporation in the statistical analyses.
A power analysis was executed in G-power (Faul, Erdfelder, Lang, & Buchner, 2007). For the correlational analyses, a sample of 59 participants was needed in order to detect a medium effect size (r = .35, two tailed) with an alpha of .05 and a power of .80. The maximum number of predictors was 4. For the regression analyses a sample of 55 participants was needed in order to detect a medium effect size (f² = .15) with an alpha of .05 and a power of .80. The maximum number of predictors was 5. G-power results showed that the total sample size in this study
10 should be at least 59 participants. By having a total of 69 participants, this
requirement was met. Procedure
This study is part of a larger academic project that aims to evaluate and enhance the different treatment programs of the institution. Also the client population is explored to detect possible gaps in treatment provision. This larger academic project includes a total of 10 self-report and structured questionnaires1, including the Jeugd Trauma Vragenlijst (JTV; Arntz & Wessel, 1996), the Reactive Proactive Questionnaire (RPQ; Raine et al., 2006) and the Youth Psychopathic Traits Inventory – short version (YPI-s; van Baardewijk et al., 2010) that were used in this particular study. Participants were told that the academic research does not affect their
registration and treatment at Inforsa and were given the option to ask questions. A follow-up appointment to complete all questionnaires was often made due to, for example, impaired concentration or time shortage. The questionnaires were administered at Inforsa (Keizersgracht, Amsterdam) or by visiting clients at their home. Approximately 80% of the participants completed the questionnaires at Inforsa. The researcher read the questions out loud when clients experienced sight or linguistic difficulties. This number is estimated at 8 participants (N = 8).
Instruments
JTV
The ‘Jeugd Trauma Vragenlijst (JTV: Arntz & Wessel, 1996) is a Dutch translation from the Childhood Trauma Questionnaire (CTQ: Bernstein & Fink, 1997; Bernstein et al., 1994), which contains 27 questions. Participants evaluate statements on a five-point Likert scale, ranging from 1 (never true) tot 5 (very often true). The JTV uses opposing questions, scoring a ‘5’ (very often) on a particular question can sometimes deny or sometimes actually indicate a traumatic component. For analytical purposes, the questions were mirrored so that scoring a ‘5’ reflected a traumatic component. Administration time of the JTV is approximately 15 minutes.The JTV differentiates five subtypes of childhood abuse: ‘physical abuse, ‘physical neglect’, ‘emotional abuse’, ‘emotional neglect’, and ‘sexual abuse’. In previous research, the JTV has found to be a valid and reliable self-report instrument to asses childhood trauma (Thombs, Bernstein, Lobbestael, & Arntz, 2009; Van Harmelen, Elzinga, Kievit, & Spinhoven, 2011). In the current study, an analysis for internal consistency in this sample showed a Cronbach’s Alpha of .80, indicating a good reliability. When viewed on a subschale level, an internal consistency analysis in this study showed Cronbach’s Alpha’s of: physical abuse (.90), physical neglect (.71), emotional abuse (.91), emotional neglect (.90), and sexual abuse (.87).
This study mainly focussed on traumatic experiences as a whole. However, postdictive analyses were conducted for trauma subtypes to identify meaningful relationships and thereby encourage further research and understanding.
RPQ
The Reactive Proactive Questionnaire (RPQ: Raine et al., 2006) is a short self-report instrument used to assess reactive and proactive aggression in youngsters and adults. This study uses the Dutch version (Cima et al., 2013). It consists of 23 Likert-type questions (11 reactive, 12 proactive) on aggressive behaviour (rated 0 = never; 1
11 = sometimes; 2 = often). The RPQ takes approximately 10 minutes to conduct. Good psychometric properties have been described for the RPQ (Cima et al., 2013). A high reliability was found in this sample with a Cronbach’s Alpha of .86 and per subschale .90 (reactive) and .87 (proactive).
Although high intercorrelations between the two types of aggression have been found in previous studies (ranging from r = .41 to .76), sufficient research indicate a two-factor structure: reactive-proactive (Raine et al., 2006).
YPI-s
The Youth Psychopathic Traits Inventory – short version (YPI-s: van Baardewijk et al., 2010) is a self-report screening tool for psychopathic personality traits, consisting of 18 questions. Participants rate the questions on a four-point Likert scale from 1 (not applicable to me) to 4 (very applicable to me). YPI-s is derived from the original YPI instrument, developed by Andershed, Kerr and Stattin (2002).
Administering the YPI-s takes approximately 10 minutes. The YPI-s is based on Cooke and Michie’s three factor structure of psychopathy: (1) Interpersonal factor, grandiose and manipulative style; (2) Affective factor, callous and unemotional style; (3) Lifestyle factor, impulsive and irresponsible behaviour (Cooke & Michie, 2001; van Baardewijk et al., 2010). The instrument provides a total psychopathy score as well.
Van Baardewijk et al. (2010) conclude that the YPI-s is a practical and valid alternative to measure psychopathic traits (including callous-unemotional traits), especially when administration time is limited. Its structure has been proven reliable and valid by a sufficient number of studies. However, a reliability analysis in this study showed a Cronbach’s Alpha of .57, which can be interpreted as
poor/questionable. This result gives reason to believe that the three factor structure does not fully grasp the psychopathy construct. Perhaps further research is needed to optimize the psychopathy construct and its measuring instruments. When viewed on a subfactor level, the Cronbach’s Alpha’s for internal consistency in this study were .80 (interpersonal factor), .71 (affective factor), and .74 (lifestyle factor).
Data-analysis
It is important to note that this study, as well as the YPI-s, approached
psychopathy in a dimensional way. People experience psychopathy traits to a greater or lesser extent. The same is true for the JTV and the RPQ. A dimensional approach, instead of categorical (cut-off point) was applied.
Data analysis was conducted using SPSS v20 statistical software. A
significance level of p < .05 was set for all analyses. Cross-sectional research shows age to influence aggressive behaviour. Aggression seems to peak in early adolescence and is less frequently observed in older age categories (Lahey et al., 2000). The current study conducted analyses with and without age as a covariate variable. A significant negative relation between age and total aggression, proactive aggression and reactive aggression was found, as described in the results section. Based on these findings, all analyses were conducted with age as a covariate variable.
Participants could not be distinguished based on index offense. An independent-samples T-tests showed no significant results for index offense and trauma (t = 0.80, p = .43) or psychopathy total (t = 1.77, p = .08), psychopathy F1 (t = 1.42, p = .16), psychopathy F2 (t = 0.48, p = .64), and psychopathy F3 (t = 1.89, p = .06). Also, no significant results were found for index offense and aggression (t =
12 1.53, p = .13), with reactive (t = 1.86, p = .07), and proactive (t = 0.95, p = .35). Hence, further classification was discarded.
This study conducted several analyses. The relationships between trauma and aggression (hypothesis 1) and trauma and psychopathy (hypothesis 3) were examined by using a Pearson partial correlation. Possible predictive effects of psychopathy on aggression (hypothesis 3) were examined by conducting Multiple Regression Analyses. The role of psychopathy in the relation between trauma and aggression (hypothesis 4) was assessed via mediation and moderation analyses. Baron and Kenny’s (1986) steps were used for the mediation analyses. For the moderation analysis, interaction variables for trauma and psychopathy were created. Both the mediation and moderation analyses were conducted using Multiple Regression models.
RESULTS
This study conducted analyses with and without age as a covariate variable. Age proved to be a significant negative predictor for aggression, with total aggression (Beta = .37, p < .01), reactive aggression (Beta = -.35, p < .01) and proactive
aggression (Beta = -.35, p < .01). This effect was taken into account by including age in the following analyses.
Trauma and aggression
A Pearson partial correlation was conducted to examine a possible connection between trauma and the two types of aggression. Age was include as a covariate. As expected, a significant correlation was found for trauma and reactive aggression (r = .34, p < .005) (Hypothesis 1a). However, contrary to the expectations, no significant correlation was found for trauma and proactive aggression (r = .29, p = .117)
(Hypothesis 1b).
Psychopathy and aggression
A Linear Multiple Regression analysis was conducted with total psychopathy score as the predictor and aggression as the dependent variable (Table 1). Again, age was included as a covariate. The model was significant with F(2, 66) = 21.80, p < .001. The linear multiple regression model supported hypothesis 2: total psychopathy was a significant predictor for aggression. In this sample 39,8% of the variation in aggression was explained by the predictors total psychopathy and age (R² = .40).
13
Table 1
Psychopathy total predicting aggression, with age as covariate.
Model Unstandardized Coefficients Standardized
Coefficients t Sig. B Std. Error Beta (Constant) 10,960 5,350 2,049 ,044 Psychopathy total ,602 ,113 ,511 5,335 ,000 Age -,338 ,095 -,339 -3,547 ,001
Note. Predictors: Psychopathy total; Age. Dependent Variable: Aggression total.
Hypotheses 2a, 2b and 2c examined psychopathy and aggression on a subfactor level. All psychopathy subfactors were included in the analyses to control for their shared variance. This way suppression effects are prevented, which increases the validity of each predictor (Hicks & Patrick, 2006). For hypothesis 2a the
independent variables included F1, F2, F3 and age, with proactive aggression as the dependent variable. The Multiple Linear Regression model was significant at F(4, 64) = 6.22, p < .001. Although the interpersonal (F1) and affective (F2) subfactors of psychopathy were hypothesized to predict proactive aggression, no significant results were found (as shown in Table 2). F1 and F2 did not predict proactive aggression. Worth mentioning though, is the borderline significant result of predictor F1.
Table 2
The predictive effects of F1 and F2 on proactive aggression.
Model Unstandardized Coefficients Standardized
Coefficients t Sig. B Std. Error Beta (Constant) 3,956 2,944 1,344 ,184 F1 ,292 ,149 ,233 1,961 ,054 F2 ,167 ,149 ,127 1,123 ,266 F3 ,225 ,142 ,184 1,588 ,117 Age -,150 ,054 -,307 -2,806 ,007
Note. Predictors F1 = Interpersonal psychopathy factor; F2 = Affective psychopathy factor; F3 = Lifestyle psychopathy factor; Age.
Dependent Variable: Proactive aggression.
Hypothesis 2b looked at the relation between the lifestyle (F3) factor of psychopathy and reactive aggression. The Multiple Linear Regression model was significant at F(4, 64) = 14.50, p < .001. The predictor variables included the three
14 psychopathy subfactors and age, with reactive aggression as the dependent variable. As hypothesized, Table 3 shows F3 being a significant predictor for reactive
aggression.
It was assumed for the interpersonal (F1) and affective (F2) factor to have predictive effect for reactive aggression (Hypothesis 2c). Results indeed showed no predictive effects, as displayed in Table 3.
Table 3
The predictive effect of F3 on reactive aggression.
Model Unstandardized Coefficients Standardized
Coefficients t Sig. B Std. Error Beta (Constant) 6,646 2,934 2,265 ,027 F1 ,196 ,148 ,134 1,319 ,192 F2 ,196 ,149 ,127 1,319 ,192 F3 ,688 ,141 ,481 4,868 ,000 Age -,192 ,053 -,336 -3,600 ,001
Note. Predictors: F1 = Interpersonal psychopathy factor; F2 = Affective psychopathy factor; F3 = Lifestyle psychopathy factor; Age.
Dependent Variable: Reactive aggression.
Trauma and psychopathy
This study predicted a correlation between trauma and total psychopathy (Hypothesis 3). Indeed, a Pearson partial correlation (correction for age) showed a significant positive correlation between trauma and total psychopathy (r = .25, p < .05). A correlation between the trauma and the interpersonal (F1) and affective (F2) factors was hypothesized (Hypothesis 3a). However, a Pearson partial correlation showed that no such correlations were found with F1 (r = .05, p = .715) and F2 (r = .20, p = .099). On the other hand, trauma was significantly correlated with the
lifestyle (F3) factor of psychopathy (r = .29, p < .05). This Pearson partial correlation supported hypothesis 3b.
Mediation of psychopathy on trauma and aggression
Baron and Kenny’s steps for mediation were used to check the requirements to perform a mediation analysis (Baron & Kenny, 1986). Step 1 requires the independent variable to be a significant predictor of the dependent variable. Results of hypothesis 1b in this study were not showing a significant relation between trauma (independent) and proactive aggression (dependent). The requirements were not met, therefore hypothesis 4a could not be conducted.
To test hypothesis 4b, Baron and Kenny’s steps for mediation were used once again. The requirements in step 1 were met, as shown by hypothesis 1a. The
independent variable (trauma) was a significant predictor of the dependent variable (reactive aggression). Also, the independent variable (trauma) was a significant predictor of the mediator (F3), as shown by hypothesis 3b (step 2). Step 3 requires the
15 mediator (F3) to be a significant predictor of the dependent variable (reactive
aggression), while controlling for the independent variable (trauma). To confirm step 3, a Hierarchical Multiple Regression Analysis was conducted. First, with trauma and age as the predictors and reactive aggression as the dependent variable. Second, the mediator (F3) was added to the model as a predictor. Table 4 shows F3 significantly predicting reactive aggression while controlling for trauma. This result was in line with the requirement as stated in step 3. Step 4 examines a possible partial or full mediation effect. A Sobel test was used to assess the significance of the mediation effect (Preacher & Leonardelli, 2001). Results showed a significant full mediation effect of F3 (Z = 2.68, p < .01) on trauma and reactive aggression (Hypothesis 4b). Moreover, the relation between trauma and reactive aggression stopped being significant when F3 was added. The explained variation in reactive aggression increased from 23,5% (Block 1: R² = .26) to 47,0% (Block 2: R² = .47) when the mediator variable (F3) was added.
Table 4
Multiple Regression Analysis in which F3 predicts the relation between trauma and reactive aggression.
Model Unstandardized Coefficients Standardized Coefficients t Sig. B Std. Error Beta 1 (Constant) 16,288 2,648 6,150 ,000 Trauma total ,102 ,033 ,340 3,086 ,003 Age -,243 ,063 -,426 -3,859 ,000 2 (Constant) 8,718 2,629 3,316 ,001 Trauma total ,056 ,029 ,188 1,939 ,057 Age -,227 ,053 -,397 -4,286 ,000 F3 ,727 ,135 ,508 5,376 ,000
Note. Model 1 predictors: Trauma total; Age.
Model 2 predictors: Trauma total; Age; F3 = Lifestyle psychopathy factor. Dependent Variable: Reactive aggression.
Moderation of psychopathy on trauma and aggression
A possible moderation effect of F1/F2 on the relation between trauma and reactive aggression was hypothesized. In other words, this study examined if the relation between trauma and reactive aggression was stronger for participants with low scores on F1/F2. To test this hypothesis (4c), a Multiple Regression Analysis was conducted with reactive aggression as the dependent variable and trauma, F1, F2, trauma × F1 and trauma × F2 as the predictor variables. The interaction variables have been centred around the mean to prevent collinearity. The model was significant at F(6, 62) = 9.44, p < .001. The results were in contrast with the expectations (as
16 shown in Table 5). Neither trauma × F1, nor trauma × F2 were significant predictors of reactive aggression.
Table 5
Multiple Regression Analysis in which F1, F2, trauma and their interaction terms predict reactive aggression.
Model Unstandardized Coefficients Standardized
Coefficients t Sig. B Std. Error Beta (Constant) 6.277 2.910 2.157 .035 Trauma total .063 .031 .210 2.006 .049 F1 .212 .146 .145 1.451 .152 F2 .153 .147 .099 1.039 .303 F3 .622 .148 .434 4.190 .000 Trauma × F1 (moderator) -.041 .540 -.008 -.076 .939 Trauma × F2 (moderator) -.696 .527 -.128 -1.322 .191 Age -.222 .054 -.389 -4.077 .000
Note. Predictors: F1 = Interpersonal psychopathy factor; F2 = Affective psychopathy factor; F3 = Lifestyle psychopathy factor; Trauma × F1 (moderator); Trauma × F2 (moderator); Age.
Dependent Variable: Reactive aggression
Exploratory analyses: Trauma subtypes
Additional analyses were added to his study by differentiating trauma subtypes. This way relations of the trauma construct could be examined in greater depth and may possibly provide etiological information. A Pearson partial correlation correcting for age, showed physical abuse to be related to trauma (r = .82, p < .01). Also, emotional abuse was related to trauma (r = .88, p < .01). The two types of abuse had no significant correlation with proactive aggression as shown by Table 6.
However they were significantly correlated with reactive aggression. These two types of trauma correlated with trauma in total, reactive aggression and psychopathy (as mentioned hereafter) and were included for further analyses.
Other noteworthy results include significant correlations of physical neglect with total, reactive and proactive aggression. Emotional neglect showed a significant correlation only with total aggression. Interestingly, sexual abuse was not related to aggression or psychopathy at all.
17 A subsequent Pearson partial correlation correcting for age, was conducted to examine possible relations between physical abuse, emotional abuse and psychopathy. Both types of trauma were significantly correlated to psychopathy, as shown in Table 7. In addition, results showed a significant correlation for physical and emotional abuse with F3. The only result in which physical abuse differed from emotional abuse, was a significantly correlation between physical abuse and F2.
Table 6
Pearson partial correlation (correcting for age) between trauma subtypes and aggression.
Corr. P.A. P.N. E.A. E.N. S.A. Tot.
Agg. Re. Agg. Pro. Agg. P.A. - P.N .505*** - E.A. .684*** .574*** - E.N .388** .657*** .558*** - S.A. .438*** .185 .339** .122 - Tot Agg. .220 .298* .297* .246* .011 - Re. Agg. .288* .283* .397** .226 .077 .940*** - Pro. Ag. .108 .270* .135 .231 -.068 .917*** .725*** - Note. P.A. = physical abuse; P.N. = physical neglect; E.A. = emotional abuse; E.N. = emotional neglect; S.A. = sexual abuse; Tot. Agg. = total aggression; Re. Agg. = reactive aggression; Pro. Agg. = proactive aggression.
* = p < .05 ** = p < .01 *** = p < .001
Table 7
Pearson partial correlation (correcting for age) between trauma subtypes and psychopathy.
Corr. P.A. P.N. E.A. E.N. S.A. Tot.
Psy. F1 F2 F3 P.A. - P.N .505*** - E.A. .684*** .574*** - E.N .388** .657*** .558*** - S.A. .438*** .185 .339** .122 - Tot. Psy. .266* .128 .340** .156 -.116 - F1 .054 -.033 .149 .038 -.166 .746*** - F2 .243* .101 .203 .149 -.034 .694*** .282* - F3 .284* .210 .387** .156 -.055 .754*** .358** .270* -
18 Note. P.A. = physical abuse; P.N. = physical neglect; E.A. = emotional abuse; E.N. = emotional neglect; S.A. = sexual abuse; Tot. Psy. = total psychopathy; F1 =
interpersonal psychopathy factor; F2 = affective psychopathy factor; F3 = lifestyle psychopathy factor.
* = p < .05 ** = p < .01 *** = p < .001
Trauma subtypes mediation
Baron and Kenny’s steps were followed to test for possible mediation effects of F3 on the relation between physical/emotional abuse and reactive aggression (Baron & Kenny, 1986). All steps were completed for both types of abuse. A Multiple Linear Regression model (F(3, 65) = 17.91, p < .001) (as shown in Table 8) and a Sobel test were conducted for physical abuse, which provided similar results (Sobel Z = 2.24, p < .05). Adding F3 to the model caused physical abuse to be nonsignificant and increased the explained variation by 22.9%. F3 functioned as a mediator (full) in the relation between physical abuse and reactive aggression.
Following, a Multiple Linear Regression model with age as covariate (F(3, 65) = 19.02, p < .001) (as shown in Table 9) and a Sobel test, showed F3 fully mediating the relation between emotional abuse and reactive aggression (Sobel Z = 2.89 , p < .005). Emotional abuse stopped being significantly related to reactive aggression when F3 was added to the model and lead to an increase in explained variation of 17.2%.
Table 8
Multiple Regression Analysis in which F3 mediates the relation between physical abuse and reactive aggression.
Model Unstandardized Coefficients Standardized
Coefficients t Sig. B Std. Error Beta 1 (Constant) 17.568 2.624 6.695 .000 Physical Abuse .269 .110 .272 2.442 .017 Age -.224 .064 -.392 -3.518 .001 2 (Constant) 9.242 2.655 3.481 .001 Physical Abuse .119 .096 .121 1.248 .217 Age -.214 .053 -.374 -4.034 .000 F3 .756 .137 .528 5.512 .000
Note. Model 1 predictors: Physical abuse; Age.
Model 2 predictors: Physical abuse; Age; F3 = Lifestyle psychopathy factor. Dependent Variable: Reactive aggression.
19
Table 9
Multiple Regression Analysis in which F3 mediates the relation between emotional abuse and reactive aggression.
Model Unstandardized Coefficients Standardized
Coefficients t Sig. B Std. Error Beta 1 (Constant) 16.956 2.517 6.736 .000 Emotional Abuse .354 .101 .376 3.517 .001 Age -.234 .061 -.409 -3.821 .000 2 (Constant) 9.507 2.619 3.629 .001 Emotional Abuse .173 .093 .184 1.849 .069 Age -.220 .052 -.384 -4.187 .000 F3 .704 .141 .491 5.003 .000
Note. Model 1 predictors: Emotional abuse; Age.
Model 2 predictors: Emotional abuse; Age; F3 = Lifestyle psychopathy factor. Dependent Variable: Reactive aggression.
Trauma subtypes moderation
Final, a moderation effect of F1/F2 on the relation between physical/emotional abuse and reactive aggression was examined. Multiple Linear Regression models were used to test for a possible effect. Age was included as a covariate variable. The models were significant for emotional abuse and the interpersonal subfactor (F(4, 64) = 8.49, p < .001) and for emotional abuse and the affective subfactor (F(4, 64) = 8.11, p < .01). However, the moderators emotional abuse × F1 (Beta = -.02, p = .84) and emotional abuse × F2 (Beta = -.16, p = .13) were not significant.
The models for physical abuse and the interpersonal subfactor (F(4, 64) = 7.05, p < .01) and physical abuse and the affective subfactor (F(4, 64) = 5.78, p < .01) were significant. No significant moderation effects were found for physical abuse × F1 (Beta = -.07, p = .53) or physical abuse × F2 (Beta = -.12, p = .28)
Results of the Multiple Linear Regression models show no significant moderation effects. F1/F2 have no protective effect on the relation between physical or emotional abuse and reactive aggression.
20
DISCUSSION
This study investigated the complex relations between trauma, aggression, and psychopathy, in order to generate further understanding on their interplay. The
relations of trauma with aggression and psychopathy were examined as well as
predictive effects of psychopathy on aggression. Overarching, the role of psychopathy in the relation between trauma and aggression was investigated.
Trauma and aggression
The first examined relation concerned trauma and aggression. This study found a correlation between trauma and reactive aggression. As mentioned, trauma is linked to numerous adverse outcomes (disorders) that may lead to displaying reactive aggression (Oosterlaan, Geurts, Knol, & Sergeant, 2005). Findings in this study give support for a relationship between trauma and reactive aggression which many other studies also found (Ford, Chapman, Connor, & Cruise, 2012). As mentioned,
postdictive analyses were added to this study to differentiate trauma subtypes. The overall relation between trauma and reactive aggression seems to be especially explained by emotional and physical abuse, but also physical neglect. Presumably, experiencing trauma may cause an increased risk for displaying reactive aggression via various arousal related problems. More specifically, maladaptive- and
hyperarousal show strong links with trauma and reactive aggression (Ford, Chapman, Connor, & Cruise, 2012). Another explanation is provided by Trickett and McBride-Chang (1995), who conclude that (physical) abuse in childhood may lead to
interpretation biases in social situations. Social cues are more frequently considered hostile in physically abused children, leading to a lower threshold for displaying reactive aggression (Crick & Dodge, 1996).
An overall link between trauma and proactive aggression was not found however. As Dodge, Lochman, Harnish, Bates and Petit (1997) demonstrated, histories of abuse and early onset problems were only found in reactive aggressive groups and not in proactive aggressive groups. They found that the proactive aggressive group distinguished themselves by anticipating positive outcomes for displaying aggression. They argue that reactive aggression has its origin in factors such as early trauma’s, as found in the current study, and adverse family conditions. Proactive aggression however, has its origin in social learning during school years (Dodge, et al., 1997). Proactive aggression is linked to the belief that aggression can be used to accomplish personal goals and desires (Hubbard, McAuliffe, Morrow, & Romano, 2010).
Although the results in the current study regarding trauma as a whole are in line with this conception, physical neglect did in fact have a significant correlation with proactive aggression. This finding is remarkable since neglect is often linked to internalizing problems, whereas abuse is linked to more externalizing problems such as aggression (Hildyard & Wolfe, 2002). Low self-esteem, social withdrawal, and severe cognitive and academic deficits have been linked to neglect (Hildyard & Wolfe, 2002). It may be that proactive aggression is learned to be convenient to achieve personal goals and desires, when cognitive and academic skills are lacking due to effects of physical neglect. Although no hard statements can be made, this notion would fit the social learning theory of Dodge et al. (1997). Noteworthy, literature shows unclear frameworks and insufficient discrimination of physical neglect, complicating theory development regarding a link between physical neglect and proactive aggression. While physical neglect stands out in the current study, the
21 overall results give no support for previous understandings that signs of blunting or acquired callousness by trauma may lead to a proactive kind of aggression. Proactive aggression could have its basis in other factors, a genetic predisposition or social learning for example.
A worth mentioning result regarding aggression, was that older participants were less inclined to report aggression (proactive as well as reactive) in comparison with younger participants. Most likely, engaging in aggressive behaviour declines when one gets older.
Psychopathy and aggression
Second, relationships between psychopathy and aggression have been
explored. This study showed psychopathy to be a significant predictor for aggression, an outcome that has been widely supported (Porter & Woodworth, 2006; Cima et al, 2008.
When viewed on a subfactor level, the lifestyle factor of psychopathy
significantly predicted reactive aggression, matching previous results (Cornell et al., 1996; Woodworth & Porter, 2002; Reidy et al., 2011). Both concepts share
characteristics with one another (e.g. impulsivity), which could explain this outcome. Reactive aggression is often linked to rage and outbursts of anger, which are strongly related to social deviance and fit the lifestyle factor of psychopathy (Patrick, 1994).
As hypothesized, the interpersonal and affective subfactor did not predict reactive aggression, paralleling several previous studies (Cornell et al., 1996; Woodworth & Porter, 2002; Reidy et al., 2011). Reidy et al. (2011) even found a negative interaction with reactive aggression. An informative insight was replicated in this study: the core features of psychopathy (the interpersonal and affective factor) were not associated with the most common form of aggression (reactive). Traits like glibness, superficial charm, callousness and shallow affect could be seen as non-coexisting with the hot-tempered and easily provoked features of reactive aggression (Reidy et al., 2011). A calm, callous and calculating mind can be considered at odds with the short tempered features of reactive aggression.
At the same time, F1 (interpersonal) and F2 (affective) had no significant predictive effect on proactive aggression as well. The interpersonal style fell just short of a significant result in this study. This finding is in contrast with this study’s
hypothesis and a majority of research. Perhaps the interpersonal and affective styles manifest themselves in different, non-aggressive ways, such as exploitation, fraud, or property crime (Flores-Mendoza, Alvarenga, Herrero, & Abad, 2008). A
methodological explanation could also be presented. People with high psychopathic traits are often prone to lying and engaging in deceitful behaviour (Hare, Hart, & Harper, 1991). While the YPI items are worded in such a way that people with psychopathic traits would see them as positive, the formulation of the RPQ questions is described in a more negative way. Perhaps participants with psychopathic traits were restrained to accurately report aggressive behaviours. It has to be said that deceit, self-serving misrepresentations and social desirability could have influenced the findings in this study.
Trauma and psychopathy
Third, the current study found associations between trauma and psychopathy. This finding is in line with the notion of Craparo, Schimmenti and Caretti (2013). They described that early exposure to traumatic experiences can play a relevant role in the development of more severe psychopathic traits. Also, Weiler and Widom (1996,
22 as cited in Poythress, Skeem, & Lilienfeld, 2006) found that abused and neglected people had significantly higher psychopathy scores, while controlling for
demographic features. Like many other studies, they found a link between trauma and psychopathy. The current study adds some important insights to the current body of knowledge by distinguishing trauma per psychopathy subfactor.
First of all, this study found a significant relation between trauma and the lifestyle factor (F3) of psychopathy. The lifestyle factor seems largely responsible for the relation between trauma and the total psychopathy construct. Furthermore, a relation between abuse (both physical and emotional) and the lifestyle factor of psychopathy (F3) was found, suggesting that the link between trauma and F3 is due to traumatic abuse.
In addition, physical abuse showed a relation with the affective factor (F2). Perhaps emotional detachment could serve as a coping method to endure physical abuse, which can lead to traits resembling the affective factor (Hawes, & Dadds, 2007; Ford et al., 2012; Kerig et al., 2012). As for the interpersonal factor (F1), no links to histories of abuse or neglect were found in this study. It is noteworthy that characteristics of F1 could interfere with a true representation of traumatic
experiences. For example, a ‘grandiose sense of self-worth’ could have prevented the disclosing of negative and/or shameful experiences and ‘Pathological lying’ may have played a role in honestly answering the questionnaires
Taking this into consideration, a more interesting insight can be given. The interpersonal and, to a lesser extent, the affective factors, are perhaps less affected by environment than the lifestyle factor (Valentine, 2001; Poythress, Skeem, &
Lilienfeld, 2006; Cima, et al., 2008; Daversa, 2010; Craparo et al., 2013). It is
believed that a genetic origin contributes to dysfunctions in emotion which is the core of the psychopathy construct (Blair, Peschardt, Budhani, Mitchell, & Pine, 2006). Results in this study give mixed support for different perspectives on the etiology of psychopathy. The notion of Krischer and Sevecke (2008) where traumatic experiences relate to deficits in certain emotional traits as described in the affective subfactor of psychopathy, is supported by a relation between physical abuse and F2. However, no links for trauma subtypes and F1 were found. This resembles the results of Poythress, Skeem and Lilienfeld (2006) who showed that abuse was unrelated to interpersonal traits, but had strong relations with the impulsive and irresponsible traits of
psychopathy. Also, a noteworthy result involved sexual abuse not relating to
psychopathy or aggression, a result often found by a majority of research (Poythress, Skeem and Lilienfeld, 2006). An explanation could be that sexual abuse is associated with less aggressive psychiatric disorders such as mood or anxiety problems (Trickett & McBride-Chang, 1995; Chen et al., 2010). In addition, disclosure regarding sexual abuse may have been extremely difficult, certainly at the very start of treatment.
Concluding thoughts on trauma and psychopathy include a relevant role of traumatic experiences in the development of psychopathy lifestyle traits. Interpersonal traits however, do need seem to be influenced by trauma and an inherited or innate basis for F1 may be carefully presumed (Poythress, Skeem, & Lilienfeld, 2006). The origin of affective traits is much more unclear. As found in the current study, physical abuse could perhaps lead to the cold and callous characteristics of psychopathy (Krischer and Sevecke, 2008), although trauma in general showed no associations with the affective factor. Future research on trauma on psychopathy subfactors is recommended for better understanding.
23 Trauma, Psychopathy and aggression
An important goal of the current study was to examine the role of
psychopathic traits in the relation between trauma and aggression. In contrast with F1 and to a lesser degree F2, the lifestyle features of psychopathy (F3) seem to be
robustly influenced by environmental factors such as trauma. Moreover, the development of such traits in response to trauma, could explain the reactive
aggression that is associated with this disorder. This study showed a full mediation of F3 on the relation between trauma and reactive aggression. More specifically, F3 also fully mediated the relation between physical abuse / emotional abuse and reactive aggression. Trauma (physical and emotional abuse) seem to be precursors of F3 traits from which reactive aggression can derive. Of course, this relation could also be explained via a genetic basis for aggression or psychopathy. Children born with psychopathy lifestyle or aggressive features may be at a heightened risk for abuse (trauma) by their environment. However, the notion of trauma being a precursor of F3, is based on convincing results in twin studies where a causal role of physical abuse in antisocial behaviour was demonstrated (Jaffee et al., 2004). Although no causal relations can be stated, the current study thus suggests that traumatisation via abuse increases the risk for developing psychopathy lifestyle factor traits, which in turn can lead to reactive aggression.
A primary goal in this study was to examine if individuals with a history of trauma and interpersonal/affective psychopathy features are protected for displaying reactive aggression. Characteristics of these psychopathy features can be considered at odds with displaying reactive aggression despite having experienced trauma (Reidy et al., 2011). Nonetheless, this notion was not supported. A diminishing effect may have been too small to have a protective influence. Another possible explanation may be that people with high interpersonal and affective psychopathy traits tend to engage less in reactive aggression compared to people with high lifestyle psychopathy traits, but a total absence (or counteracting effect) however is not quite accurate.
An interesting conclusion that could be drawn from this study is that trauma in general relates to hyperarousal instead of hypoarousal (Kendall-Tackett, 2000;
Daversa, 2010). Results in this study show relations for trauma with the psychopathy lifestyle factor and reactive aggression: traits which are associated with hyperarousal (Raine et al., 2006). Variables like the psychopathy affective factor and proactive aggression, which are rather associated with hypoarousal (Reidy et al., 2011), had no relations with overall trauma (although trauma subtypes showed different results). The overall conclusion is in line with Woodworth and Porter (2002), who argue that psychopaths and proactive aggression are accompanied with low emotional arousal, while nonpsychopaths and reactive aggression are linked to high emotional arousal. Presumably, trauma increases the risk for hyperarousal related problems and
disorders.
Primary vs secondary psychopathy?
Results in this study may give rise to the idea of different variations of psychopathy in its heterogeneous construct, for example the primary and secondary psychopathy differentiation (Karpman, 1941, as cited in Poythress, Skeem, & Lilienfeld, 2006). Karpman’s distinction centralized around the level of neuroticism. Secondary psychopaths (described as neurotic psychopaths) would experience higher levels of anxiety and fear, which could cause them to react emotionally and
impulsively. They lack emotional and behavioural regulation, which could result in antisocial and impulsive behaviour. Hate, revenge, but also depression, are themes
24 that are often present. Primary psychopaths on the other hand, act in a more deliberate and pragmatical way and showed states of fearlessness (a term conceptualised by Lykken, 1995) and less anxiety (Karpman, 1948; Gowlett, 2014). According to Karpman (1948), primary psychopathy had a hereditary basis, while secondary psychopathy would rather be caused by environmental influence. Lykken (1995) also distinguished psychopathy by etiological differences. It is presumed for secondary psychopathy to be caused by adverse environmental factors such as early abuse or abandonment (Lykken, 1995; Porter, 1996). Several studies linked primary
psychopathy to traits of emotional detachment (resembling F2), whereas secondary psychopathy was associated with social deviance and impulsivity (resembling F3) (Levenson et al., 1995; Mealey, 1995; as described in Skeem, Johansson, Andershed Kerr, & Louden, 2007)
This study did not explore a primary vs secondary psychopathy distinction and can only speculate to what degree the results align with such a differentiation. Studies have provided sufficient support for a subfactorial validity of psychopathy.
Psychopathy subfactors are related but differentiable psychopathologies (Hare, 1991; Cooke, & Michie, 2001; Hicks, Markon, Patrick, Krueger, & Newman, 2004;
Daversa, 2010; Reidy et al., 2011; Olver, Neumann, Wong, & Hare, 2013). It is interesting that the primary vs secondary distinction mirrors the factor structure of the YPI (Kimonis, Frick, Cauffman, Goldweber, & Skeem, 2012). The interpersonal and affective subfactor show similarities with primary psychopathy, while the lifestyle subfactor shows resemblances to secondary psychopathy (Kimonis et al., 2012).
A correlation between the lifestyle factor and trauma in comparison with no correlations between the interpersonal/affective subfactors with trauma in this study, are in line with the idea of secondary psychopathy resulting from adverse
environmental factors (Lykken, 1995; Porter, 1996; Kimonis et al., 2012). Future studies could explore how primary vs secondary psychopathy fits in the associations between trauma, psychopathy, aggression and their subtypes, to help clarify
etiological issues.
Limitations
Several limitations mighty have influenced the outcomes in the current study. First of all, underreporting may have been in issue. Participants complete the
questionnaires in the presence of the researcher. Although anonymity is emphasized, social desirability cannot be precluded. The majority of Inforsa’s clients are legally obliged to go to therapy and some are reluctant to receive therapy. They might believe that concealing or withholding certain information can be beneficial for their
treatment and judicial process, though it’s explicitly stated that participation is unrelated to their treatment. Self-serving misrepresentations and social desirability (disadvantages of self-report questionnaires) may be malefactors when the results are not in line with the expectations.
Second, the psychopathy construct is still heavily debated. It’s unclear if the three factor structure fully grasps all of its components. Also, this study uses the YPI-s which can be YPI-seen aYPI-s leYPI-sYPI-s reliably and leYPI-sYPI-s thoroughly than the frequently uYPI-se
Psychopathy Checklist – Revised (PCL-R). The methodology to measure psychopathy could have its flaws.
Finally, the more ‘extreme’ and distressed clients are often unwilling to participate in this study. Therefore it is not impossible that results are somewhat more moderate, since the peaks and troughs are not included in this study.
25
Recommendations
This study supports several links between trauma, psychopathy and aggression which stress the importance of prevention and treatment of trauma. Specific
interventions could be beneficial to prevent psychopathic traits and asocial and aggressive behaviour. Individuals with a history of trauma and aggression could benefit from trauma focused treatments such as EMDR therapy, as this may tackle the root of the problem rather than its consequences. Specifying trauma subtypes would be important here, since subtypes seem to have different correlates. In addition, psychopathy lifestyle traits may be treated to decrease chances of engaging in aggressive behaviour. Future research could concentrate on trauma and psychopathy treatment efficiency in aggressive populations.
26
REFERENCES
Arntz, A. & Wessel, I. (1996). Jeugd trauma vragenlijst (Dutch version of the childhood trauma questionnaire). Maastricht: Author.
Berkowitz, L. (1998). Frustrations, appraisals, and aversively stimulated aggression. Aggressive Behavior, 14, 3-11.
Baron, R.M., & Kenny, D.A. (1986). The moderator–mediator variable distinction in social psychological research: Conceptual, strategic, and statistical
considerations. Journal of Personality and Social Psychology, 51(6), 1173-1182.
Bernstein, D.P, & Fink, L. (1998). Childhood Trauma Questionnaire: A retrospective self-report manual. San Antonio, TX: The Psychological Corporation.
Bernstein D. P., Fink L., Handelsman L., Foote J., Lovejoy M., Wenzel K., et al. (1994). Initial reliability and validity of a new retrospective measure of child-abuse and neglect. American Journal of Psychiatry, 151, 1132–1136.
Blair, R.J.R., Peschardt, K.S., Budhani, S., Mitchell, D.G.V., & Pine, D.S. (2006). The development of psychopathy. Journal of Child Psychology and Psychiatry, 47(3), 262-275.
Chen, L.P., Murad, H.M., Paras, M.L., Colbenson, K.M., Sattler, A.L., Goranson, E.N., Elamin, M.B., Seime, R.J., Shinozaki, G., Prokop, L.J., & Zirakzadeh, A. (2010). Sexual abuse and lifetime diagnosis of psychiatric disorders: Systematic review and meta-analysis. Mayo Clinic Proceedings, 85(7), 618– 629.
Cima, M., Raine, A., Meesters, C., & Popma, A. (2013). Validation of the Dutch reactive proactive questionnaire (RPQ): differential correlates of reactive and proactive aggression from childhood to adulthood. Aggressive Behavior, 39(2), 99-113.
Cima, M., Smeets, T., & Jellicic, M. (2008). Self-reported trauma, cortisol levels, and aggression in psychopathic and non-psychopathic prison inmates. Biological Psychology 78(1), 175–86.
Cooke, D. J., & Michie, C. (2001). Refining the construct of psychopathy: Towards a hierarchical model. Psychological Assessment, 13(2), 171-188.
Cornell, D. G., Warren, J., Hawk, G., Stafford, E., Oram, g., & Pine, D. (1996). Psychopathy in instrumental and reactive violent offenders. Journal of Consulting and Clinical Psychology, 64(4), 783-790.
Craparo, G., Schimmenti, A., & Carreti, V. (2013) Traumatic experiences in childhood and psychopathy: a study on a sample of violent offenders from Italy. European Journal of Psychotraumatology, 4. Accessed on 03-04-2015,
http://dx.doi.org/10.3402/ejpt.v4i0.21471.
Crick, N. R., & Dodge, K. A. (1996). Social information-processing mechanisms in reactive and proactive aggression. Child Development, 67(3), 993-1102. Dadds, M. R., Hawes, D. J., Frost, A. D. J., Vassalo, S., Bunn, P., Hunter, K., & Mer,
S. (2009). Learning to ‘talk the talk’: the relationship of psychopathic traits to deficits in empathy across childhood. Journal of Child Psychology and Psychiatry, 50(5), 599-606.
Daversa, M.T. (2010). Early Environmental Predictors of the Affective and
Interpersonal Constructs of Psychopathy. International Journal of Offender Therapy and Comperative Criminology, 54 (1), 6-21.
Dodge, K.A., Lochman, J.E., Harnish, J.D., Bates, J.E., & Petit, G.S. (1997). Reactive and proactive aggression in school children and psychiatrically impaired
27 chronically assaultive youth. Journal of Abnormal Psychology, 106 (1), 37-51.
Falkenbach, D.M. (2004). The subtypes of psychopathy and their relationship to hostile and instrumental aggression (Doctoral dissertation). University of South Florida, Florida.
Farrington, D. P. (2005). The importance of child and adolescent psychopathy. Journal of Abnormal Child Psychology, 33(4), 489-497.
Faul, F., Erdfelder, E., Lang, A.-G., & Buchner, A. (2007). G*Power 3: A flexible statistical power analysis program for the social, behavioural, and biomedical sciences. Behavior Research Methods, 39, 175-191.
Flores-Mendoza, C. E., Alvarenga, M. A. S., Herrero, O., & Abad, F. J. (2008). Factor structure and behavioural correlates of the Psychopathy Checklist-Revised [PCL-R] in a Brazilian prisoner sample. Personality and Individual
Differences, 45, 584-590.
Ford, J.D., Chapman, J., Connor, D.F., & Cruise, K. R. (2012). Complex trauma and aggression in secure juvenile justice settings. Criminal Justice and Behavior, 39(6), 694-724.
Gowlett, C.T. (2014). Does secondary psychopathy exist? Exploring
conceptualisations of psychopathy and evidence for the existence of a
secondary variant of psychopathy. Internet Journal of Criminology. Accessed on 23-12-2015, at
http://www.internetjournalofcriminology.com/Gowlett_Does_Secondary_Psyc hopa thy_Exist_IJC_Jan_2014.pdf.
Hare, R. D., Hart, S. D., & Harpur, T. J. (1991). Psychopathy and the DSM-IV criteria for antisocial personality disorder. Journal of Abnormal Psychology, 100(3), 391-398.
Hawes, D. J., Dadds, M. R. (2007). Stability and malleability of callous-unemotional traits during treatment for childhood conduct problems. Journal of Clinical Child and Adolescent Psychology, 36(3), 437-355.
Hicks, B.M., & Patrick, C. J. (2006). Psychopathy and negative emotionality:
Analyses of suppressor effects reveal distinct relations with emotional distress, fearfulness, and anger-hostility. Journal of Abnormal Psychology, 115(2), 276-287.
Hicks, B. M., Markon, K. E., Patrick, C. J., Krueger, R. F., & Newman, J. P. (2004). Identifying psychopathy subtypes on the basis of personality structure. Psychological Assessment, 16(3), 276-288.
Hildyard, K. L., & Wolfe, D. A. (2002). Child neglect: Developmental issues and outcomes. Child Abuse & Neglect, 26(6-7), 679-695.
Hubbard, J. A., McAuliffe, M. D., Morrow, M. T., & Romano, L. J. (2010). Reactive and proactive aggression in childhood and adolescence: Precursors, outcomes, processes, experiences, and measurement. Journal of Personality, 78(1), 95-118.
Jaffee, S.R., Caspi, A., Moffit, T.E., Dodge, K.A., Rutter, M., Taylor, A., & Tully, L.A. (2005). Nature × nurture: Genetic vulnerabilities interact with physical maltreatment to promote conduct problems. Developmental Psychopathology, 17(1), 67-84.
Karpman, B. (1948). The myth of the psychopathic personality. American Journal of Psychiatry, 524-534.
