High levels of psychopathic traits increase the risk of transferring reactive aggression to innocent people after provocation: Evidence from an ERP study

Highlights

• Participants with different levels of psychopathic traits completed a modified TAP.

• Psychopathic traits enhanced aggressive behavior to the innocents.

• A more negative N2 and a smaller P3 reflected more aggression to the innocents.

Abstract

This study investigated the neuropsychological underpinnings of reactive aggression toward innocent people in a student population with different levels of psychopathic traits. While recording event-related potentials, participants (divided into high/low psychopathic [HP/LP] traits groups) competed against two fictitious opponents in a modified Taylor Aggression Paradigm. We found that the HP group compared to the LP group selected more often high-intensity punishment for the second innocent opponent after being provoked by the first opponent. Further, a more negative N2 and a smaller P3 was found in the HP group while punishing the innocents—reflecting a tendency on antisocial-aggressive behavior. Finally, both groups showed a more negative FRN for losing than winning trials when seeing the outcome of the game. Our results suggest that high psychopathic traits increase the risk of transferring provoked aggression to innocent people—offering a psychophysiological perspective for explaining and predicting aggression against the innocents in social interactions.

Introduction

Psychopathy is an intriguing concept for both researchers and laypeople. At the same time, however, it is one of the most disturbing and hardly treatable mental disorders in the world (Vien, 2006). Psychopathy is associated with a collection of affective, interpersonal, and behavioral features, including shallow affect, lack of remorse or guilt, callousness, grandiosity, pathological lying, manipulation, irresponsibility, impulsivity, poor self-control, and sensation seeking (Cleckley, 1951; Hare, 1980; Harpur, Hare, & Hakstian, 1989). About 15∼20% of the forensic population and about 1% of the general population (Glenn, Kurzban, & Raine, 2011; Hare, 1991) meets the diagnostic criteria for psychopathy. However, it has been recognized that psychopathic features also exist as personality traits (Edens, Marcus, Lilienfeld, & Poythress, 2006; Glenn et al., 2011; Guay, Ruscio, Knight, & Hare, 2007). High levels of psychopathic traits can be readily found in normal-functioning non-forensic samples without fulfilling the criteria of psychopathy (Benning, Patrick, Hicks, Blonigen, & Krueger, 2003; Ross, Benning, Patrick, Thompson, & Thurston, 2009).

Psychopathic traits are one of the most widely recognized risk factors for aggressive behavior (Guerra & White, 2017). Aggression can be divided into proactive aggression (PA; or instrumental aggression) and reactive aggression (RA; or impulsive aggression) based on the functions of behavior (Poulin & Boivin, 2000; Preston, Anestis, Watts, & Lilienfeld, 2018). Proactive aggression has been described as a goal-directed, forethoughtful and organized act without previous provocation. This form of aggressive behavior functions as a means to gain resources or attains domination and intimidation (Berkowitz, 1983; Dodge & Coie, 1987). In contrast, reactive aggression refers to an impulsive and emotion-driven act that involves high affective-physiological arousal. This form of aggressive behavior occurs as a hostile response to interpersonal provocation or threat (Chase, O’Leary, & Heyman, 2001; Dodge & Coie, 1987; Poulin & Boivin, 2000).

Extensive research exists on the relationship between psychopathy and proactive aggression—yielding a consistent and robust finding across forensic and non-forensic populations. The psychopathic measures from either clinical diagnostic tools or self-report personality scales are positively related to proactive aggression (Cima & Raine, 2009; Reidy, Zeichner, Miller, & Martinez, 2007; Walsh, Swogger, & Kosson, 2009). However, less consistent findings are available on the association between psychopathic traits and reactive aggression across different population samples. Non-forensic individuals with higher psychopathic scores showed significantly greater reactive aggression in a laboratory aggression paradigm (Lotze, Veit, Anders, & Birbaumer, 2007; Reidy et al., 2007; Reidy, Zeichner, & Martinez, 2008; Reidy, Zeichner, & Seibert, 2011). On the contrary, forensic individuals with higher psychopathic scores showed significantly less reactive aggression using the same experimental paradigm (Veit et al., 2010). Other research on both youth and adult offenders demonstrated that psychopathy is more associated with proactive aggression than reactive aggression (Bezdjian, Tuvblad, Raine, & Baker, 2011; Cima & Raine, 2009; Lehmann & Ittel, 2012). However, a recent meta-analysis study revealed that an increased level of an individual’s psychopathic traits is associated with an enhanced propensity of both proactive aggression and reactive aggression—although the effect size is significantly larger for self-report scales in general population than that for clinical assessments in forensic population (Blais, Solodukhin, & Forth, 2014). Therefore, whether higher psychopathic traits facilitate reactive aggression remains inconclusive. Further research is needed to clarify the relationship between psychopathic traits and the risk of reactive aggression.

To date, reactive aggression has been widely studied in controlled laboratory settings using reliable and well-validated behavioral approaches (Krämer, Jansma, Tempelmann, & Munte, 2007; Krämer, Riba, Richter, & Münte, 2011; Lotze et al., 2007). The Taylor Aggression Paradigm (TAP) is one of the most frequently used measures of reactive aggression (Taylor, 1967). One recognized version of the TAP involves participants competing against a fictitious opponent in a reaction-time game in which they are asked to respond as fast as possible to win—whereby the outcome of the game is predetermined by the experimenter (Krämer, Büttner, Roth, & Münte, 2008). For each single trial of the game, one player has to select the intensity of a prospective punishment for the other player. A high-intensity punishment selected by the fictitious opponent is regarded as a provocation; it can elicit reactive aggression in participants—administering a similar high-level punishment against the opponent. Different types of aversive stimuli are used as effective punishment, including electric shock, unpleasant noise, and pneumatic pressure (Krämer et al., 2007, 2008; Lotze et al., 2007; Taylor, 1967; Veit et al., 2010).

A recent study further split the TAP into a passive and an active phase. For the passive phase, participants can only receive punishment when losing but without an opportunity to punish the opponent when winning, whereas in the active phase participants can punish the opponent when winning but without receiving any punishment even when losing (Beyens, Yu, Han, Zhang, & Zhou, 2015). This modification enhances the effectiveness of provocation and highly motivates subsequent reactive aggression since participants continuously experience high-level punishment without being able for immediate revenge.

A plethora of TAP studies have consistently demonstrated that people adopt a tit-for-tat strategy and exhibit significantly more reactive aggression after being provoked (Beyer, Munte, Gottlich, & Kramer, 2015; Brunnlieb, Munte, Kramer, Tempelmann, & Heldmann, 2013; Krämer et al., 2007, 2011). Previous studies allowed participants to retaliate against provokers, which mirrored most revenge situations occurring in real daily lives (Beyens et al., 2015; Chester & DeWall, 2016; Ping et al., 2018). However, as known from some other types of violence—including school shootings—some offenders, when cornered, might take revenge on innocent people to vent their frustration caused by someone else. Although a few studies have investigated such a phenomenon in a controlled laboratory setting (Reijntjes, Kamphuis, Thomaes, Bushman, & Telch, 2013; Reijntjes, Thomaes et al., 2013), it remains unknown what kind of individuals are more inclined to retaliate against innocent people after provocation. Considering that psychopathic traits are closely associated with aggressive behavior (Lotze et al., 2007; Reidy et al., 2007, 2008; Reidy et al., 2011), individual differences of psychopathic traits might have an impact on transferring reactive aggression to innocent people. However, as far as we know, few studies have investigated this phenomenon. Hence, a need exists in identifying the behavioral and especially the neuropsychological signatures for antisocial-aggressive behavior toward the innocents by people with different levels of psychopathic traits.

Measuring different event-related potentials (ERPs) components while performing the TAP allows investigating the neuropsychological underpinnings of reactive aggression. First, the N2 component—a negative deflection distributed around 200∼300 ms post-onset—has been frequently associated with reactive aggression as measured with the TAP (Beyens et al., 2015; Krämer et al., 2008; Nieuwenhuis, Slagter, von Geusau, Heslenfeld, & Holroyd, 2005). For example, N2 amplitudes induced by a high relative to a low provocation are only present in individuals with higher aggressive traits (Krämer et al., 2008).

Second, the feedback-related negativity (FRN) component, occurring at about 250 ms post-stimulus, is usually observed during the TAP outcome stage (Beyens et al., 2015; Krämer et al., 2008; Ping et al., 2018; Wiswede et al., 2011). The FRN is being known as an index of a feedback valence evaluation (Müller, Möller, Rodriguez-Fornells, & Münte, 2005; Nieuwenhuis et al., 2005). A more negative FRN is expected when people experience or observe a negative outcome (e.g., monetary loss, game loss) (Beyens et al., 2015; Fukushima & Hiraki, 2006; Heldmann, Russeler, & Munte, 2008; Ping et al., 2018; Yeung, Holroyd, & Cohen, 2005). Further, a more negative FRN elicited by losing a trial is only found in individuals with high compared to low aggressive behavior (Krämer et al., 2008).

Third, the P3 component is a positive deflection observed at about 300∼600 ms post-stimulus onset (Kim & Jung, 2014) and is associated with attention-orienting processes, memory operations, and inhibitory control (Friedman, Cycowicz, & Gaeta, 2001; Polich, 2007). A decreased P3 component is closely associated with antisocial and aggressive behavior (Bingren, Chanchan, Qisha, Guorong, & Wei, 2016; Gao & Raine, 2009; Pasion, Fernandes, Pereira, & Barbosa, 2018; Verona & Bresin, 2015). Moreover, an atypical P3 component is also observed for individuals with higher psychopathic traits in forensic and non-forensic samples (Anderson & Stanford, 2012; Carlson, Thai, & McLarnon, 2009; Drislane, Vaidyanathan, & Patrick, 2013; Kim & Jung, 2014).

In this study, we combined a modified TAP with ERP recordings to examine how individual differences of psychopathic traits are related to the behavior of transferring reactive aggression to innocent people after strong provocation. We also asked the same participants to complete a subsequent behavioral experiment with a low provocation to examine whether individuals with higher psychopathic traits would show more aggression toward unrelated people regardless of the level of provocation. Participants from an undergraduate student population were divided into a high psychopathic (HP) traits group and a low psychopathic (LP) traits group based on the Levenson Self-Report Psychopathy Scale (LSRP) (Levenson, Kiehl, & Fitzpatrick, 1995). They were asked to compete with two fictitious opponents in a reaction-time game during two phases. For the passive phase, participants were provoked by the first opponent who selected high-intensity punishment most of the time and were punished after losing a trial—but were unable to punish the opponent even after winning a trial. During the active phase, they played against the second opponent, who was not responsible for the provocation in the passive phase.

At the behavioral level, we hypothesized that the HP group (compared to the LP group) selects more often a high-intensive punishment for innocent opponents in the active phase only after being strongly provoked in the passive phase. At the neural level, we predicted that the HP group (compared to the LP group) displays a larger N2 amplitude and a smaller P3 amplitude—reflecting the level of aggressiveness and impulsive-antisocial behavior—while deciding the punishment intensity for the innocent opponent. Further, we expected a more negative FRN for losing than for winning trials in the HP group (compared to the LP group) after receiving feedback about the outcome of the trial—reflecting the negative valence of being not able to punish after losing the trial.