Don't look back in anger: Neural correlates of reappraisal, analytical rumination, and angry rumination during recall of an anger-inducing autobiographical memory

Abstract

Despite the enormous costs associated with unrestrained anger, little is known about the neural mechanisms underlying anger regulation. Behavioral evidence supports the effectiveness of reappraisal in reducing anger, and demonstrates that rumination typically maintains or augments anger. To further understand the effects of different anger regulation strategies, during functional magnetic resonance imaging 21 healthy male and female undergraduates recalled an anger-inducing autobiographical memory. They then engaged in three counterbalanced anger regulation strategies: reappraisal, analytical rumination, and angry rumination. Reappraisal produced the least self-reported anger followed by analytical rumination and angry rumination. Rumination was associated with increased functional connectivity of the inferior frontal gyrus with the amygdala and thalamus. Understanding how neural regions interact during anger regulation has important implications for reducing anger and violence.

Introduction

When recalling an anger-inducing event, the extent to which one becomes angered depends on how one mentally processes the event (Mauss et al., 2007, Memedovic et al., 2010, Ray et al., 2008). This process of emotion regulation refers to how we experience and express our emotions. Failing to regulate frequent anger can lead to violence, social dysfunction, and poor mental and physical health (Anderson and Bushman, 2002, Friedman and Booth-Kewley, 1987, Houston and Vavak, 1991, John and Gross, 2004). In the present functional magnetic resonance imaging (fMRI) study, participants cognitively processed an anger-inducing autobiographical memory in three ways: namely, by engaging in reappraisal, analytical rumination, and angry rumination. Our goal was to understand the neural mechanisms involved in these three forms of anger regulation.

Of these three types of emotion regulation, the most effective in reducing anger is cognitive reappraisal (Denson et al., 2011a, Mauss et al., 2007, Memedovic et al., 2010, Ray et al., 2008). Cognitive reappraisal involves reinterpreting an emotional event in order to reduce its negative emotional impact (Gross, 1998, Gross, 2001). For instance, one may think about an anger-eliciting event from the perspective of a neutral third party. Behavioral studies have shown that reappraisal decreases anger and increases healthy patterns of cardiovascular responding (Denson et al., 2011a, Denson et al., in press, Mauss et al., 2007, Memedovic et al., 2010, Ray et al., 2008); however, the neural regions involved in reappraising an anger-inducing event have not been investigated. Outside of the anger context, reappraising negative affective stimuli increases activation in regions implicated in meaning processing, self-reflection, cognitive control, and reward. The most robust finding is that reappraisal activates the dorsal and/or ventral lateral prefrontal cortex (PFC) (for a review, see Ochsner and Gross, 2008). The inferior frontal gyrus (IFG) in the lateral PFC has been implicated in a variety of tasks requiring cognitive and inhibitory control including emotion regulation (Lieberman, 2007, Tabibnia et al., 2011). Other regions often activated by reappraisal include the medial PFC (mPFC), lateral and medial orbitofrontal cortex (OFC), anterior cingulate cortex (ACC), amygdala, and caudate (McRae et al., 2008, McRae et al., 2009, Ochsner and Gross, 2008, Ochsner et al., 2002, Ochsner et al., 2004).

In addition to reappraisal, we investigated two forms of rumination. Analytical rumination involves focusing on why an event occurred by analyzing the event's causes, consequences, and meaning (for a review, see Watkins, 2008). When analytical rumination is conducted from a “cool”, self-distanced perspective, it reduces anger and cardiovascular reactivity relative to “hot”, emotionally evocative, self-immersed angry rumination (Ayduk and Kross, 2008, Kross et al., 2005). Similarities have been drawn between self-distanced analytical rumination and reappraisal, such that both involve an attempt to create meaning. Furthermore, taking a distanced perspective has been conceptualized as an aspect of reappraisal (Ayduk and Kross, 2010, Ochsner and Gross, 2008). By contrast, angry rumination involves focusing on one's angry feelings and thoughts of revenge, and results in increased anger and aggression (Bushman, 2002, Caprara, 1986, Denson et al., 2006, Sukhodolsky et al., 2001). Angry rumination increases or maintains anger, aggression, blood pressure (Bushman, 2002, Bushman et al., 2005, Denson et al., 2011b, Pedersen et al., 2011) and increases activation in the insula, ACC, mPFC, and dorsal and ventral lateral PFC (Denson et al., 2009).

To our knowledge, only three studies have examined the neural correlates of anger regulation, and they have primarily done so indirectly (Alia-Klein et al., 2007, Alia-Klein et al., 2009, Denson et al., 2009). Alia-Klein et al. (2007) asked men to listen to the word no, which tends to be associated with anger, and yes, which does not tend to be associated with anger. Relative to neutral words, hearing the word yes increased activity in the lateral OFC, whereas hearing the word no decreased activity in the lateral OFC. Self-reported trait anger control was positively correlated with lateral OFC activity while listening to the word no. Additional research suggests that OFC lesions are associated with deficits in self-regulation and impulsive aggression (Blair, 2004, Grafman et al., 1996, Kringelbach and Rolls, 2004). Consistent with this notion, Mehta and Beer (2010) found that blunted activation in the medial OFC was correlated with increased reactive aggression in an economic bargaining game. The lateral and medial OFC have also been implicated in regulating general negative affect (Banks et al., 2007, Ochsner et al., 2002, Ochsner et al., 2004, Phan et al., 2005). Ochsner et al. (2004) found that using reappraisal to decrease negative affect was associated with increased lateral PFC and OFC activation. In sum, the research to date suggests a role for top-down control processes supported by the lateral PFC and OFC in anger regulation.

In this experiment, we identified regions commonly activated during reappraisal, analytical rumination, and angry rumination. We then tested two complementary hypotheses that may account for the differential effectiveness of each strategy in reducing anger. The mean level hypothesis is that mean levels of activation in the regions differ as a function of type of anger regulation. For instance, reappraisal may initiate relatively greater activation in regions implicated in cognitive control and reward (e.g., IFG, OFC, dACC, caudate); whereas angry rumination may initiate greater activation in regions implicated in negative emotions and arousal (e.g., amygdala, insula, thalamus). To test the mean level hypothesis, we conducted a conjunction analysis to identify regions that are active during all three anger regulation strategies relative to a baseline period in which participants were asked to relax. We then examined mean differences in activity during the three strategies.

A second possibility is the functional connectivity hypothesis which is that the emotion regulation strategies differ in terms of functional connections that exist between neural regions. Past research on reappraising negative affect has implicated downregulation of subcortical regions by the PFC (Banks et al., 2007, Heatherton, 2011, Ochsner et al., 2002, Urry et al., 2006). In contrast to reappraisal, rumination may involve upregulation of subcortical limbic activation by the PFC. For instance, thinking about the anger experienced and planning revenge may be supported by positive connectivity between cortical and subcortical regions. Such a reciprocal feedback loop in which participants become increasingly “worked up” could explain how rumination (or at least angry rumination) increases arousal and anger. To test the functional connectivity hypothesis, we performed psychophysiological interaction (PPI) analyses examining activity during each emotion regulation strategy. We expected that regions associated with top-down prefrontal control would modulate activity in subcortical regions.