Predicting anger in social anxiety: The mediating role of rumination

Abstract

Anger and the way that anger is expressed have been linked to attrition and poorer treatment outcomes in patients with Social Anxiety Disorder (e.g., Erwin, B. A., Heimberg, R. G., Schneier, F. R. & Liebowitz, M. R. (2003). Anger experience and expression in social anxiety disorder: Pretreatment profile and predictors of attrition and response to cognitive-behavioral treatment. Behavior Therapy, 34, 331–350). Understanding the connection between social anxiety and anger may be one way to improve outcomes in this population. A cross-sectional regression design was used in a sample of 363 undergraduates to examine the suggestion that ruminative thought is a critical factor linking social anxiety to anger. In support of this hypothesis, brooding fully mediated the relationship between social anxiety and trait anger and partially mediated the relationship between social anxiety and outward anger expression. The relationship between brooding and anger suppression became non-significant after depression was controlled. In contrast, reflective pondering partially mediated the relationship between social anxiety and anger suppression. These results suggest that addressing rumination may be useful in the treatment of socially anxious patients who struggle with anger. They also support the utility of considering multiple forms of rumination and multiple anger outcomes in a single study.

Introduction

Social anxiety is a common and complex problem that affects people's interpersonal, educational, and occupational lives. While fear has long been recognized as a key characteristic of social anxiety, social anxiety has also been linked to higher levels of trait anger and maladaptive anger expression (Erwin, Heimberg, Schneier, & Liebowitz, 2003). Erwin and her colleagues found that trait anger and the tendency to suppress anger (i.e., anger-in) reduced treatment completion and response in patients with Social Anxiety Disorder. The connections between anger, maladaptive anger expression, and treatment outcomes point to the need to better understand the association between social anxiety and anger. DiGiuseppe (2007) proposed that rumination is the link between these two variables. The present study evaluates this proposition by assessing whether rumination mediates the association between social anxiety and anger and, further, whether variations in the characteristics of ruminative thought influence this relationship.

While various writers have proposed somewhat differing definitions and measures of rumination, there is general agreement that rumination is a term used to describe repetitive, aversive, and uncontrollable thought (Segerstrom, Stanton, Alden, & Shortridge, 2003). Moreover, research suggests that measures of this construct generally correlate positively with one another (e.g., Segerstrom, Tsao, Alden, & Craske, 2000). Rumination is found in a number of clinical disorders and is generally linked to poor health and mental health outcomes (e.g., Segerstrom et al., 2003).

Connections between social anxiety, negative repetitive thought, and anger are consistent with DiGiuseppe's (2007) proposed mediating role of rumination. As mentioned above, social anxiety has been linked to higher levels of trait anger and maladaptive anger expression (Erwin et al., 2003). In addition, socially anxious people are more likely to engage in post-event rumination (i.e., post-event processing) and are less likely to distract following anxiety-provoking social events relative to non-anxious controls (Kocovski et al., 2005, Kocovski and Rector, 2007, Mellings and Alden, 2000). Social anxiety has also been linked to the tendency to brood, a resentful and self-critical form of rumination (Joormann, Dkane, & Gotlib, 2006).

Rumination, in turn, has been linked to anger outcomes. Ruminating on anger-provoking events has been shown to increase anger (Rusting & Nolen-Hoeksema, 1998), and individuals who use rumination as an emotion regulation strategy are more likely to experience anger (Martin & Dahlen, 2005). Anger may be particularly likely if thoughts focus on themes of revenge, the justness of events (i.e., they should not have happened), or the justness of others' actions (i.e., others are unfair and blameworthy; DiGiuseppe & Froh, 2002). These effects appear to be specific to rumination and do not occur when anger-provoking events are simply brought to mind.¹

Rumination also influences anger expression. Participants asked to ruminate about an anger-provoking incident reacted more aggressively than participants who did not ruminate (Bushman, 2002). Conversely, rumination has also been associated with the tendency to suppress anger (Martin & Dahlen, 2005), suggesting a complex relationship between rumination and anger that requires elucidation. Important for our purposes, social anxiety is associated with the tendency to use rumination as an anger regulation strategy, an approach that the studies discussed above suggest is ineffective and may even intensify feelings of anger (Weber, Wiedig, Freyer, & Gralher, 2004). While these findings are consistent with DiGiuseppe's (2007) proposal that rumination may mediate the link between social anxiety and anger, research is needed to directly evaluate a mediation model.

Segerstrom et al. (2003) found that repetitive thought styles vary along several dimensions, most notably valence (positive versus negative content) and purpose (searching versus solving). These researchers suggested that variations in the valence and purpose of negative repetitive thought may affect the outcome of rumination. While there is growing recognition of the multifaceted nature of rumination, few studies have examined how these facets affect the experience and expression of emotion.

The Ruminative Responses Scale (RRS; Nolen-Hoeksema & Morrow, 1991), a widely used measure of rumination, has been shown to comprise two subscales, brooding (i.e., the tendency to respond to negative affect with resentful and self-critical thinking) and reflective pondering (i.e., the tendency to withdraw and analyze negative situations and mood states; Treynor, Gonzalez, & Nolen-Hoeksema, 2003). These subscales differ on both the valence and searching dimensions identified by Segerstrom et al. (2003). Brooding maps onto the negative pole of thought content, whereas pondering appears to reflect a search for meaning and includes an element of withdrawal (e.g., ‘Go away by yourself and think about why you feel this way’) not found in the brooding items. These subscales therefore offer an opportunity to examine how variations in the content and purpose of ruminative thought affect anger outcomes.

Just as rumination is multifaceted, the experience and expression of anger have been explored from a variety of different perspectives. Trait anger reflects individual differences in the tendency to experience anger in a variety of situations (Allan & Gilbert, 2002) and has been linked to long-term maladaptive personality patterns (e.g., paranoia and hostility; Conger, Conger, Edmondson, Tescher, & Smolin, 2003). Variations in anger expression have also been described. Anger-out refers to how often an individual expresses angry feelings through verbally or physically aggressive behavior and is associated with different pathological outcomes relative to anger-in (e.g., hostility and somatization; Conger et al., 2003). Anger-in, on the other hand, refers to how often an individual experiences but holds in (i.e., suppresses) angry feelings and is associated with a variety of internalizing symptoms (e.g., anxiety, depression, interpersonal sensitivity; Conger et al., 2003). These different forms of anger expression may affect individuals and their social relationships in different ways, suggesting that it is important to consider multiple anger outcomes.

Depression has been linked to both rumination and anger (e.g., Erwin et al., 2003, Kashdan and Roberts, 2007), and some writers suggest that socially anxious individuals may be particularly likely to engage in negative post-event rumination when depressive symptoms are also present (Kashdan & Roberts, 2007). Moreover, in a study by Moscovitch, McCabe, Antony, Rocca, and Swinson (2008), differences in anger experience and expression found between anxious patients and non-clinical controls were largely eliminated when depression was included as a covariate. These findings indicate that it is important to examine the role of depression when evaluating rumination as a potential mediator of the link between social anxiety and anger.

The present study was designed to evaluate the potential mediating role of rumination in the relationship between social anxiety and anger, focusing on two subtypes of rumination: brooding and reflective pondering. Three anger outcomes were considered: trait anger, anger-out (i.e., outward anger expression), and anger-in (i.e., anger suppression). It appears plausible that the subscales of the RRS, and the cognitive variations that they reflect, may differentially relate to anger outcomes. Brooding, with its negative, self-critical focus and implicit theme of struggling against the reality of events (Rude, Maestas, & Neff, 2007) may foster greater anger experience and greater outward expression of that anger as individuals struggle against their circumstances. Reflective pondering, on the other hand, may lack the extremely negative, self-critical element required to generate and outwardly express anger. Its introspective focus may lead to greater anger suppression as individuals respond to angry affect via withdrawal and contemplation. Understanding how variations in negative thinking affect anger expression might therefore guide cognitive treatments. Based on this reasoning, we predicted that brooding would significantly mediate the relationship between social anxiety and both trait anger and outward anger expression while reflective pondering would mediate the relationship between social anxiety and anger suppression.