Child Abuse & Neglect 51 (2016) 276–283
Contents lists available at ScienceDirect
Child Abuse & Neglect
Research article
Childhood physical abuse and aggression: Shame and narcissistic vulnerability Amanda C. Keene a , James Epps b,∗ a b
Department of Psychology, University of South Florida, 4202 East Fowler Ave, PCD4118G, Tampa, FL 33620, USA Department of Mental Health Law and Policy, University of South Florida, 13301 Bruce B. Downs Blvd, Tampa, FL 33612, USA
a r t i c l e
i n f o
Article history: Received 17 April 2015 Received in revised form 16 July 2015 Accepted 16 September 2015 Available online 11 November 2015 Keywords: Childhood physical abuse Adult aggression Shame Narcissism
a b s t r a c t This study examined narcissistic vulnerability and shame-proneness as potential mediators between childhood physical abuse (CPA) and adult anger and aggression. Participants were 400 undergraduate students, 134 of whom had a history of CPA. All participants completed self-report questionnaires assessing history of CPA, shame-proneness, narcissistic vulnerability, physical aggression, trait anger, and hostility. Results indicated abused participants were more angry and aggressive and experienced higher levels of shame-proneness and narcissistic vulnerability than nonabused participants. Multiple mediation analyses showed that narcissistic vulnerability, but not shame-proneness, partially mediated the relation between abuse and physical aggression. However, narcissistic vulnerability and shameproneness both emerged as partial mediators between abuse and the anger and hostility variables. These findings suggest that narcissistic vulnerability and shame-proneness may function as mediators of adjustment following childhood maltreatment. Study limitations and recommendations for future research are discussed. © 2015 Elsevier Ltd. All rights reserved.
Introduction Childhood physical abuse (CPA) has been identified as a major risk factor for antisocial outcomes in adulthood, including substance abuse, criminal behavior, and aggression (Egeland, Yates, Appleyard, & van Dulmen, 2002; Norman et al., 2012). In particular, a positive relationship between maltreatment and adult aggression has been well-established (Epps, Carlin, & Ward, 1999; Lansford et al., 2007; Maneta, Cohen, Schulz, & Waldinger, 2012; Scarpa, Haden, & Abercromby, 2010). Though definitions vary, the terms “abuse” and “maltreatment” will be used synonymously throughout this article. Studies have identified many factors that may influence the relationship between CPA and adult anger/aggression, including maltreatment characteristics, environmental factors, and cognitive and emotional processes (Chen, Coccaro, Lee, & Jacobson, 2012; Verona & Sachs-Ericsson, 2005). Specifically, poor emotion regulation has been found to underlie aggressive tendencies in adults with a history of CPA (Gratz, Paulson, Jakupcak, & Tull, 2009; Stevens et al., 2013; Teisl & Cicchetti, 2008). The hostile and invalidating environment under which CPA occurs hinders the development of emotion regulation skills (Briere, 2002). Previous studies have shown that victims of childhood maltreatment are less adept at identifying and regulating their emotions in a constructive manner (Briere, 2002; Chen et al., 2012; Stevens et al., 2013; Teisl & Cicchetti, 2008). Poor emotion regulation is associated with greater emotional intensity, difficulty understanding and recognizing emotions,
∗ Corresponding author. http://dx.doi.org/10.1016/j.chiabu.2015.09.012 0145-2134/© 2015 Elsevier Ltd. All rights reserved.
A.C. Keene, J. Epps / Child Abuse & Neglect 51 (2016) 276–283
277
and negative reactivity to certain emotions (Eisenberg, 2000; Lemerise & Arsenio, 2000). As a result, these individuals with a history of maltreatment are particularly prone to experiencing negative emotions, such as shame and anger (Bennett, Sullivan, & Lewis, 2005; Gold, Sullivan, & Lewis, 2011). Over time, repeated experiences of painful emotions, such as shame, can predispose an individual to experiencing shame more readily than other emotions (Lemerise & Arsenio, 2000; Malatesta & Wilson, 1988). Children and adults prone to experiencing shame are more likely to respond with greater anger and aggression toward perceived sources of threat than less shame-prone individuals (Hejdenberg & Andrews, 2011). Furthermore, shame has been identified as an important mediator of adjustment following maltreatment. Bennett et al. (2005) found that physical abuse was related to shame-proneness, which in turn was associated with increased anger in young children. Similarly, Gold et al. (2011) found that among adults, abusive parenting was related to shame-proneness and violent behavior. The relationship between shame, anger, and aggression is further evident in studies of pathological narcissism (Ritter et al., 2013; Schoenleber & Berenbaum, 2012; Thomaes, Bushman, Stegge, & Olthof, 2008; Thomaes, Stegge, Olthof, Bushman, & Nezlek, 2011). Narcissism has traditionally been described in terms of overt grandiosity; however, recent clinical conceptualizations describe pathological narcissism as a heterogeneous construct consisting of grandiose and vulnerable forms (Cain, Pincus, & Ansell, 2008; Dickinson & Pincus, 2003; Hendin & Cheek, 1997; Miller & Campbell, 2008; Pincus et al., 2009; Pincus & Lukowitsky, 2010; Ronningstam, 2010). While narcissistic grandiosity and narcissistic vulnerability have some correlates in common, they manifest divergent relationships with a variety of constructs. Both forms are associated with feelings of entitlement, mistrust, and neuroticism (Cain et al., 2008; Wright et al., 2013; Zeigler-Hill, Green, Arnau, Sisemore, & Meyers, 2011). However, narcissistic grandiosity mirrors the diagnostic criteria for narcissistic personality disorder in the Diagnostic and Statistical Manual of Mental Disorders (American Psychiatric Association, 2013) and is characterized by a sense of superiority to others, and exploitative and entitled behaviors. Narcissistic vulnerability is marked by proneness to shame, hypersensitivity, anger, and aggression (Cain et al., 2008; Pincus et al., 2009). Additionally, narcissistic vulnerability is strongly associated with reactive aggression whereas narcissistic grandiosity is only weakly associated with instrumental aggression (Malkin, Zeigler-Hill, Barry, & Southard, 2012). Furthermore, limited evidence suggests that narcissistic vulnerability is positively associated with self-reported childhood physical, emotional, and sexual abuse whereas narcissistic grandiosity is unrelated (Miller et al., 2010; Roche, Pincus, Lukowitsky, Ménard, & Conroy, 2013). Narcissistic vulnerability is also associated with other abuse sequelae such as depression, anxiety, paranoia, and hostility (Miller et al., 2011). Overview and Predictions The present study aimed to extend previous research on the link between CPA and adult anger and aggression by examining the potential mediating roles of narcissistic vulnerability and shame-proneness. Consistent with previous research, it was expected that adults with a history of CPA would experience more anger and would report more aggression than nonabused adults. It was hypothesized that adults with a history of CPA would be more shame-prone and would exhibit more narcissistic vulnerability, but not narcissistic grandiosity, than nonabused adults. Finally, narcissistic vulnerability and shame-proneness, taken together, were hypothesized to mediate the relationship between history of CPA and aggression and anger. Methods Participants This study recruited 400 undergraduate students (321 female, 79 male; age, M = 21.39 years; age range: 18–49 years) attending a large southern university who received credit toward an undergraduate psychology course in return for their participation. Participants were recruited from the psychology department’s research participant pool. After providing informed consent electronically, participants completed questionnaires anonymously via an online delivery system. Due to the sensitive nature of some of the questions, all participants were informed that they may decline to answer any questions and that they may withdraw from the study at any time without penalty. Measures Demographic Form. The demographic form was used to collect the following information from the participant: age, gender, parents’ marital status, and the number of siblings. Emotional and Physical Abuse Questionnaire. The EPAB (Carlin et al., 1994) consists of 32 items that describe childhood experiences of discipline and abuse. Participants respond on a 6-point Likert-type scale of frequency of occurrence ranging from 0 (never) to 5 (very frequently). The physical abuse subscale consisted of 14 items. Participants were determined to have a history of physical abuse if they endorsed one of the nine major assault items (e.g., bones broken, teeth knocked out, being purposefully burned) or if one of the five other items was endorsed as having occurred frequently (e.g., pinched,
278
A.C. Keene, J. Epps / Child Abuse & Neglect 51 (2016) 276–283
shaken, spanked with a board or stick or wire; Carlin et al., 1994). In the current sample, internal consistency was good for the physical abuse subscale (˛ = .83). Pathological Narcissism Inventory. The PNI (Pincus et al., 2009) is a 52-item questionnaire developed to assess vulnerable and grandiose expressions of pathological narcissism. Each item is rated on a five-point Likert-type scale ranging from 0 (not at all like me) to 5 (very much like me). The PNI measures seven dimensions of pathological narcissism: contingent self-esteem, exploitative, self-sacrificing self-enhancement, hiding the self, grandiose fantasy, devaluing, and entitlement rage. These seven dimensions load onto two higher-order factors: narcissistic grandiosity (exploitative, self-sacrificing selfenhancement, and grandiose fantasy) and narcissistic vulnerability (contingent self-esteem, hiding the self, entitlement rage, and devaluing; Wright, Lukowitsky, Pincus, & Conroy, 2010). In the current sample, internal consistency was good for the vulnerable subscale (˛ = .96) and the grandiose subscale (˛ = .89). Experience of Shame Scale. The ESS (Andrews, Qian, & Valentine, 2002) is a 25-item self-report questionnaire that assesses shame-proneness related to one’s own character, behavior, and body. Each item is rated on a 4-point Likert-type scale ranging from 1 (not at all) to 4 (very much), indicating the frequency of thinking, experiencing, and avoiding any of the three areas of shame during the past year. The ESS has good construct validity, high internal consistency and test–retest reliability (Andrews et al., 2002). In the current sample, internal consistency for the total ESS was strong (˛ = .95). State and Trait Anger Scales. The State and Trait Anger scales of the State Trait Personality Inventory (Spielberger, 1979) consist of 20 items that assess the tendency to experience anger. Although both scales were administered, only the Trait Anger (T-Anger) scale was thought to be pertinent to the current study. Each item of the trait anger subscale is rated on a 4-point Likert-type scale ranging from 1 (almost never) to 4 (always). The Trait Anger scale demonstrated good internal consistency in the current sample (˛ = .87). Aggression Questionnaire. The AQ (Buss & Perry, 1992) is a 29-item scale that measures an individual’s propensity for physical aggression, verbal aggression, anger, and hostility. Each item is rated on a 5-point Likert-type scale ranging from 1 (extremely uncharacteristic of me) to 5 (extremely characteristic of me). In the current sample, the subscales demonstrated good internal consistency (physical aggression, ˛ = .84; verbal aggression, ˛ = .79; hostility, ˛ = .85). Data Analysis To examine differences between physically abused and nonabused participants on the dependent measures, data were analyzed using multivariate analysis of covariance (MANCOVA) with gender as a covariate. All multivariate effects were examined using Hotelling’s Trace statistic, followed by univariate tests post hoc. Although emotional abuse data were available, those data were not used as covariates in these or any other analyses. The literature suggests that CPA exists against a backdrop of family dysfunction, frequently including emotional abuse (Claussen & Crittenden, 1991; Edwards, Holden, Felitti, & Anda, 2003). Briere and colleagues (Briere, 1988; Briere & Elliott, 2003) suggest that “partialling out” the effects of variables with high degrees of multicollinearity creates unlikely scenarios of limited ecological validity (in this case, the physically abusive parent who is not emotionally abusive). Mediation analyses were conducted to determine whether the relationship between history of CPA and adult anger and aggression were mediated by shame-proneness and narcissistic vulnerability. To examine multiple mediators simultaneously, analyses were conducted using a bootstrapping macro for SPSS developed by Preacher and Hayes (2008). This macro estimates path coefficients and generates bias-corrected (BC) confidence intervals for the total and specific indirect effects of multiple mediators. In order to demonstrate full mediation the indirect effect must be significant and the direct effect non-significant. However, partial mediation may be demonstrated when the indirect effect is significant and the direct effects pathway is reduced in absolute size, but not reduced to non-significant levels. Indirect effects were considered significant when the 95% BC confidence intervals did not include zero. This method was preferred over Baron and Kenny’s (1986) causal steps approach because it allows multiple mediators to be tested simultaneously, reducing the risk of Type I error. Due to missing data on some variables, the sample size varied slightly across analyses. Missing data analysis indicated that only 0.2% of all data points were missing. Therefore, missing data were handled with pairwise deletion. Results Descriptive Statistics Among the 400 participants who completed the EPAB, 134 (33%) reported a history of CPA. Abused and nonabused participants did not differ significantly in terms of demographic variables (i.e., age, sex, and parents’ marital status; see Table 1), and there were no differences between participants with and without missing data. MANCOVA results showed significant differences between groups, F(8, 368) = 7.17, p < .001; Wilks’ = .87; 2p = .14. This difference despite the significance of gender as a covariate (F(8, 368) = 4.02, p < .001; Wilks’ = .92; 2p = .08). Using a Bonferroni adjusted p-value of .006 (.05/9),
A.C. Keene, J. Epps / Child Abuse & Neglect 51 (2016) 276–283
279
Table 1 Demographic characteristics of physically abused and non-abused participants. Not abused Age, M (SD) Gender, n (%) Male Female Parents’ marital status, n (%) Single, never married Divorced Widowed Married Cohabitating
Abused
21.29 (3.9)
p
21.60 (4.03)
47 (17.7) 218 (82.3)
30 (22.6) 103 (77.4)
87 (33.0) 52 (19.7) 5 (1.9) 112 (42.4) 8 (3.0)
50 (37.6) 25 (18.8) 1 (0.8) 55 (41.4) 2 (1.5)
.34 .28
.67
Note: Differences between groups were examined using t-tests and chi-square tests. Table 2 Means and standard deviations for physically abused and non-abused participants on dependent variables. Measure
AQ Physical aggression AQ Anger AQ Hostility AQ Verbal aggression Trait anger Narcissistic vulnerability Narcissistic grandiosity Shame-proneness * **
Not abused
Abused
M (SD)
M (SD)
F(1,363)
2p
16.72 (5.72) 12.92 (4.82) 16.13 (6.40) 12.05 (3.90) 16.09 (4.81) 2.88 (1.02) 3.60 (0.92) 48.06 (15.41)
21.24 (7.10) 15.72 (5.34) 19.09 (6.49) 12.87 (4.34) 18.13 (5.72) 3.23 (0.94) 3.86 (0.86) 53.08 (15.35)
42.34** 22.75** 17.47** 3.1 12.07** 10.85** 6.96* 8.81**
.10 .06 .05 .01 .03 .03 .02 .02
p < .05. p < .006, Bonferroni adjusted.
univariate analyses indicated that participants from the abused group reported significantly more physical aggression, trait anger, and hostility, but not verbal aggression. Abused participants also reported experiencing significantly more shame and narcissistic vulnerability, though differences in narcissistic grandiosity did not meet corrected criteria for significance. Table 2 presents the means, standard deviations, and effect sizes of the two groups on the dependent measures. Correlations Because narcissistic vulnerability and narcissistic grandiosity were positively correlated (r = .68, p < .05), partial correlations were conducted to determine whether they showed unique associations with the dependent variables (see Table 3). After controlling for narcissistic grandiosity, the relations between narcissistic vulnerability and all dependent variables remained significant. However, after controlling for narcissistic vulnerability, the only significant association for narcissistic grandiosity was with hostility while all other associations were no longer significant. As such, narcissistic vulnerability was the superior predictor of each dependent variable, and was retained for meditational analyses. Mediation Analyses For the mediation analyses, CPA was the independent variable, narcissistic vulnerability and shame-proneness were entered simultaneously as mediators, and narcissistic grandiosity and gender were included as covariate (see Fig. 1). Separate Table 3 Partial correlations of narcissistic vulnerability and grandiosity to self-report variables. Narcissistic vulnerability
AQ Physical aggression AQ Anger AQ Hostility AQ Verbal aggression Trait anger Shame-proneness * ** ***
p < .05. p < .01. p < .001.
Narcissistic grandiosity
Zero-order
Partial
Zero-order
Partial
.30*** .39*** .60*** .25*** .52*** .68***
.19*** .36*** .54*** .18** .44*** .60***
.24*** .20*** .34*** .18** .33*** .42***
.05 −.11* −.14** .00 −.06 −.12*
280
A.C. Keene, J. Epps / Child Abuse & Neglect 51 (2016) 276–283
.16*
Childhood Physical Abuse
Narcissistic Vulnerability
b1
c (c')
Aggression Variables b2
3.01* Shame
Fig. 1. Narcissistic vulnerability and shame as mediators between childhood physical abuse and aggression variables while controlling for gender and narcissistic grandiosity. Note: Path coefficients are unstandardized. * p < .05.
mediation analyses were conducted for each dependent variable (see Table 4) and are reported below. Per Hayes’ (2009) recommendations, we specified 5,000 bootstrap resamples and calculated 95% BC confidence intervals for the total and specific indirect effects. Physical Aggression. Prior to entering both mediators, the total effect (path c) of CPA on physical aggression was significant (see Table 4) and the direct effect (path c ) remained significant after entering the two mediators. The total indirect effect through narcissistic vulnerability and shame-proneness was significant, indicating partial mediation. The specific indirect effect of each mediator was examined to determine whether either variable significantly contributed to the indirect (i.e., mediated) effect of CPA on physical aggression while also accounting for the other mediator. Narcissistic vulnerability was the only significant indirect pathway partially mediating the effects of CPA on physical aggression (point estimate = 0.19, BC 95% CI [.01, .58]). Shame-proneness did not significantly contribute to the total indirect effect above and beyond narcissistic vulnerability (point estimate = .04, BC 95% CI [−.11, .33]). Hostility. Before entering both mediators, the total effect of CPA on hostility was significant (path c) and the direct effect remained significant after entering the two mediators (path c ). The total indirect effect through narcissistic vulnerability and shame-proneness was significant, indicating partial mediation. Inspection of the specific indirect effect of each mediator indicated that both narcissistic vulnerability (point estimate = 0.56, BC 95% CI [.02, 1.15]) and shame-proneness (point estimate = 0.21, BC 95% CI [.02, .57]) significantly contributed to the total indirect effect. Trait Anger. Before entering both mediators, the total effect of CPA on trait anger was significant and the direct effect remained significant after entering the two mediators. The total indirect effect through narcissistic vulnerability and shame-proneness was significant indicating partial mediation. Inspection of the specific indirect effect of each mediator indicated that both narcissistic vulnerability (point estimate = 0.29, BC 95% CI [.02, .64]) and shame-proneness (point estimate = 0.22, BC 95% CI [.01, .58]) significantly contributed to the total indirect effect. Verbal Aggression. Although the model predicting verbal aggression was significant, r2 = 0.05, F(5, 384) = 5.35, p = 0.001, the total and direct effects of CPA on verbal aggression did not reach significance. As such, no further analyses were conducted.
Table 4 Summary of results for four multiple mediation models with childhood physical abuse as the independent variable. Total effect [c path]
Direct effect [c path]
1.24*** 0.01
3.95***
3.72***
.23 (.02, .61)
Narcissistic vulnerability Shame-proneness
3.62*** 0.07***
2.46***
1.69**
.77 (.08, 1.50)
Trait anger
Narcissistic vulnerability Shame-proneness
1.93*** 0.08***
1.46**
0.94*
.52 (.06, 1.03)
Verbal aggression
Narcissistic vulnerability Shame-proneness
0.56
0.44
.12 (−.02, .35)
Dependent variable
Mediators
Physical aggression
Narcissistic vulnerability Shame-proneness
Hostility
Note: Path coefficients are unstandardized. * p < .05. ** p < .01. *** p < .001.
Effect of mediators on dependent variable [b paths]
0.95** −0.01
Total indirect effect (bias-corrected CIs)
A.C. Keene, J. Epps / Child Abuse & Neglect 51 (2016) 276–283
281
Discussion The purpose of this study was to extend previous research on the relationship between childhood physical abuse and adult anger and aggression by exploring the roles of narcissistic vulnerability and shame-proneness. Consistent with previous research, our study found that physically abused participants reported significantly higher levels of anger and aggression than nonabused participants (Carter, Crabtree, Epps, & Roberts-Davis, 2014; Egeland et al., 2002; Springer, Sheridan, Kuo, & Carnes, 2007). More specifically, participants with self-reported histories of CPA had significantly higher scores on two measures of anger, physical aggression, and hostility. As predicted, abused participants also reported significantly higher levels of narcissistic vulnerability and shame-proneness than nonabused participants. These relationships remained significant even when variance accounted for by gender was taken into account. The mediation analyses suggest that the combined effect of narcissistic vulnerability and shame-proneness may play a role in the angry and aggressive behaviors observed in victims of CPA. Both narcissistic vulnerability and shame-proneness emerged as partial mediators in the relation between CPA and trait anger and hostility. Surprisingly, however, the relation between CPA and physical aggression was partially mediated only by narcissistic vulnerability. Although CPA was associated with shame-proneness, shame-proneness was not associated with physical aggression. It is theoretically interesting that shame-proneness was not a mediator in the relation between CPA and physical aggression. These findings are consistent with other research showing a positive relationship between shame-proneness and anger, but no direct relationship to physical aggression (Farmer & Andrews, 2009; Tangney, Wagner, Fletcher, & Gramzow, 1992). For example, Tangney, Wagner, HillBarlow, Marschall, and Gramzow (1996) reported positive correlations between shame-proneness and both physical and verbal aggression. However, Tangney et al. (1992) found no relationship between shame-proneness and overt expressions of hostility, although shame was positively related to anger arousal and indirect expressions of hostility. It may be that the relationship between shame-proneness and more overt aggressive behaviors is attenuated by the presence of other factors. Taken together, these findings extend previous research by incorporating narcissistic vulnerability and shame-proneness as potential mediators of adjustment in individuals with a history of maltreatment. Narcissistic vulnerability and narcissistic grandiosity demonstrated divergent relations on a variety of measures assessing maladaptive emotions and behaviors. In particular, when narcissistic grandiosity was partialled out, narcissistic vulnerability remained positively correlated with anger, aggression, and shame-proneness. In contrast, when narcissistic vulnerability was partialled out, the only significant association for narcissistic grandiosity was a negative correlation with hostility. These findings highlight the dyadic nature of pathological narcissism and provide further evidence for the discriminant validity of these dimensions. Furthermore, these findings also indicate that variance in the associations between grandiosity and anger and aggression variables may be accounted for by vulnerability, as the significance of these associations dissipate when controlling for vulnerability. Taken together, this suggests that future studies of narcissism and aggression may be more precise if narcissism is viewed as a bipartite construct. This distinction should also be reflected in the models from which such studies are derived. The unique combination of narcissistic vulnerability and shame-proneness may further indicate that adults with a history of maltreatment are especially vulnerable to poor emotion regulation. Although speculative, our findings are consistent with theoretical accounts that the anger and aggression exhibited by victims of CPA may serve a self-protective function by shielding the individual from painful emotions, such as shame. This may help to maintain tenuously positive self-perceptions. Previous research has supported this view, demonstrating that maltreated children are more likely to experience emotion regulation deficits both as children and adults (Gratz et al., 2009; Shields & Cicchetti, 1997; Stevens et al., 2013; Teisl & Cicchetti, 2008). Furthermore, these findings are consistent with the view that pathological narcissism may stem from early childhood maltreatment, which may shape the expression of pathological narcissism. In particular, the unique associations for narcissistic vulnerability are consistent with accounts that frame narcissistic vulnerability as more maladaptive than narcissistic grandiosity (Dickinson & Pincus, 2003). This study is the first to examine the roles of narcissistic vulnerability and shame-proneness collectively on anger and aggression in adults with a history of CPA. Our findings extend the literature on anger and aggression in victims of CPA by showing that self-reported histories of CPA were associated with elevated levels of narcissistic vulnerability and shameproneness, which, in turn, were associated with higher levels of anger and aggression. Although significant, the results of this study should be interpreted with caution due to several limitations. First, because CPA and other forms of abuse so frequently occur together, the effects herein attributed to physical abuse could be a function of emotional or sexual abuse. Second, this study used a sample of convenience. Undergraduate students differ from the wider population in several ways, which limits the generalizability of our findings. Third, the cross sectional design of this study precludes any definitive conclusions regarding causality or temporal sequence between variables. The concurrent measurement of variables is not optimal in the mediation models. Fourth, this study relied on retrospective and self-report data. Finally, although our results were significant, effect sizes between the abused group and the nonabused group on measures of narcissism and shame were small. Although small, they are consistent with other studies of abuse and shame (Gold et al., 2011), and, upon replication, can have important implications for understanding and treating victims of physical abuse. Despite these limitations, the findings suggest a potentially significant area for further examination of the role of childhood physical abuse as an etiological factor in adult anger and aggression. If the results of this study are replicated, then mental health professionals working with aggressive individuals may wish to assess for narcissistic vulnerability and shame, and provide ancillary treatment where appropriate.
282
A.C. Keene, J. Epps / Child Abuse & Neglect 51 (2016) 276–283
Conflict of interest The authors declare that they have no conflict of interest. Ethical approval All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. Informed consent Informed consent was obtained from all individual participants included in the study. References American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Washington, DC: Author. Andrews, B., Qian, M., & Valentine, J. D. (2002). Predicting depressive symptoms with a new measure of shame: The experience of shame scale. British Journal of Clinical Psychology, 41, 29–42. 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, 1173–1182. Bennett, D. S., Sullivan, M. W., & Lewis, M. (2005). Young children’s adjustment as a function of maltreatment, shame and anger. Child Maltreatment, 10, 311–323. Briere, J. (1988). Controlling for family variables in abuse effects research: A critique of the “partialling” approach. Journal of Interpersonal Violence, 3, 80–89. Briere, J. (2002). Treating adult survivors of severe childhood abuse: Further development of an integrative model. In J. E. B Myers, L. Berliner, J. Briere, C. T. Hendrix, T. Reid, & C. Jenny (Eds.), The APSAC handbook on child maltreatment (2nd ed., pp. 175–204). Newbury Park, CA: Sage Publications. Briere, J., & Elliott, D. M. (2003). Prevalence and psychological sequelae of self-reported childhood physical and sexual abuse in a general population sample of men and women. Child Abuse & Neglect, 27, 1205–1222. Buss, A. H., & Perry, M. (1992). The aggression questionnaire. Journal of Personality and Social Psychology, 63, 452–459. Cain, N. M., Pincus, A. L., & Ansell, E. B. (2008). Narcissism at the crossroads: Phenotypic description of pathological narcissism across clinical theory, social/personality psychology, and psychiatric diagnosis. Clinical Psychology Review, 28, 638–656. Carlin, A. S., Kemper, K., Ward, N. G., Sowell, H., Gustafson, B., & Stevens, N. (1994). The effect of differences in objective and subjective definitions of childhood physical abuse on estimates of its incidence and relationship to psychopathology. Child Abuse & Neglect, 18, 393–399. Carter, S. D., Crabtree, N. L., Epps, J., & Roberts-Davis, A. M. (2014). Age of onset of physical abuse: Implications for adult anger and aggression. Universal Journal of Psychology, 2, 131–135. Chen, P., Coccaro, E. F., Lee, R., & Jacobson, K. C. (2012). Moderating effects of childhood maltreatment on associations between social information processing and adult aggression. Psychological Medicine, 42, 1293–1304. Claussen, A. H., & Crittenden, P. M. (1991). Physical and psychological maltreatment: Relations among types of maltreatment. Child Abuse & Neglect, 15, 5–18. Dickinson, K. A., & Pincus, A. L. (2003). Interpersonal analysis of grandiose and vulnerable narcissism. Journal of Personality Disorders, 17, 188–207. Edwards, V. J., Holden, G. W., Felitti, V. J., & Anda, R. F. (2003). Relationship between multiple forms of childhood maltreatment and adult mental health in community respondents: Results from the adverse childhood experiences study. American Journal of Psychiatry, 160, 1453–1460. Egeland, B., Yates, T., Appleyard, K., & van Dulmen, M. (2002). The long term consequences of maltreatment in the early years: A developmental pathway model to antisocial behavior. Children’s Services: Social Policy, Research, and Practice, 5, 249–260. Eisenberg, N. (2000). Emotion, regulation, and moral development. Annual Review of Psychology, 51, 665–697. Epps, J., Carlin, A. S., & Ward, N. G. (1999). Adult anger expression and childhood physical abuse history: The effects of event memory vs. self-labeling. Journal of Mental Health and Aging, 5, 175–185. Farmer, E., & Andrews, B. (2009). Shameless yet angry: Shame and its relationship to anger in male young offenders and undergraduate controls. Journal of Forensic Psychiatry & Psychology, 20, 48–65. Gold, J., Sullivan, M. W., & Lewis, M. (2011). The relation between abuse and violent delinquency: The conversion of shame to blame in juvenile offenders. Child Abuse & Neglect, 35, 459–467. Gratz, K. L., Paulson, A., Jakupcak, M., & Tull, M. T. (2009). Exploring the relationship between childhood maltreatment and intimate partner abuse: Gender differences in the mediating role of emotion dysregulation. Violence & Victims, 24, 68–82. Hayes, A. F. (2009). Beyond Baron and Kenny: Statistical mediation analysis in the new millennium. Communication Monographs, 76, 408–420. Hejdenberg, J., & Andrews, B. (2011). The relationship between shame and different types of anger: A theory-based investigation. Personality and Individual Differences, 50, 1278–1282. Hendin, H. M., & Cheek, J. M. (1997). Assessing hypersensitive narcissism: A reexamination of Murray’s Narcism Scale. Journal of Research in Personality, 31, 588–599. Lansford, J. E., Miller-Johnson, S., Berlin, L. J., Dodge, K. A., Bates, J. E., & Pettit, G. S. (2007). Early physical abuse and later violent delinquency: A prospective longitudinal study. Child Maltreatment, 12, 233–245. Lemerise, E. A., & Arsenio, W. F. (2000). An integrated model of emotion processes and cognition in social information processing. Child Development, 71, 107–118. Malatesta, C. Z., & Wilson, A. (1988). Emotion cognition interaction in personality development: A discrete emotions, functionalist analysis. British Journal of Social Psychology, 27, 91–112. Malkin, M. L., Zeigler-Hill, V., Barry, C. T., & Southard, A. C. (2012). The view from the looking glass: How are narcissistic individuals perceived by others? Journal of Personality, 81, 1–15. Maneta, E., Cohen, S., Schulz, M., & Waldinger, R. J. (2012). Links between childhood physical abuse and intimate partner aggression: The mediating role of anger expression. Violence and Victims, 27, 315–328. Miller, J. D., & Campbell, W. K. (2008). Comparing clinical and social-personality conceptualizations of narcissism. Journal of Personality, 76, 449–476. Miller, J. D., Dir, A., Gentile, B., Wilson, L., Pryor, L. R., & Campbell, W. K. (2010). Searching for a vulnerable dark triad: Comparing factor 2 psychopathy, vulnerable narcissism, and borderline personality disorder. Journal of Personality, 78, 1529–1564. Miller, J. D., Hoffman, B. J., Gaughan, E. T., Gentile, B., Maples, J., & Campbell, W. K. (2011). Grandiose and vulnerable narcissism: A nomological network analysis. Journal of Personality, 79, 1013–1042. Norman, R. E., Byambaa, M., De, R., Butchart, A., Scott, J., & Vos, T. (2012). The long-term health consequences of child physical abuse, emotional abuse, and neglect: A systematic review and meta-analysis. PLoS Medicine, 9, e1001349.
A.C. Keene, J. Epps / Child Abuse & Neglect 51 (2016) 276–283
283
Pincus, A. L., Ansell, E. B., Pimentel, C. A., Cain, N. M., Wright, A. G. C., & Levy, K. N. (2009). Initial construction and validation of the pathological narcissism inventory. Psychological Assessment, 21, 365–379. Pincus, A. L., & Lukowitsky, M. R. (2010). Pathological narcissism and narcissistic personality disorder. Annual Review of Clinical Psychology, 6, 421–446. Preacher, K. J., & Hayes, A. F. (2008). Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models. Behavior Research Methods, 40, 879–891. Ritter, K., Vater, A., Rüsch, N., Schröder-Abé, M., Schütz, A., & Fydrich, T. (2013). Shame in patients with narcissistic personality disorder. Psychiatry Research, 215, 429–437. Roche, M. J., Pincus, A. L., Lukowitsky, M. R., Ménard, K. S., & Conroy, D. E. (2013). An integrative approach to the assessment of narcissism. Journal of Personality Assessment, 95, 237–248. Ronningstam, E. F. (2010). Narcissistic personality disorder: A current review. Current Psychiatry Review, 12, 68–75. Scarpa, A., Haden, S. C., & Abercromby, J. M. (2010). Pathways linking child physical abuse, depression, and aggressiveness across genders. Journal of Aggression, Maltreatment, & Trauma, 19, 757–776. Schoenleber, M., & Berenbaum, H. (2012). Shame regulation in personality pathology. Journal of Abnormal Psychology, 121, 433–446. Shields, A., & Cicchetti, D. (1997). Emotion regulation among school-age children: The development and validation of a new criterion q-sort scale. Developmental Psychology, 33, 906–916. Spielberger, C. D. (1979). Preliminary manual for the State-Trait Personality Inventory. Tampa, Florida: Unpublished manual, University of South Florida. Springer, K. W., Sheridan, J., Kuo, D., & Carnes, M. (2007). Long-term physical and mental health consequences of childhood physical abuse: Results from a large population-based sample of men and women. Child Abuse & Neglect, 31, 517–530. Stevens, N. R., Gerhart, J., Goldsmith, R. E., Heath, N. M., Chesney, S. A., & Hobfoll, S. E. (2013). Emotion regulation difficulties, low social support, and interpersonal violence mediate the link between childhood abuse and posttraumatic stress symptoms. Behavior Therapy, 44, 152–161. Tangney, J. P., Wagner, P., Fletcher, C., & Gramzow, R. (1992). Shamed into anger? The relation of shame and guilt to anger and self-reported aggression. Journal of Personality and Social Psychology, 62, 669–675. Tangney, J. P., Wagner, P., Hill-Barlow, D., Marschall, D. E., & Gramzow, R. (1996). Relation of shame and guilt to constructive versus destructive responses to anger across the lifespan. Journal of Personality and Social Psychology, 70, 707–809. Teisl, M., & Cicchetti, D. (2008). Physical abuse, cognitive and emotional processes, and aggressive/disruptive behavior problems. Social Development, 17, 1–23. Thomaes, S., Bushman, B. J., Stegge, H., & Olthof, T. (2008). Trumping shame by blasts of noise: Narcissism, self-esteem, shame, and aggression in young adolescents. Child Development, 79, 1792–1801. Thomaes, S., Stegge, H., Olthof, T., Bushman, B. J., & Nezlek, J. B. (2011). Turning shame inside-out: “Humiliated fury” in young adolescents. Emotion, 11, 786–793. Verona, E., & Sachs-Ericsson, N. (2005). The intergenerational transmission of externalizing behaviors in adult participants: The mediating role of childhood abuse. Journal of Consulting and Clinical Psychology, 73, 1135–1145. Wright, A. G. C., Lukowitsky, M. R., Pincus, A. L., & Conroy, D. E. (2010). The higher order factor structure and gender invariance of the Pathological Narcissism Inventory. Assessment, 17, 467–483. Wright, A. G. C., Pincus, A. L., Thomas, K. M., Hopwood, C. J., Markon, K. E., & Krueger, R. F. (2013). Conceptions of narcissism and the DSM-5 pathological personality traits. Assessment, 20, 339–352. Zeigler-Hill, V., Green, B. A., Arnau, R. C., Sisemore, T. B., & Meyers, E. M. (2011). Trouble ahead, trouble behind: Narcissism and early maladaptive schemas. Journal of Behavior Therapy and Experimental Psychiatry, 42, 96–103.
Contents lists available at ScienceDirect
Child Abuse & Neglect
Research article
Childhood physical abuse and aggression: Shame and narcissistic vulnerability Amanda C. Keene a , James Epps b,∗ a b
Department of Psychology, University of South Florida, 4202 East Fowler Ave, PCD4118G, Tampa, FL 33620, USA Department of Mental Health Law and Policy, University of South Florida, 13301 Bruce B. Downs Blvd, Tampa, FL 33612, USA
a r t i c l e
i n f o
Article history: Received 17 April 2015 Received in revised form 16 July 2015 Accepted 16 September 2015 Available online 11 November 2015 Keywords: Childhood physical abuse Adult aggression Shame Narcissism
a b s t r a c t This study examined narcissistic vulnerability and shame-proneness as potential mediators between childhood physical abuse (CPA) and adult anger and aggression. Participants were 400 undergraduate students, 134 of whom had a history of CPA. All participants completed self-report questionnaires assessing history of CPA, shame-proneness, narcissistic vulnerability, physical aggression, trait anger, and hostility. Results indicated abused participants were more angry and aggressive and experienced higher levels of shame-proneness and narcissistic vulnerability than nonabused participants. Multiple mediation analyses showed that narcissistic vulnerability, but not shame-proneness, partially mediated the relation between abuse and physical aggression. However, narcissistic vulnerability and shameproneness both emerged as partial mediators between abuse and the anger and hostility variables. These findings suggest that narcissistic vulnerability and shame-proneness may function as mediators of adjustment following childhood maltreatment. Study limitations and recommendations for future research are discussed. © 2015 Elsevier Ltd. All rights reserved.
Introduction Childhood physical abuse (CPA) has been identified as a major risk factor for antisocial outcomes in adulthood, including substance abuse, criminal behavior, and aggression (Egeland, Yates, Appleyard, & van Dulmen, 2002; Norman et al., 2012). In particular, a positive relationship between maltreatment and adult aggression has been well-established (Epps, Carlin, & Ward, 1999; Lansford et al., 2007; Maneta, Cohen, Schulz, & Waldinger, 2012; Scarpa, Haden, & Abercromby, 2010). Though definitions vary, the terms “abuse” and “maltreatment” will be used synonymously throughout this article. Studies have identified many factors that may influence the relationship between CPA and adult anger/aggression, including maltreatment characteristics, environmental factors, and cognitive and emotional processes (Chen, Coccaro, Lee, & Jacobson, 2012; Verona & Sachs-Ericsson, 2005). Specifically, poor emotion regulation has been found to underlie aggressive tendencies in adults with a history of CPA (Gratz, Paulson, Jakupcak, & Tull, 2009; Stevens et al., 2013; Teisl & Cicchetti, 2008). The hostile and invalidating environment under which CPA occurs hinders the development of emotion regulation skills (Briere, 2002). Previous studies have shown that victims of childhood maltreatment are less adept at identifying and regulating their emotions in a constructive manner (Briere, 2002; Chen et al., 2012; Stevens et al., 2013; Teisl & Cicchetti, 2008). Poor emotion regulation is associated with greater emotional intensity, difficulty understanding and recognizing emotions,
∗ Corresponding author. http://dx.doi.org/10.1016/j.chiabu.2015.09.012 0145-2134/© 2015 Elsevier Ltd. All rights reserved.
A.C. Keene, J. Epps / Child Abuse & Neglect 51 (2016) 276–283
277
and negative reactivity to certain emotions (Eisenberg, 2000; Lemerise & Arsenio, 2000). As a result, these individuals with a history of maltreatment are particularly prone to experiencing negative emotions, such as shame and anger (Bennett, Sullivan, & Lewis, 2005; Gold, Sullivan, & Lewis, 2011). Over time, repeated experiences of painful emotions, such as shame, can predispose an individual to experiencing shame more readily than other emotions (Lemerise & Arsenio, 2000; Malatesta & Wilson, 1988). Children and adults prone to experiencing shame are more likely to respond with greater anger and aggression toward perceived sources of threat than less shame-prone individuals (Hejdenberg & Andrews, 2011). Furthermore, shame has been identified as an important mediator of adjustment following maltreatment. Bennett et al. (2005) found that physical abuse was related to shame-proneness, which in turn was associated with increased anger in young children. Similarly, Gold et al. (2011) found that among adults, abusive parenting was related to shame-proneness and violent behavior. The relationship between shame, anger, and aggression is further evident in studies of pathological narcissism (Ritter et al., 2013; Schoenleber & Berenbaum, 2012; Thomaes, Bushman, Stegge, & Olthof, 2008; Thomaes, Stegge, Olthof, Bushman, & Nezlek, 2011). Narcissism has traditionally been described in terms of overt grandiosity; however, recent clinical conceptualizations describe pathological narcissism as a heterogeneous construct consisting of grandiose and vulnerable forms (Cain, Pincus, & Ansell, 2008; Dickinson & Pincus, 2003; Hendin & Cheek, 1997; Miller & Campbell, 2008; Pincus et al., 2009; Pincus & Lukowitsky, 2010; Ronningstam, 2010). While narcissistic grandiosity and narcissistic vulnerability have some correlates in common, they manifest divergent relationships with a variety of constructs. Both forms are associated with feelings of entitlement, mistrust, and neuroticism (Cain et al., 2008; Wright et al., 2013; Zeigler-Hill, Green, Arnau, Sisemore, & Meyers, 2011). However, narcissistic grandiosity mirrors the diagnostic criteria for narcissistic personality disorder in the Diagnostic and Statistical Manual of Mental Disorders (American Psychiatric Association, 2013) and is characterized by a sense of superiority to others, and exploitative and entitled behaviors. Narcissistic vulnerability is marked by proneness to shame, hypersensitivity, anger, and aggression (Cain et al., 2008; Pincus et al., 2009). Additionally, narcissistic vulnerability is strongly associated with reactive aggression whereas narcissistic grandiosity is only weakly associated with instrumental aggression (Malkin, Zeigler-Hill, Barry, & Southard, 2012). Furthermore, limited evidence suggests that narcissistic vulnerability is positively associated with self-reported childhood physical, emotional, and sexual abuse whereas narcissistic grandiosity is unrelated (Miller et al., 2010; Roche, Pincus, Lukowitsky, Ménard, & Conroy, 2013). Narcissistic vulnerability is also associated with other abuse sequelae such as depression, anxiety, paranoia, and hostility (Miller et al., 2011). Overview and Predictions The present study aimed to extend previous research on the link between CPA and adult anger and aggression by examining the potential mediating roles of narcissistic vulnerability and shame-proneness. Consistent with previous research, it was expected that adults with a history of CPA would experience more anger and would report more aggression than nonabused adults. It was hypothesized that adults with a history of CPA would be more shame-prone and would exhibit more narcissistic vulnerability, but not narcissistic grandiosity, than nonabused adults. Finally, narcissistic vulnerability and shame-proneness, taken together, were hypothesized to mediate the relationship between history of CPA and aggression and anger. Methods Participants This study recruited 400 undergraduate students (321 female, 79 male; age, M = 21.39 years; age range: 18–49 years) attending a large southern university who received credit toward an undergraduate psychology course in return for their participation. Participants were recruited from the psychology department’s research participant pool. After providing informed consent electronically, participants completed questionnaires anonymously via an online delivery system. Due to the sensitive nature of some of the questions, all participants were informed that they may decline to answer any questions and that they may withdraw from the study at any time without penalty. Measures Demographic Form. The demographic form was used to collect the following information from the participant: age, gender, parents’ marital status, and the number of siblings. Emotional and Physical Abuse Questionnaire. The EPAB (Carlin et al., 1994) consists of 32 items that describe childhood experiences of discipline and abuse. Participants respond on a 6-point Likert-type scale of frequency of occurrence ranging from 0 (never) to 5 (very frequently). The physical abuse subscale consisted of 14 items. Participants were determined to have a history of physical abuse if they endorsed one of the nine major assault items (e.g., bones broken, teeth knocked out, being purposefully burned) or if one of the five other items was endorsed as having occurred frequently (e.g., pinched,
278
A.C. Keene, J. Epps / Child Abuse & Neglect 51 (2016) 276–283
shaken, spanked with a board or stick or wire; Carlin et al., 1994). In the current sample, internal consistency was good for the physical abuse subscale (˛ = .83). Pathological Narcissism Inventory. The PNI (Pincus et al., 2009) is a 52-item questionnaire developed to assess vulnerable and grandiose expressions of pathological narcissism. Each item is rated on a five-point Likert-type scale ranging from 0 (not at all like me) to 5 (very much like me). The PNI measures seven dimensions of pathological narcissism: contingent self-esteem, exploitative, self-sacrificing self-enhancement, hiding the self, grandiose fantasy, devaluing, and entitlement rage. These seven dimensions load onto two higher-order factors: narcissistic grandiosity (exploitative, self-sacrificing selfenhancement, and grandiose fantasy) and narcissistic vulnerability (contingent self-esteem, hiding the self, entitlement rage, and devaluing; Wright, Lukowitsky, Pincus, & Conroy, 2010). In the current sample, internal consistency was good for the vulnerable subscale (˛ = .96) and the grandiose subscale (˛ = .89). Experience of Shame Scale. The ESS (Andrews, Qian, & Valentine, 2002) is a 25-item self-report questionnaire that assesses shame-proneness related to one’s own character, behavior, and body. Each item is rated on a 4-point Likert-type scale ranging from 1 (not at all) to 4 (very much), indicating the frequency of thinking, experiencing, and avoiding any of the three areas of shame during the past year. The ESS has good construct validity, high internal consistency and test–retest reliability (Andrews et al., 2002). In the current sample, internal consistency for the total ESS was strong (˛ = .95). State and Trait Anger Scales. The State and Trait Anger scales of the State Trait Personality Inventory (Spielberger, 1979) consist of 20 items that assess the tendency to experience anger. Although both scales were administered, only the Trait Anger (T-Anger) scale was thought to be pertinent to the current study. Each item of the trait anger subscale is rated on a 4-point Likert-type scale ranging from 1 (almost never) to 4 (always). The Trait Anger scale demonstrated good internal consistency in the current sample (˛ = .87). Aggression Questionnaire. The AQ (Buss & Perry, 1992) is a 29-item scale that measures an individual’s propensity for physical aggression, verbal aggression, anger, and hostility. Each item is rated on a 5-point Likert-type scale ranging from 1 (extremely uncharacteristic of me) to 5 (extremely characteristic of me). In the current sample, the subscales demonstrated good internal consistency (physical aggression, ˛ = .84; verbal aggression, ˛ = .79; hostility, ˛ = .85). Data Analysis To examine differences between physically abused and nonabused participants on the dependent measures, data were analyzed using multivariate analysis of covariance (MANCOVA) with gender as a covariate. All multivariate effects were examined using Hotelling’s Trace statistic, followed by univariate tests post hoc. Although emotional abuse data were available, those data were not used as covariates in these or any other analyses. The literature suggests that CPA exists against a backdrop of family dysfunction, frequently including emotional abuse (Claussen & Crittenden, 1991; Edwards, Holden, Felitti, & Anda, 2003). Briere and colleagues (Briere, 1988; Briere & Elliott, 2003) suggest that “partialling out” the effects of variables with high degrees of multicollinearity creates unlikely scenarios of limited ecological validity (in this case, the physically abusive parent who is not emotionally abusive). Mediation analyses were conducted to determine whether the relationship between history of CPA and adult anger and aggression were mediated by shame-proneness and narcissistic vulnerability. To examine multiple mediators simultaneously, analyses were conducted using a bootstrapping macro for SPSS developed by Preacher and Hayes (2008). This macro estimates path coefficients and generates bias-corrected (BC) confidence intervals for the total and specific indirect effects of multiple mediators. In order to demonstrate full mediation the indirect effect must be significant and the direct effect non-significant. However, partial mediation may be demonstrated when the indirect effect is significant and the direct effects pathway is reduced in absolute size, but not reduced to non-significant levels. Indirect effects were considered significant when the 95% BC confidence intervals did not include zero. This method was preferred over Baron and Kenny’s (1986) causal steps approach because it allows multiple mediators to be tested simultaneously, reducing the risk of Type I error. Due to missing data on some variables, the sample size varied slightly across analyses. Missing data analysis indicated that only 0.2% of all data points were missing. Therefore, missing data were handled with pairwise deletion. Results Descriptive Statistics Among the 400 participants who completed the EPAB, 134 (33%) reported a history of CPA. Abused and nonabused participants did not differ significantly in terms of demographic variables (i.e., age, sex, and parents’ marital status; see Table 1), and there were no differences between participants with and without missing data. MANCOVA results showed significant differences between groups, F(8, 368) = 7.17, p < .001; Wilks’ = .87; 2p = .14. This difference despite the significance of gender as a covariate (F(8, 368) = 4.02, p < .001; Wilks’ = .92; 2p = .08). Using a Bonferroni adjusted p-value of .006 (.05/9),
A.C. Keene, J. Epps / Child Abuse & Neglect 51 (2016) 276–283
279
Table 1 Demographic characteristics of physically abused and non-abused participants. Not abused Age, M (SD) Gender, n (%) Male Female Parents’ marital status, n (%) Single, never married Divorced Widowed Married Cohabitating
Abused
21.29 (3.9)
p
21.60 (4.03)
47 (17.7) 218 (82.3)
30 (22.6) 103 (77.4)
87 (33.0) 52 (19.7) 5 (1.9) 112 (42.4) 8 (3.0)
50 (37.6) 25 (18.8) 1 (0.8) 55 (41.4) 2 (1.5)
.34 .28
.67
Note: Differences between groups were examined using t-tests and chi-square tests. Table 2 Means and standard deviations for physically abused and non-abused participants on dependent variables. Measure
AQ Physical aggression AQ Anger AQ Hostility AQ Verbal aggression Trait anger Narcissistic vulnerability Narcissistic grandiosity Shame-proneness * **
Not abused
Abused
M (SD)
M (SD)
F(1,363)
2p
16.72 (5.72) 12.92 (4.82) 16.13 (6.40) 12.05 (3.90) 16.09 (4.81) 2.88 (1.02) 3.60 (0.92) 48.06 (15.41)
21.24 (7.10) 15.72 (5.34) 19.09 (6.49) 12.87 (4.34) 18.13 (5.72) 3.23 (0.94) 3.86 (0.86) 53.08 (15.35)
42.34** 22.75** 17.47** 3.1 12.07** 10.85** 6.96* 8.81**
.10 .06 .05 .01 .03 .03 .02 .02
p < .05. p < .006, Bonferroni adjusted.
univariate analyses indicated that participants from the abused group reported significantly more physical aggression, trait anger, and hostility, but not verbal aggression. Abused participants also reported experiencing significantly more shame and narcissistic vulnerability, though differences in narcissistic grandiosity did not meet corrected criteria for significance. Table 2 presents the means, standard deviations, and effect sizes of the two groups on the dependent measures. Correlations Because narcissistic vulnerability and narcissistic grandiosity were positively correlated (r = .68, p < .05), partial correlations were conducted to determine whether they showed unique associations with the dependent variables (see Table 3). After controlling for narcissistic grandiosity, the relations between narcissistic vulnerability and all dependent variables remained significant. However, after controlling for narcissistic vulnerability, the only significant association for narcissistic grandiosity was with hostility while all other associations were no longer significant. As such, narcissistic vulnerability was the superior predictor of each dependent variable, and was retained for meditational analyses. Mediation Analyses For the mediation analyses, CPA was the independent variable, narcissistic vulnerability and shame-proneness were entered simultaneously as mediators, and narcissistic grandiosity and gender were included as covariate (see Fig. 1). Separate Table 3 Partial correlations of narcissistic vulnerability and grandiosity to self-report variables. Narcissistic vulnerability
AQ Physical aggression AQ Anger AQ Hostility AQ Verbal aggression Trait anger Shame-proneness * ** ***
p < .05. p < .01. p < .001.
Narcissistic grandiosity
Zero-order
Partial
Zero-order
Partial
.30*** .39*** .60*** .25*** .52*** .68***
.19*** .36*** .54*** .18** .44*** .60***
.24*** .20*** .34*** .18** .33*** .42***
.05 −.11* −.14** .00 −.06 −.12*
280
A.C. Keene, J. Epps / Child Abuse & Neglect 51 (2016) 276–283
.16*
Childhood Physical Abuse
Narcissistic Vulnerability
b1
c (c')
Aggression Variables b2
3.01* Shame
Fig. 1. Narcissistic vulnerability and shame as mediators between childhood physical abuse and aggression variables while controlling for gender and narcissistic grandiosity. Note: Path coefficients are unstandardized. * p < .05.
mediation analyses were conducted for each dependent variable (see Table 4) and are reported below. Per Hayes’ (2009) recommendations, we specified 5,000 bootstrap resamples and calculated 95% BC confidence intervals for the total and specific indirect effects. Physical Aggression. Prior to entering both mediators, the total effect (path c) of CPA on physical aggression was significant (see Table 4) and the direct effect (path c ) remained significant after entering the two mediators. The total indirect effect through narcissistic vulnerability and shame-proneness was significant, indicating partial mediation. The specific indirect effect of each mediator was examined to determine whether either variable significantly contributed to the indirect (i.e., mediated) effect of CPA on physical aggression while also accounting for the other mediator. Narcissistic vulnerability was the only significant indirect pathway partially mediating the effects of CPA on physical aggression (point estimate = 0.19, BC 95% CI [.01, .58]). Shame-proneness did not significantly contribute to the total indirect effect above and beyond narcissistic vulnerability (point estimate = .04, BC 95% CI [−.11, .33]). Hostility. Before entering both mediators, the total effect of CPA on hostility was significant (path c) and the direct effect remained significant after entering the two mediators (path c ). The total indirect effect through narcissistic vulnerability and shame-proneness was significant, indicating partial mediation. Inspection of the specific indirect effect of each mediator indicated that both narcissistic vulnerability (point estimate = 0.56, BC 95% CI [.02, 1.15]) and shame-proneness (point estimate = 0.21, BC 95% CI [.02, .57]) significantly contributed to the total indirect effect. Trait Anger. Before entering both mediators, the total effect of CPA on trait anger was significant and the direct effect remained significant after entering the two mediators. The total indirect effect through narcissistic vulnerability and shame-proneness was significant indicating partial mediation. Inspection of the specific indirect effect of each mediator indicated that both narcissistic vulnerability (point estimate = 0.29, BC 95% CI [.02, .64]) and shame-proneness (point estimate = 0.22, BC 95% CI [.01, .58]) significantly contributed to the total indirect effect. Verbal Aggression. Although the model predicting verbal aggression was significant, r2 = 0.05, F(5, 384) = 5.35, p = 0.001, the total and direct effects of CPA on verbal aggression did not reach significance. As such, no further analyses were conducted.
Table 4 Summary of results for four multiple mediation models with childhood physical abuse as the independent variable. Total effect [c path]
Direct effect [c path]
1.24*** 0.01
3.95***
3.72***
.23 (.02, .61)
Narcissistic vulnerability Shame-proneness
3.62*** 0.07***
2.46***
1.69**
.77 (.08, 1.50)
Trait anger
Narcissistic vulnerability Shame-proneness
1.93*** 0.08***
1.46**
0.94*
.52 (.06, 1.03)
Verbal aggression
Narcissistic vulnerability Shame-proneness
0.56
0.44
.12 (−.02, .35)
Dependent variable
Mediators
Physical aggression
Narcissistic vulnerability Shame-proneness
Hostility
Note: Path coefficients are unstandardized. * p < .05. ** p < .01. *** p < .001.
Effect of mediators on dependent variable [b paths]
0.95** −0.01
Total indirect effect (bias-corrected CIs)
A.C. Keene, J. Epps / Child Abuse & Neglect 51 (2016) 276–283
281
Discussion The purpose of this study was to extend previous research on the relationship between childhood physical abuse and adult anger and aggression by exploring the roles of narcissistic vulnerability and shame-proneness. Consistent with previous research, our study found that physically abused participants reported significantly higher levels of anger and aggression than nonabused participants (Carter, Crabtree, Epps, & Roberts-Davis, 2014; Egeland et al., 2002; Springer, Sheridan, Kuo, & Carnes, 2007). More specifically, participants with self-reported histories of CPA had significantly higher scores on two measures of anger, physical aggression, and hostility. As predicted, abused participants also reported significantly higher levels of narcissistic vulnerability and shame-proneness than nonabused participants. These relationships remained significant even when variance accounted for by gender was taken into account. The mediation analyses suggest that the combined effect of narcissistic vulnerability and shame-proneness may play a role in the angry and aggressive behaviors observed in victims of CPA. Both narcissistic vulnerability and shame-proneness emerged as partial mediators in the relation between CPA and trait anger and hostility. Surprisingly, however, the relation between CPA and physical aggression was partially mediated only by narcissistic vulnerability. Although CPA was associated with shame-proneness, shame-proneness was not associated with physical aggression. It is theoretically interesting that shame-proneness was not a mediator in the relation between CPA and physical aggression. These findings are consistent with other research showing a positive relationship between shame-proneness and anger, but no direct relationship to physical aggression (Farmer & Andrews, 2009; Tangney, Wagner, Fletcher, & Gramzow, 1992). For example, Tangney, Wagner, HillBarlow, Marschall, and Gramzow (1996) reported positive correlations between shame-proneness and both physical and verbal aggression. However, Tangney et al. (1992) found no relationship between shame-proneness and overt expressions of hostility, although shame was positively related to anger arousal and indirect expressions of hostility. It may be that the relationship between shame-proneness and more overt aggressive behaviors is attenuated by the presence of other factors. Taken together, these findings extend previous research by incorporating narcissistic vulnerability and shame-proneness as potential mediators of adjustment in individuals with a history of maltreatment. Narcissistic vulnerability and narcissistic grandiosity demonstrated divergent relations on a variety of measures assessing maladaptive emotions and behaviors. In particular, when narcissistic grandiosity was partialled out, narcissistic vulnerability remained positively correlated with anger, aggression, and shame-proneness. In contrast, when narcissistic vulnerability was partialled out, the only significant association for narcissistic grandiosity was a negative correlation with hostility. These findings highlight the dyadic nature of pathological narcissism and provide further evidence for the discriminant validity of these dimensions. Furthermore, these findings also indicate that variance in the associations between grandiosity and anger and aggression variables may be accounted for by vulnerability, as the significance of these associations dissipate when controlling for vulnerability. Taken together, this suggests that future studies of narcissism and aggression may be more precise if narcissism is viewed as a bipartite construct. This distinction should also be reflected in the models from which such studies are derived. The unique combination of narcissistic vulnerability and shame-proneness may further indicate that adults with a history of maltreatment are especially vulnerable to poor emotion regulation. Although speculative, our findings are consistent with theoretical accounts that the anger and aggression exhibited by victims of CPA may serve a self-protective function by shielding the individual from painful emotions, such as shame. This may help to maintain tenuously positive self-perceptions. Previous research has supported this view, demonstrating that maltreated children are more likely to experience emotion regulation deficits both as children and adults (Gratz et al., 2009; Shields & Cicchetti, 1997; Stevens et al., 2013; Teisl & Cicchetti, 2008). Furthermore, these findings are consistent with the view that pathological narcissism may stem from early childhood maltreatment, which may shape the expression of pathological narcissism. In particular, the unique associations for narcissistic vulnerability are consistent with accounts that frame narcissistic vulnerability as more maladaptive than narcissistic grandiosity (Dickinson & Pincus, 2003). This study is the first to examine the roles of narcissistic vulnerability and shame-proneness collectively on anger and aggression in adults with a history of CPA. Our findings extend the literature on anger and aggression in victims of CPA by showing that self-reported histories of CPA were associated with elevated levels of narcissistic vulnerability and shameproneness, which, in turn, were associated with higher levels of anger and aggression. Although significant, the results of this study should be interpreted with caution due to several limitations. First, because CPA and other forms of abuse so frequently occur together, the effects herein attributed to physical abuse could be a function of emotional or sexual abuse. Second, this study used a sample of convenience. Undergraduate students differ from the wider population in several ways, which limits the generalizability of our findings. Third, the cross sectional design of this study precludes any definitive conclusions regarding causality or temporal sequence between variables. The concurrent measurement of variables is not optimal in the mediation models. Fourth, this study relied on retrospective and self-report data. Finally, although our results were significant, effect sizes between the abused group and the nonabused group on measures of narcissism and shame were small. Although small, they are consistent with other studies of abuse and shame (Gold et al., 2011), and, upon replication, can have important implications for understanding and treating victims of physical abuse. Despite these limitations, the findings suggest a potentially significant area for further examination of the role of childhood physical abuse as an etiological factor in adult anger and aggression. If the results of this study are replicated, then mental health professionals working with aggressive individuals may wish to assess for narcissistic vulnerability and shame, and provide ancillary treatment where appropriate.
282
A.C. Keene, J. Epps / Child Abuse & Neglect 51 (2016) 276–283
Conflict of interest The authors declare that they have no conflict of interest. Ethical approval All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. Informed consent Informed consent was obtained from all individual participants included in the study. References American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Washington, DC: Author. Andrews, B., Qian, M., & Valentine, J. D. (2002). Predicting depressive symptoms with a new measure of shame: The experience of shame scale. British Journal of Clinical Psychology, 41, 29–42. 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, 1173–1182. Bennett, D. S., Sullivan, M. W., & Lewis, M. (2005). Young children’s adjustment as a function of maltreatment, shame and anger. Child Maltreatment, 10, 311–323. Briere, J. (1988). Controlling for family variables in abuse effects research: A critique of the “partialling” approach. Journal of Interpersonal Violence, 3, 80–89. Briere, J. (2002). Treating adult survivors of severe childhood abuse: Further development of an integrative model. In J. E. B Myers, L. Berliner, J. Briere, C. T. Hendrix, T. Reid, & C. Jenny (Eds.), The APSAC handbook on child maltreatment (2nd ed., pp. 175–204). Newbury Park, CA: Sage Publications. Briere, J., & Elliott, D. M. (2003). Prevalence and psychological sequelae of self-reported childhood physical and sexual abuse in a general population sample of men and women. Child Abuse & Neglect, 27, 1205–1222. Buss, A. H., & Perry, M. (1992). The aggression questionnaire. Journal of Personality and Social Psychology, 63, 452–459. Cain, N. M., Pincus, A. L., & Ansell, E. B. (2008). Narcissism at the crossroads: Phenotypic description of pathological narcissism across clinical theory, social/personality psychology, and psychiatric diagnosis. Clinical Psychology Review, 28, 638–656. Carlin, A. S., Kemper, K., Ward, N. G., Sowell, H., Gustafson, B., & Stevens, N. (1994). The effect of differences in objective and subjective definitions of childhood physical abuse on estimates of its incidence and relationship to psychopathology. Child Abuse & Neglect, 18, 393–399. Carter, S. D., Crabtree, N. L., Epps, J., & Roberts-Davis, A. M. (2014). Age of onset of physical abuse: Implications for adult anger and aggression. Universal Journal of Psychology, 2, 131–135. Chen, P., Coccaro, E. F., Lee, R., & Jacobson, K. C. (2012). Moderating effects of childhood maltreatment on associations between social information processing and adult aggression. Psychological Medicine, 42, 1293–1304. Claussen, A. H., & Crittenden, P. M. (1991). Physical and psychological maltreatment: Relations among types of maltreatment. Child Abuse & Neglect, 15, 5–18. Dickinson, K. A., & Pincus, A. L. (2003). Interpersonal analysis of grandiose and vulnerable narcissism. Journal of Personality Disorders, 17, 188–207. Edwards, V. J., Holden, G. W., Felitti, V. J., & Anda, R. F. (2003). Relationship between multiple forms of childhood maltreatment and adult mental health in community respondents: Results from the adverse childhood experiences study. American Journal of Psychiatry, 160, 1453–1460. Egeland, B., Yates, T., Appleyard, K., & van Dulmen, M. (2002). The long term consequences of maltreatment in the early years: A developmental pathway model to antisocial behavior. Children’s Services: Social Policy, Research, and Practice, 5, 249–260. Eisenberg, N. (2000). Emotion, regulation, and moral development. Annual Review of Psychology, 51, 665–697. Epps, J., Carlin, A. S., & Ward, N. G. (1999). Adult anger expression and childhood physical abuse history: The effects of event memory vs. self-labeling. Journal of Mental Health and Aging, 5, 175–185. Farmer, E., & Andrews, B. (2009). Shameless yet angry: Shame and its relationship to anger in male young offenders and undergraduate controls. Journal of Forensic Psychiatry & Psychology, 20, 48–65. Gold, J., Sullivan, M. W., & Lewis, M. (2011). The relation between abuse and violent delinquency: The conversion of shame to blame in juvenile offenders. Child Abuse & Neglect, 35, 459–467. Gratz, K. L., Paulson, A., Jakupcak, M., & Tull, M. T. (2009). Exploring the relationship between childhood maltreatment and intimate partner abuse: Gender differences in the mediating role of emotion dysregulation. Violence & Victims, 24, 68–82. Hayes, A. F. (2009). Beyond Baron and Kenny: Statistical mediation analysis in the new millennium. Communication Monographs, 76, 408–420. Hejdenberg, J., & Andrews, B. (2011). The relationship between shame and different types of anger: A theory-based investigation. Personality and Individual Differences, 50, 1278–1282. Hendin, H. M., & Cheek, J. M. (1997). Assessing hypersensitive narcissism: A reexamination of Murray’s Narcism Scale. Journal of Research in Personality, 31, 588–599. Lansford, J. E., Miller-Johnson, S., Berlin, L. J., Dodge, K. A., Bates, J. E., & Pettit, G. S. (2007). Early physical abuse and later violent delinquency: A prospective longitudinal study. Child Maltreatment, 12, 233–245. Lemerise, E. A., & Arsenio, W. F. (2000). An integrated model of emotion processes and cognition in social information processing. Child Development, 71, 107–118. Malatesta, C. Z., & Wilson, A. (1988). Emotion cognition interaction in personality development: A discrete emotions, functionalist analysis. British Journal of Social Psychology, 27, 91–112. Malkin, M. L., Zeigler-Hill, V., Barry, C. T., & Southard, A. C. (2012). The view from the looking glass: How are narcissistic individuals perceived by others? Journal of Personality, 81, 1–15. Maneta, E., Cohen, S., Schulz, M., & Waldinger, R. J. (2012). Links between childhood physical abuse and intimate partner aggression: The mediating role of anger expression. Violence and Victims, 27, 315–328. Miller, J. D., & Campbell, W. K. (2008). Comparing clinical and social-personality conceptualizations of narcissism. Journal of Personality, 76, 449–476. Miller, J. D., Dir, A., Gentile, B., Wilson, L., Pryor, L. R., & Campbell, W. K. (2010). Searching for a vulnerable dark triad: Comparing factor 2 psychopathy, vulnerable narcissism, and borderline personality disorder. Journal of Personality, 78, 1529–1564. Miller, J. D., Hoffman, B. J., Gaughan, E. T., Gentile, B., Maples, J., & Campbell, W. K. (2011). Grandiose and vulnerable narcissism: A nomological network analysis. Journal of Personality, 79, 1013–1042. Norman, R. E., Byambaa, M., De, R., Butchart, A., Scott, J., & Vos, T. (2012). The long-term health consequences of child physical abuse, emotional abuse, and neglect: A systematic review and meta-analysis. PLoS Medicine, 9, e1001349.
A.C. Keene, J. Epps / Child Abuse & Neglect 51 (2016) 276–283
283
Pincus, A. L., Ansell, E. B., Pimentel, C. A., Cain, N. M., Wright, A. G. C., & Levy, K. N. (2009). Initial construction and validation of the pathological narcissism inventory. Psychological Assessment, 21, 365–379. Pincus, A. L., & Lukowitsky, M. R. (2010). Pathological narcissism and narcissistic personality disorder. Annual Review of Clinical Psychology, 6, 421–446. Preacher, K. J., & Hayes, A. F. (2008). Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models. Behavior Research Methods, 40, 879–891. Ritter, K., Vater, A., Rüsch, N., Schröder-Abé, M., Schütz, A., & Fydrich, T. (2013). Shame in patients with narcissistic personality disorder. Psychiatry Research, 215, 429–437. Roche, M. J., Pincus, A. L., Lukowitsky, M. R., Ménard, K. S., & Conroy, D. E. (2013). An integrative approach to the assessment of narcissism. Journal of Personality Assessment, 95, 237–248. Ronningstam, E. F. (2010). Narcissistic personality disorder: A current review. Current Psychiatry Review, 12, 68–75. Scarpa, A., Haden, S. C., & Abercromby, J. M. (2010). Pathways linking child physical abuse, depression, and aggressiveness across genders. Journal of Aggression, Maltreatment, & Trauma, 19, 757–776. Schoenleber, M., & Berenbaum, H. (2012). Shame regulation in personality pathology. Journal of Abnormal Psychology, 121, 433–446. Shields, A., & Cicchetti, D. (1997). Emotion regulation among school-age children: The development and validation of a new criterion q-sort scale. Developmental Psychology, 33, 906–916. Spielberger, C. D. (1979). Preliminary manual for the State-Trait Personality Inventory. Tampa, Florida: Unpublished manual, University of South Florida. Springer, K. W., Sheridan, J., Kuo, D., & Carnes, M. (2007). Long-term physical and mental health consequences of childhood physical abuse: Results from a large population-based sample of men and women. Child Abuse & Neglect, 31, 517–530. Stevens, N. R., Gerhart, J., Goldsmith, R. E., Heath, N. M., Chesney, S. A., & Hobfoll, S. E. (2013). Emotion regulation difficulties, low social support, and interpersonal violence mediate the link between childhood abuse and posttraumatic stress symptoms. Behavior Therapy, 44, 152–161. Tangney, J. P., Wagner, P., Fletcher, C., & Gramzow, R. (1992). Shamed into anger? The relation of shame and guilt to anger and self-reported aggression. Journal of Personality and Social Psychology, 62, 669–675. Tangney, J. P., Wagner, P., Hill-Barlow, D., Marschall, D. E., & Gramzow, R. (1996). Relation of shame and guilt to constructive versus destructive responses to anger across the lifespan. Journal of Personality and Social Psychology, 70, 707–809. Teisl, M., & Cicchetti, D. (2008). Physical abuse, cognitive and emotional processes, and aggressive/disruptive behavior problems. Social Development, 17, 1–23. Thomaes, S., Bushman, B. J., Stegge, H., & Olthof, T. (2008). Trumping shame by blasts of noise: Narcissism, self-esteem, shame, and aggression in young adolescents. Child Development, 79, 1792–1801. Thomaes, S., Stegge, H., Olthof, T., Bushman, B. J., & Nezlek, J. B. (2011). Turning shame inside-out: “Humiliated fury” in young adolescents. Emotion, 11, 786–793. Verona, E., & Sachs-Ericsson, N. (2005). The intergenerational transmission of externalizing behaviors in adult participants: The mediating role of childhood abuse. Journal of Consulting and Clinical Psychology, 73, 1135–1145. Wright, A. G. C., Lukowitsky, M. R., Pincus, A. L., & Conroy, D. E. (2010). The higher order factor structure and gender invariance of the Pathological Narcissism Inventory. Assessment, 17, 467–483. Wright, A. G. C., Pincus, A. L., Thomas, K. M., Hopwood, C. J., Markon, K. E., & Krueger, R. F. (2013). Conceptions of narcissism and the DSM-5 pathological personality traits. Assessment, 20, 339–352. Zeigler-Hill, V., Green, B. A., Arnau, R. C., Sisemore, T. B., & Meyers, E. M. (2011). Trouble ahead, trouble behind: Narcissism and early maladaptive schemas. Journal of Behavior Therapy and Experimental Psychiatry, 42, 96–103.
