Childhood Emotional Abuse and Attachment Processes in the Dyadic Adjustment of Dating Couples Shelley A. Riggs, Angela M. Cusimano, and Karen M. Benson University of North Texas In an effort to improve understanding of the mechanisms that link early maltreatment to later outcomes, this study investigated the mediation effects of adult attachment processes on the association between childhood emotional abuse and later romantic relationships among heterosexual couples. College students and their dating partners (N 310 155 couples) completed the Childhood Trauma Question- naire, Experiences in Close Relationship Scale, and Dyadic Adjustment Scale. Using the Actor���Partner Interdependence Model (Kenny, Kashy, & Cook, 2006), multilevel modeling results indicated that memories of childhood emotional abuse reported by both students and their partners were significantly associated with attachment strategies, as well as romantic relationship quality. Findings supported hypothesized mediation effects of attachment anxiety and avoidance. Keywords: emotional abuse, adult attachment, romantic relationship, actor���partner effects Only recently has emotional abuse1 been acknowledged as a distinct form of child maltreatment, which often accompanies physical or sexual abuse but also frequently occurs in isolation, making it the most pervasive and chronic type of maltreatment (Davis, Petretic-Jackson, & Ting, 2001 Gauthier, Stollak, Messe ��, & Aronoff, 1996). Emotional maltreatment in childhood may engender psychological disorganization (Cassidy & Mohr, 2001) that contributes to emotional, cognitive, and relational disturbances in later life (Bifulco, Moran, Baines, Bunn, & Stanford, 2002 Ferguson & Dacey, 1997 Messman-Moore & Coates, 2007). Consequently, a substantial number of individ- uals with histories of childhood emotional abuse may present for psychotherapy at some point in their lives. Yet research on childhood emotional abuse and later adult adjustment is rela- tively limited, particularly with respect to mediating or moder- ating processes (Wright, 2007). Attachment theory offers a useful, empirically supported framework for understanding out- comes of child maltreatment (Cassidy & Mohr, 2001 Riggs, 2010). Guided by a developmental-systemic perspective, the current study was designed to test a theoretical model suggest- ing that the attachment system represents a key mediational link between childhood emotional abuse and romantic relationship outcomes (Riggs, 2010). In addition to testing direct and me- diation paths, the study examined mutual influences between dating partners and investigated the moderating effects of sex on associations between childhood emotional abuse and rela- tionship adjustment in heterosexual dating couples. The Attachment System A long-lived idea in the developmental and clinical litera- tures is the notion that childhood experiences in the family of origin affect adults��� romantic relationships (W. A. Collins & Sroufe, 1999 Dattilio, 2006). In the context of early attachment relationships with parents, children learn about mutual trust and about receiving and giving care, as well as negotiating close- ness/distance and power. It is also within this first relational bond that the capacity for empathy and intimacy emerges (W. A. Collins & Sroufe, 1999). Individuals who experience emotionally abusive attachment relationships in childhood are at a distinct disadvantage in interpersonal contexts because they develop a distorted understanding of what loving and caring relationships involve (Riggs, 2010). To effectively treat clients with histories of emotional abuse, it is important for counselors to understand that maltreatment undermines the adaptive devel- opment of the attachment system (N. L. Collins, Guichard, Ford, & Feeney, 2004 Riggs, 2010). Developmental Outcomes of Childhood Emotional Abuse On the basis of a review of existing theory and research, Riggs (2010) outlined a conceptual model that traces the developmental outcomes of childhood emotional abuse and identifies the attach- ment system as a principal conduit to adult functioning in close 1 Although the terms and definitions vary in the literature, these differ- ences are beyond the scope of this article. For our purposes, the terms emotional abuse and psychological abuse or maltreatment are used inter- changeably. This article was published Online First December 13, 2010. Shelley A. Riggs, Angela M. Cusimano, and Karen M. Benson, Depart- ment of Psychology, University of North Texas. We wish express our great appreciation to David Kenny for excellent training at the DATIC workshop and for his continued generosity in providing ongoing statistical consultation for this study. We also acknowl- edge Mike Devine for his assistance with the online data collection, as well as Zane Dodd and Sara Pollard for data management support. Correspondence concerning this article should be addressed to Shelley A. Riggs, Department of Psychology, University of North Texas, 1155 Union Circle #311280, Denton, TX 76203-5017. E-mail: riggs@unt.edu Journal of Counseling Psychology �� 2010 American Psychological Association 2011, Vol. 58, No. 1, 126���138 0022-0167/10/$12.00 DOI: 10.1037/a0021319 126

relationships. The first assumption of the model is that emotional abuse by attachment figures in infancy and early childhood con- tributes to insecure attachment. This premise is supported by research that consistently implicates insensitive parenting charac- terized by rejection, intrusiveness, or inconsistency as a major contributor to insecurity of attachment manifested in insecure- avoidant or insecure-ambivalent/resistant behavioral strategies (DeWolff & van IJzendoorn, 1997). More extreme forms of par- enting that would be considered emotionally abusive contribute to disorganized/disoriented infant attachment. For example, Lyons- Ruth and colleagues (e.g., Lyons-Ruth, Melnick, Bronfman, Sherry, & Llanas, 2004) reported significant links between disor- ganized infant attachment and hostile parenting marked by antip- athy and harsh rejection, excessive intrusiveness/control, frighten- ing or threatening actions, or an incongruous mixture of these behaviors that both elicit and reject the infant���s attachment-related emotions and behaviors. Children���s attachment insecurity then creates a vulnerability to later dysfunction because it impedes emotional regulation and engenders negative internal working models of self and others. Specifically, when an abusive caregiver is simultaneously ���the source of and the solution to its alarm��� (Main & Hesse, 1990, p. 163), a child is unable to resolve fear and reduce anxiety, leading to chronic activation of the fight-or-flight mechanism and a dys- regulation of arousal (Lyons-Ruth et al., 2004). As a result, chil- dren do not learn how to manage their emotional states effectively and develop negative views of the self as inept, damaged, and contemptible, as well as negative views of others as hurtful, betraying, and cruel (Riggs, 2010). This is consistent with research indicating that early emotional abuse creates substantial risk for poor self-concept, low self-esteem, and disorders of emotional regulation and impulse control (Mullen, Martin, Anderson, Romans, & Herbison, 1996 Showers, Zeigler-Hill, & Limke, 2006 Wenar & Kerig, 2000 Widom, Kahn, Kaplow, Sepulveda- Kozakowski, & Wilson, 2007). According to the model (Riggs, 2010), these deficits foster ineffective coping responses and maladaptive interpersonal schemas that subsequently disrupt peer relations and contribute to insecurity in the adult attachment system. The model further suggests that all of these variables (e.g., emotional dysregula- tion, negative internal models, maladaptive coping, and inse- cure adult attachment) may function as mediators of childhood emotional abuse and mental health, which in turn may mediate their associations with romantic relationship functioning. This developmental trajectory is suggested by research documenting direct associations between various forms of psychological dysfunction (e.g., depression, anxiety, eating disorder, dissoci- ation, aggression, and suicidal behavior) and both childhood emotional abuse (Bifulco et al. 2002 Ferguson & Dacey, 1997 Morimoto & Sharma, 2004 Mullen et al., 1996) and insecure attachment (for reviews, see Sroufe, Carlson, Levy, & Egeland, 1999, and Riggs et al., 2007). In addition, retrospective evi- dence indicates that memories of childhood emotional abuse are significantly related to poor interpersonal skills, insecure at- tachment style, and relationship dysfunction in adulthood (Davis et al., 2001 Messman-Moore & Coates, 2007 Varia & Abidin, 1999). Adult Attachment Internal working models are dynamic representations that en- able individuals to predict and interpret a partner���s behavior and to plan their own response (W. A. Collins & Sroufe, 1999). Due to their influence on selective attention and encoding and other cognitive-affective processes, internal working models function as one mechanism through which the attachment system maintains continuity when the environment remains stable from infancy through adulthood (Cassidy, 2000 N. L. Collins et al., 2004). Unlike infant attachment, however, the adult romantic attachment system interacts with the sexual mating system and is symmetrical because caregiving is reciprocal whereby each partner looks to the other as a source of security (Zeifman & Hazan, 2008). The reciprocity of the romantic attachment bond implies that each partner affects the other and must be considered when addressing relationship problems in therapy. Adult romantic attachment varies along two dimensions of attachment anxiety and avoidance that traditionally have been thought to reflect internal working models of self and other, respectively (Bartholomew & Horowitz, 1991 Brennan, Clark, & Shaver, 1998). Although this correspondence has been debated recently (Mikulincer & Shaver, 2007), for the purposes of this study, the behavioral strategies of attachment anxiety and avoid- ance are viewed as surface indicators of the underlying internal working models of self and other (Shaver & Mikulincer, 2002). With positive internal working models, secure adults experience low levels of attachment anxiety or avoidance and generally em- ploy adaptive coping skills, such as help seeking (Lopez, Melen- dez, Sauer, Berger, & Wyssmann, 1998 Riggs, Jacobvitz, & Hazen, 2002). In contrast, high levels of attachment anxiety (neg- ative self model) are associated with emotion-focused coping and a hyperactivating strategy that exaggerates emotions and help- seeking behaviors, whereas high levels of attachment avoidance (negative other model) are associated with instrumental coping and a deactivating strategy that minimizes the importance of emotions and close relationships (Mikulincer & Shaver, 2003). Romantic Relationships Attachment Research As one of the most well-supported and sophisticated theories of normal and abnormal development, attachment theory offers coun- selors insight into the history and current functioning of clients. The theory is especially salient when working with clients pre- senting with relationship issues. Ample evidence links adult at- tachment patterns to various aspects of romantic relationships, including dyadic functioning and satisfaction, the capacity for intimacy, partner caretaking and support, sexual behavior, conflict management, and relational aggression (Creasey, 2002 Dutton, Saunders, Starzomski, & Bartholomew, 1994 Gallo & Smith, 2001 Shi, 2003 Simpson, Winterheld, Rholes, & Orina, 2007). Generally, attachment security is associated with adaptive rela- tional functioning, whereas attachment insecurity is associated with less desirable, maladaptive behaviors in romantic relation- ships. Insecure adults are less likely to be supportive of their partners and more likely to be controlling (Creasey, 2002 B. C. 127 EMOTIONAL ABUSE, ATTACHMENT, DATING RELATIONSHIPS