ImpactMessagesofDepressedOutpatientsasPerceivedbyTheirSignificant Others: Profiles, Therapeutic Change, and Relationship to Outcome Martin Grosse Holtforth, 1 David Altenstein, 1 Emily Ansell, 2 Claudia Schneider, 3 and Franz Caspar 4 1 University of Zurich 2 Yale University School of Medicine 3 Friedrich-Schiller-University, Jena 4 University of Bern Whereas previous interpersonal research in depression has frequently used self-reports, patients’ impact on others is rarely analyzed. We analyzed impacts of 180 depressed psychotherapy outpatients out of 832 diagnostically heterogeneous patients as rated by their significant others. Depressed patients were perceived as more submissive, hostile-submissive, and friendly-submissive, and as less dominant and friendly-dominant than patients with other principle disorders. After therapy, the 59 depressed patients whose significant others also provided data after treatment were perceived as less submissive (friendly-submissive, submissive, hostile-submissive) and more dominant and friendly- dominant. Whereas a decrease in submissiveness and hostile-submissiveness was associated with positive outcomes, decrease in friendly-submissiveness was unrelated. Cluster analyses suggested four distinct interpersonal subgroups. We discuss these results in terms of interpersonal theory and interpersonal assessment in depression therapy. 2012. Keywords: interpersonal; psychotherapy; impact message; personality; depression Depressive disorders are regularly associated with interactional problems (Horowitz, Rosenberg, Baer, Uren˜o, & Villasen˜or, 1988; Joiner & Timmons, 2009) and the relationship between interpersonal behaviors and psychopathology can generally be assumed to be bidirectional: Problematic interpersonal behavior may contribute to psychopathology, and psychological disorders may maintain or even strengthen interpersonal problems (Barrett & Barber, 2007; Cain, Pincus, & Grosse Holtforth, 2010; Pincus & Wright, 2010). Several authors have formulated theoretical expectations about the interpersonal characteristics of depressed people, which lie most frequently in the submissive range (Coyne, 1976; Hammen, 2006; Joiner, 2000; McCullough, 2003). Accordingly, interpersonal behaviors that have been empirically associated with a risk for depression include insecure attachment (Eberhardt & Hammen, 2006), shyness and social withdrawal (Alfano, Joiner, & Perry, 1994), lack of assertiveness (Ball, Otto, Pollack, & Rosenbaum, 1994), or excessive reassurance seeking (Joiner & Metalsky, 2001). However, there is relatively little empirical research on interpersonal patterns in depression (Barrett & Barber, 2007). Empirical research on interpersonal factors in psychopathology commonly uses methods developed within the interpersonal tradition (Pincus, Lukowitsky, & Wright, 2010). The central assumption is that all interpersonal behavior can be accounted for by a combination of two central descriptive dimensions: agency and communion, which span a two-dimensional space referred to as the interpersonal circumplex (IPC; Alden, Wiggins, & Pincus, 1990; Leary, 1957). The interpersonal space is commonly subdivided into eight ‘‘subspaces’’ (octants) representing combinations of agency and communion, such as hostile-submissive or friendly dominant (Pincus & Gurtman, 2006). Most interpersonal Correspondence concerning this article should be addressed to: Dr. Martin Grosse Holtforth, Gesellschaftsstrasse 49, Bern 3012; e-mail: grosse@psychologie.uzh.ch & 2011 Wiley Periodicals, Inc. Published online in Wiley Online Library (wileyonlinelibrary.com/journal/jclp). DOI: 10.1002/jclp.20854 – 319 333, & 2011 Wiley Periodicals, Inc. J. Clin. Psychol. 68: JOURNAL OF CLINICAL PSYCHOLOGY, Vol. 68(3), 319--333 (2012)