Attachment theory: A framework for understanding symptoms and interpersonal relationships in psychosis Katherine Berry * , Christine Barrowclough, Alison Wearden School of Psychological Sciences, University of Manchester, 2nd Floor, Zochonis Building, Brunswick Street, Manchester M13 9PL, UK article info Article history: Received 25 March 2008 Received in revised form 26 August 2008 Accepted 29 August 2008 Keywords: Psychosis Schizophrenia Attachment Symptoms Interpersonal abstract We investigated associations between adult attachment, symptoms and interpersonal functioning, including therapeutic relationships in 96 patients with psychosis. Using a prospective design, we also assessed changes in attachment in both psychiatrically unstable and stable groups. We measured attachment using the Psychosis Attachment Measure (PAM) and interpersonal problems and therapeutic relationships were assessed from both psychiatric staff and patient perspectives. Avoidant attachment was associated with positive symptoms, negative symptoms and paranoia. Attachment ratings were relatively stable over time, although changes in attachment anxiety were positively correlated with changes in symptoms. Predicted associations between high levels of attachment anxiety and avoidance and interpersonal problems were supported, and attachment avoidance was associated with difficulties in therapeutic relationships. Findings suggest that adult attachment style is a meaningful individual difference variable in people with psychosis and may be an important predictor of symptoms, inter- personal problems and difficulties in therapeutic relationships over and above severity of illness. Ó 2008 Elsevier Ltd. All rights reserved. Introduction Attachment theory is a lifespan developmental theory that proposes that there is a universal need to form close affectional bonds and that attachment behaviour functions as a homeostatic mechanism for modulating distress in adulthood as well as in childhood (Bowlby, 1980). The theory postulates that earlier inter- personal experiences influence future interpersonal functioning and methods of regulating distress via ‘working models’, or representations about the self and others in relationships. If care- givers are responsive and sensitive to distress the individual develops a secure attachment style, which is associated with a positive self-image, a capacity to manage distress, comfort with autonomy and in forming relationships with others. Conversely, if caregivers are insensitive or unresponsive to distress, the individual either escalates levels of distress to get their attachment needs met (insecure anxious or ambivalent attachment) or deactivates their attachment system which is associated with low levels of affect and an avoidance of close relationships (insecure avoidant attachment) (Shaver & Mikulincer, 2002). Although there are different ways of conceptualising adult attachment, Brennan, Clarke, and Shaver’s (1998) factor analysis of more than 320 self-report measures administered to a large sample of respondents revealed two dimensions of attachment. Attach- ment anxiety is associated with a negative self-image and an overly demanding interpersonal style, coupled with a fear of rejection and high levels of negative affect. Attachment avoidance is associated with a negative image of others, defensive minimisation of affect, interpersonal hostility and social withdrawal (Bartholomew & Horowitz, 1991; Mikulincer, Shaver, & Pereg, 2003). Empirical research has also shown associations between insecure adult attachments and a wide range of psychiatric disorders (Dozier, Stovall, & Albus, 1999) and there is evidence from longitudinal research to suggest that insecure attachment styles predict the onset of psychiatric symptoms in high-risk samples (Bifulco et al., 2006). Psychosis is a significant mental health problem and is charac- terised by high levels of interpersonal difficulties (Penn et al., 2004). The attachment system is likely to be particularly important in psychosis, as it is triggered by and determines individuals’ approaches to seeking help during periods of psychological stress, and psychotic experiences are often highly distressing (Bendall, McGorry, & Krstev, 2006). There is increasing recognition of the role of interpersonal factors in predicting the course of psychosis as well as influencing vulnerability (Read, van Os, Morrison, & Ross, 2005). Cognitive models propose that for some individuals difficulties in earlier relationships with significant others and interpersonal traumas lead to the formation of negative beliefs about the self and others, such as ‘I’m vulnerable’ and ‘other people are * Corresponding author. Present address: Clinical Psychology Unit, Department of Psychology, University of Sheffield, Western Bank, Sheffield S10 2TP, UK. Tel.: þ44 (0) 114 2226632; fax: þ44 (0) 114 2226610. E-mail address: katherinelberry@yahoo.co.uk (K. Berry). Contents lists available at ScienceDirect Behaviour Research and Therapy journal homepage: www.elsevier.com/locate/brat 0005-7967/$ – see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.brat.2008.08.009 Behaviour Research and Therapy 46 (2008) 1275–1282