Research report A population-based analysis of distinguishers of bipolar disorder from major depressive disorder Ayal Schaffer a, ⁎, John Cairney b , Scott Veldhuizen b , Paul Kurdyak b , Amy Cheung a,b , Anthony Levitt a a Mood and Anxiety Disorders Program, Department of Psychiatry, Sunnybrook Health Sciences Centre, University of Toronto, 2075 Bayview Avenue, Room FG 29 Toronto, Ontario, Canada M4N 3M5 b Health Systems Research and Consulting Unit, Centre for Addiction and Mental Health, Department of Psychiatry, University of Toronto, Toronto, Canada article info abstract Article history: Received 14 October 2009 Received in revised 11 January 2010 Accepted 16 February 2010 Available online 12 March 2010 Background: Many people with bipolar disorder (BD) in the community are misdiagnosed with major depressive disorder (MDD). A probabilistic model has been proposed to assist in the identification of BD among patients with depressive symptoms, however there are limited population-based data on the key distinguishers of BD from MDD. The objective of this study was to identify distinguishers of BD from MDD in a population-based sample. Methods: Population-based data were extracted from the Canadian Community Health Survey: Mental Health and Well-Being. Sociodemographic variables, clinical variables, and depressive symptomatology were compared between subjects with BD (N =467) and MDD (N = 4145). Logistic regression analysis was used to identify significant correlates of BD, and areas under the receiver operating characteristic curves (AUCs) were determined for each model. Results: BD and MDD subjects differed across a number of characteristics. Clinical variables significantly associated with BD included greater number of lifetime depressive episodes, earlier age of first depressive episode, lifetime anxiety disorder, problematic substance use, and lifetime suicide attempt. Symptoms significantly more common during a major depressive episode among BD subjects included agitation, suicidal ideation, anxious symptoms, and irritability. AUCs for these models ranged from 0.72 to 0.81. Limitations: Data were not available for all potential distinguishers; subgroups of BD could not be determined; cross-sectional data. Conclusions: These population-based results reinforce the effort to establish a generalizable probabilistic model that incorporates clinical and symptom variables in order to assist clinicians in the diagnostic assessment of BD. © 2010 Elsevier B.V. All rights reserved. Keywords: Bipolar disorder Major depressive disorder Epidemiology 1. Introduction Modern diagnostic systems such as DSM-IV and ICD-10 classify bipolar disorder (BD) and major depressive disorder (MDD) as distinct illnesses with overlapping clinical presenta- tions. Both illnesses are frequently characterized by predominant depressive symptoms (Joffe et al., 2004; Judd et al., 1998), and undiagnosed BD patients commonly present as outpatients with complaints of depression (Das et al., 2005; Hirschfeld et al., 2003b). In this clinical circumstance, there are challenges in establishing the accurate diagnosis of BD that stem from the time-consuming and often imprecise nature of retrospective identification of hypomanic or manic symptoms (Manning et al., 1998; Perlis, 2005; Yen et al., 2007). Many people with BD in the community are therefore misdiagnosed, most commonly as having MDD (Baldessarini et al., 1999; Calabrese et al., 2006; Das et al., 2005; Lish et al., 1994; Mantere et al., 2008; Perlis, 2005). There are significant individual and public health ramifi- cations to the misdiagnosis of BD (Hirschfeld et al., 2003b). Journal of Affective Disorders 125 (2010) 103–110 ⁎ Corresponding author. Tel.: + 1 416 480 4070; fax: +1 416 480 6878. E-mail address: ayal.schaffer@sunnybrook.ca (A. Schaffer). 0165-0327/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.jad.2010.02.118 Contents lists available at ScienceDirect Journal of Affective Disorders journal homepage: www.elsevier.com/locate/jad