Predictors of expressed emotion in relatives of patients with bipolar disorder

Open Access
Antle, Jennifer Marie
Area of Honors:
Human Development and Family Studies
Bachelor of Science
Document Type:
Thesis Supervisors:
  • Steffany Jane Fredman, Thesis Supervisor
  • Charles Geier, Honors Advisor
  • Bipolar disorder
  • expressed emotion
  • emotional overinvolvement
  • criticism
  • predictors
The construct of expressed emotion (EE) refers to emotional attitudes held by the relatives of patients with psychiatric illness, with a focus on relatives’ criticism, hostility, and/or emotional overinvolvement (Leff & Vaughn, 1985). Emotional overinvolvement (EOI) refers to the relatives’ intrusiveness, self-sacrifice, and emotional distress related to the patient’s well being (Leff & Vaughn, 1985) and appears to represent a mix of relatives’ appropriate and inappropriate emotional engagement with patients (Fredman, Baucom, Miklowitz, & Stanton, 2008). EE is a robust predictor of relapse and generally poorer outcomes across a range of psychopathologies (Butzlaff & Hooley, 1998). To identify families at high risk for EE and associated family dysfunction, researchers have sought to identify patient characteristics that might predict relatives’ EE. The majority of these studies have been conducted using samples of individuals with schizophrenia and have not differentiated between relatives’ criticism and EOI when classifying a family as high versus low EE. The current study sought to examine patient predictors of relatives’ EE in the context of bipolar disorder, with the critical comments and EOI dimensions considered separately as assessed with the Camberwell Family Interview (Vaughn & Leff, 1976). The ability to identify patient characteristics that might be associated with relatives’ EE increases the possibility of identifying high-risk families and steering them toward family- based treatments for bipolar disorder in the hopes of improving patient outcomes. Using a treatment-seeking sample of patients with bipolar disorder (N = 115) and caregiving relatives (Fredman, Baucom, Miklowitz, & Stanton, 2008), we explored associations between clinician-rated patient symptom severity (mania, depression, psychosis), social functioning, and illness history variables (premorbid functioning, age at onset, number of years ill, number of episodes, number of hospitalizations), and relatives’ EOI and criticism. Results indicated that current symptom severity and functioning had small-to-medium size associations with EOI. Depressive and psychotic symptom severity were each positively and significantly associated with relatives’ EOI (rs. = 248 and .251, respectively; ps = .016 and .015 respectively), and current social functioning was negatively and significantly correlated with EOI (r = -.323, p = .002). Psychosis was also significantly inversely associated with relatives’ criticism (r = -.274, p = .008). Illness history variables, however, were not associated with EOI or criticism with the exception that the number of prior hospitalizations was negatively and significantly correlated with CC (r = -.221, p = .021). These findings suggest that the degree to which patients appear sick or impaired may be a salient characteristic in pulling for relatives’ emotional engagement with patients rather than criticism.