Full Text

Indexes of expressed emotion (EE) in 58 relatives of patients with schizophrenia were related to those relatives' spontaneously expressed causal beliefs about the illness and about related symptoms and behaviors. Relatives made attributions predominantly to factors external, universal, and uncontrollable from their own perspective, and to factors internal, universal, and uncontrollable from the patient's perspective. Low-EE relatives were similar in their attributions to emotionally overinvolved relatives. Compared with these two groups, critical and/or hostile relatives made more attributions to factors personal to and controllable by the patient. Subsequent analyses suggested that hostile relatives were further characterized by making more attributions to factors internal to the patient and by making attributions with fewer causal elements.

Relapse from schizophrenia is associated with the presence of certain emotional characteristics in patients' relatives. A number of studies have shown that patients relapse more frequently when relatives are high in expressed emotion (EE), an index reflecting criticism, hostility, or emotional overinvolvement (e.g., Karno et al., 1987; Nuechterlein et al., 1986; Vaughn & Leff, 1976a; Vaughn, Snyder, Freeman, Jones, & Falloon, 1984). These findings have received further support from interventions aimed specifically at reducing expressed emotion, which on the whole have succeeded in correspondingly reducing relapse rates (e.g., Leff et al., 1989; Leff, Kuipers, Berkowitz, Eberlein-Vries, & Sturgeon, 1982). Less is known, however, about the origins and correlates of these emotional characteristics. This study was designed to test the hypothesis that relatives' emotions would be related to their beliefs about the causes of the patients' illness, behavior, and symptoms.

Causal beliefs are prominent and easily elicited from patients and accident victims (e.g., Brewin & Antaki, 1987) and, more generally, from anyone faced with an unexpected or unwelcome event (Weiner, 1985). Several authors have speculated that attributing symptoms to the patient rather than to the illness is related to high expressed emotion in relatives (