Our analyses indicate that characteristics of the parent and of the parent-patient dyad seem to be the most important determinants of EOI. EOI is probably not linked to psychotic relapse, but rather to affective disturbances in the patient.

Relatives' criticism and hostility are important risk factors for relapse in schizophrenia. In order to explain these attitudes, we examined a Norwegian sample of 47 recently hospitalized patients (with schizophrenia or related psychoses) and 72 relatives. Relatives' expressed emotion was assessed by means of the Camberwell Family Interview. Demographic and clinical data were used as predictor variables in confirmatory regression analyses. The most robust predictors of high levels of criticism were, on the part of the patient, lack of paid employment, more than 3 previous hospital admissions, more troublesome behaviours reported by relatives, especially anxiety/depression, and better cognitive functioning at admission. Robust predictors of hostility were lack of employment and more than 3 previous hospitalizations. Interventions to reduce criticism should include employing patients and working with relatives' unrealistic expectations.

Despite the importance of relatives' emotional warmth for outcome in schizophrenia, no studies to date have addressed demographic and clinical predictors of warmth. We examined a Norwegian sample of 47 recently hospitalized patients (with schizophrenia or schizophreniform disorder) and 72 key relatives. Relatives' emotional warmth was assessed by means of the Camberwell Family Interview. Regression analyses showed that no substance abuse (especially amphetamines), better premorbid adjustment (12-15 years), a chronic social security status, and the relative not being a parent were the strongest predictors of emotional warmth. Emotional warmth was not related to patients' symptoms.

The expressed emotion (EE) index may not be as stable as it was once believed to be. The aim of this study was to identify variables associated with spontaneous change from low to high and from high to low levels of EE and EE subscales--critical comments (CC), hostility (H), emotional overinvolvement (EOI). Using a longitudinal, prospective study design, of 59 relatives having at least weekly face-to-face contact with 40 patients with an acute episode or relapse into schizophrenia (DSM-III-R) were interviewed by means of the Camberwell Family Interview (CFI) at admission and at 4 1/2 months after discharge. The results showed that high-high or unstable levels of CC, H or EE were associated with the patient not working or studying prior to admission. Relatives with low-high and high-high EOI patterns had more weekly face-to-face contact with the patient prior to admission than relatives with a low-low EOI pattern. Patients whose relatives had low-high CC and EE patterns were less ill at admission than patients whose relatives had low-low patterns. Higher perceived family burden was associated with, at admission, an unstable pattern of CC, and at follow-up, high-high EOI or EE patterns rather than low-low patterns. Our study suggests that it is possible to identify which relatives will have a stable and which a changing EE level, allowing for more focused intervention.

This study indicates that LOC beliefs may be determinants of emotional overinvolvement and criticism, and should be taken into account in family work that aims at modifying relatives' EE.