Expressed emotion in an 8 year follow-up

Caregiving processes and expressed emotion in psychosis, a cross-cultural, meta-analytic review

The construct of Expressed Emotion (EE) is a reliable predictor of relapse in psychotic disorders globally. However, cultural differences in the level and manifestation of EE have been reported. This review was conducted in line with PRISMA guidelines to demonstrate the distribution of EE and its domains cross-culturally as well as its relationship with relapse in psychosis. Ninety-six studies reported global EE scores and/or separate EE domains amongst caregivers of a family member with psychosis and used the Camberwell Family Interview (CFI) to measure EE. In the meta-analysis (k = 34, n = 1982), exposure to high EE was indicative of a 95% increased likelihood of relapse compared to low EE. However, no significant effect of geographical region on global EE scores (high/low) or EE domains was found. Several adjustments to the scoring of the CFI were highlighted based on cultural norms, particularly relevant to the domains of emotional over-involvement, warmth and criticism. Although this made meaningful quantitative comparisons across studies difficult, it nonetheless highlighted cultural considerations that need to be taken into account when interpreting EE and understanding its relationship to clinical outcomes. There is not a universal normative EE experience, with cultural variation in the scoring and interpretation of EE existing as evidenced by adjusted cut off scores and conceptualisation of EE constructs. Thus, it is important for clinical practitioners to have an awareness of different cultural norms in relation to caregiving and care receiving behaviours, which can inform adaptations to clinical interventions in multicultural settings.

The predictive validity of expressed emotions (EE) in schizophrenia. A 20-year prospective study

BACKGROUND

The level of expressed emotions in the family (EE) and components thereof (emotional over-involvement EOI, and critical comments CC) have been found to be related to treatment outcomes in psychotic patients. One point of interest is the dynamic of various outcomes in relation to initial EE, CC and EOI levels.

METHOD

43 individuals with diagnosis of schizophrenia were assessed at 1, 3, 7, 12 and 20-year follow-ups. The EE indicator was evaluated during the index hospitalization. Outcome indicators included: number of relapses, number and duration of in-patient rehospitalizations, and severity of psychopathological symptoms.

RESULTS

1) A high EE indicator was associated with a higher number of relapses in every follow-up and differences between the groups with low and high EE indicators proved significant after 3, 7, 12 and 20 years of the illness. 2) A high EE indicator was associated with a higher number of rehospitalizations after 3 and 7 years, and showed as a clear trend after 12 and 20 years. 3) A high level of CC was associated with a higher number of relapses. 4) The severity of positive symptoms increased only in the group with high CC. 5) The negative syndrome was not associated with EE, EOI or CC.

CONCLUSIONS

The EE indicator may be considered a valid prolonged predictor of relapses and rehospitalizations. A high level of CC was associated with a higher number of relapses and intensification of the positive syndrome in both the short-term and long-term course of the illness.

Expressed emotion is not associated with disorder severity in first-episode mental disorder

A family atmosphere characterized by expressed emotion (EE) is a robust predictor of clinical outcome of patients with schizophrenia and mood disorders. However, there is ongoing discussion as to whether EE is more a cause of clinical outcome or a parental reaction to disorder severity. This cross-sectional study examines a sample of 42 consecutive first-episode patients from a defined geographical area with severe mental disorders (schizophrenia-related disorders, psychotic mood disorders, and non-psychotic mood disorders). Their 42 relatives were interviewed, and the relationships between EE variables derived with the five-minute speech sample method (FMSS) and the patients' demographic, premorbid and clinical measures were analyzed. A high EE score was found in 40% of the relatives. High EE was associated with the interviewed relative's not being a spouse and the patient's being young and unmarried. It was not associated with premorbid characteristics, symptom dimensions or the diagnostic group of the patient. These results do not support the hypothesis that EE is a reaction to the clinical features of the patient. Instead, demographic factors may partly mediate the effect of EE on prognosis.