Expressed emotion of families and the course of mood disorders: a cohort study in Japan

• Published in: Journal of affective disorders 2001-03, Vol.63 (1), p.43-49
• Main Authors: Mino, Yoshio, Shimodera, Shinji, Inoue, Shimpei, Fujita, Hirokazu, Tanaka, Shuichi, Kanazawa, Susumu
• Format: Article
• Language: English
• Subjects: Adult; Adult and adolescent clinical studies; Aged; Biological and medical sciences; Cohort Studies; Depression; Expressed Emotion; Family; Family Health; Female; Humans; Japan - ethnology; Male; Medical sciences; Middle Aged; Miscellaneous; Mood disorder; Mood disorders; Mood Disorders - ethnology; Mood Disorders - psychology; Predictive Value of Tests; Psychiatric Status Rating Scales; Psychology. Psychoanalysis. Psychiatry; Psychopathology. Psychiatry; Recurrence; Risk Factors; Tropical medicine
• ISSN: 0165-0327; 1573-2517
• DOI: 10.1016/S0165-0327(00)00185-3

Background: Findings on expressed emotion (EE) of the family and the course of mood disorders have not been consistent. There has also been no report on these problems from Asia. Methods: The subjects were 32 patients diagnosed to have mood disorders on the basis of DSM-IV and ICD-10 and 36 principal members of their families. EE was evaluated using Camberwell Family Interview (CFI). A cohort study was conducted for 9 months after discharge of the patients. The patients were divided into a high-EE group and a low-EE group using a few cut-off points concerning the number of critical comments (CCs) and emotional over-involvement (EOI), the 9-month relapse risks were compared, and the relapse risk ratio and its 95% confidence interval were calculated. Also, the sensitivity, specificity, positive predictive value, and negative predictive value were calculated, and the optimal cut-off point was evaluated. Multiple logistic regression analysis was performed to control the effects of potential confounding factors. Results: When the subjects with three or more CCs, or an EOI score of three or higher were regarded as a high-EE group, and the others as a low-EE group, the nine-month relapse risks were 83.3% (5/6) and 19.2% (5/26), respectively, and the relapse risk ratio (95% confidence interval) was 4.3 (1.8–12.2). The values of the validity parameters were the highest with these cut-off points. The effect of EE was also significant on multiple logistic regression analysis. Limitations: Small sample size and severity of the patients. Conclusions: Families’ EE based on CFI correlated with relapse also in Japan.