Expressed emotion. The family environment of a person with psychosis has been extensively measured using the concept of expressed emotion (EE). The methodology associated with this concept was initially used to investigate and describe the effects of different social environments after individuals with psychosis returned home from inpatient admissions (Xxxxx & Xxxxxx, 1996). In EE, the emotional aspects of close relationships are measured using prosodic variables, pitch, tone and emphasis (Xxxxxxx, 1979). EE measures key aspects of interpersonal relationships including criticism, hostility, warmth, positive comments, and emotional over-involvement (Xxxxxxx, Tarrier, Barrowclough, Zastowny, & Xxxxxx, 2000). Carers can be classified as being either high or low on EE. Carers described as high in EE typically express above threshold levels of critical, hostile and/or emotionally over-involved behaviours towards the individual with psychosis. Conversely, those below the threshold are rated as low EE. High EE in carers has been found to be a robust predictor of relapse in individuals with psychosis (Xxxxxxxxxx & Xxxxxxx, 1994; Xxxxxxxx & Xxxxxx, 1998). In a meta-analysis of 25 worldwide studies of EE and patient outcomes, individuals returning to live with high EE families after a hospital admission had a 50% relapse rate nine months later, compared with 22% in individuals returning to low EE families (Xxxxxxxxxx & Xxxxxxx, 1994). High EE is also predictive of outcome in other health and psychiatric conditions (see Xxxxxxx et al., 2000, for a review), such as diabetes (Xxxxxxx, Xxxx, Xxxxxxxxxxxx, Xxxxxxx, & Xxxxxx, 2006) and bipolar affective disorder (Xxx & Xxxxxxxxx, 2004). Given the level of involvement of carers in the care and treatment of service users with psychosis and the importance of EE to patient outcomes, EE remains an important concept to measure and understand.
Expressed emotion. Five Minute Speech Samples (FMSS; Xxxxxx et al., 1986) were transcribed and then rated by the author. Seventeen speech samples (55%) were then rated by the author, an independent research assistant and primary supervisor (Xxxxxxx Xxxxxxxx, XX), an experienced rater of EE and FMSS. Discussions were held between the three raters, initial disagreements discussed and a final decision reached regarding the ratings. A comparison in ratings between XX and the author for the double rated speech samples (n = 17) was made prior to discussions to agree on ratings. These double rated speech samples had 76% agreement of the EE classification; hence there was deemed to be good inter-rater reliability. 42% of participants (n = 13) were rated as being high EE on the FMSS. A detailed breakdown of categories and dimensions is presented in Tables 8 and 9. Table 8 Expressed emotion categories for participants excluding missing data for one case (n = 31) EE Category n (%) Low EE ratings Low 6 (19%) Borderline EOI 4 (13%) Borderline critical 7 (23%) Borderline EOI and borderline critical 1 (3%) High EE ratings Critical 7 (23%) Critical and borderline EOI 1 (3%) EOI 1 (3%) EOI and borderline critical 2 (6.5%) EOI and critical 2 (6.5%) Table 9 Expressed emotion dimensions as rated using the Five Minute Speech Sample excluding missing data for one case (n = 31) EE Dimension High EE n = 13 Low EE n = 18 Initial statement Positive 1 (8%) 3 (17%) Neutral 11 (85%) 15 (83%) Negative 1 (8%) 0 (0%) Relationship Positive 0 (0%) 5 (28%) Neutral 11 (85%) 13 (72%) Negative 2 (15%) 0 (0%) Critical comments 0 4 (31%) 17 (100%) ≥ 1 9 (69%) 0 (0%) Dissatisfaction Present 9 (69%) 8 (44%) Absent 4 (31%) 10 (56%) Emotional Overinvolvement Absent 8 (62%) 17 (100%) Present 5 (38%) 0 (0%) Positive comments 0 - 4 12 (92%) 14 (78%) ≥5 1 (8%) 4 (22%)
Expressed emotion. In considering the transmission of psychopathology in families, the construct of expressed emotion (EE) has been particularly important. EE is a method of measuring family members’ responses to an individual and refers to a family member’s expressed criticism, hostility and/or emotional over-involvement (EOI). EE is seen to reflect the emotional quality of interactions between individuals and is a general risk factor for a wide range of psychopathology as well as high maternal EOI showing association with anxiety disorders. Measures of EE are considered to be less prone to bias than self-report in tapping aspects of parenting known to be linked to childhood anxiety, such as an overinvolved or more specifically an overprotective parenting style (Xxxxxxxx and Xxxxxx 1998; Xxxxx, Xxxxxx et al. 2009). Such parenting styles are also associated with withdrawn and inhibited child behaviour (Xxx, Xxxxxxxxx et al. 2005) which can cause the child and parent to fall into a cycle where the inhibited behaviour of a child elicits more protective parenting which leads to the child being more inhibited and so on. This is an important area of research developmentally given that behavioural inhibition may precipitate other forms of psychopathology (Xxxxx, Xxxxx et al. 1995). Indeed one criticism that can be leveled at much of the EE literature is that it deals at a disorder level rather than examining the more complex effects on children that are relevant to the pathways that may lead to disorder, or phenomena seen in sub-clinical populations. Associations between high EOI and child anxiety disorders have also been found in an epidemiological sample (Xxxxxx, Xxxxxx et al. 1993). Other research has indicated a relationship between child anxiety and critical parenting (Xxx, Xxxxxxxxx et al.
