Identifying the socio-demographic and clinical determinants of family functioning in Greek patients with psychosis

Abstract

Background:

Studies on determinants affecting family functioning of patients with psychosis are still limited in Greece.

Aim:

The aim of this study was to describe the socio-demographic and clinical characteristics associated with family functioning in patients with schizophrenia and bipolar disorder in Crete, Greece.

Methods:

A total of 100 patients and their caregivers agreed to participate in the study. Family functioning was assessed in terms of cohesion, adaptability, communication and satisfaction dimensions (Family Adaptability and Cohesion Evaluation Scale IV Package), expressed emotion (Family Questionnaire), family burden (Family Burden Scale) and caregivers’ psychological distress (General Health Questionnaire-28). Multivariate linear regression models were implemented to examine the associations between each one of the family measures and different social and clinical characteristics.

Results:

With regard to the caregivers’ characteristics, gender, employment status, origin, residence, financial status, relation to the patient, contact with the patient and family structure were among the most significant determinants of family functioning. Also, patients’ socio-demographic characteristics, including age, education, origin, residence and employment status, as well as illness-related factors, such as onset of mental illness, number of hospitalisations, last hospitalisation, longer hospitalisation and clinical diagnosis impacted intrafamilial relationships.

Conclusion:

The results of this study suggest that a number of social and clinical factors contributed to the family environment of patients with psychosis. Identifying the determinants of family functioning in psychosis is instrumental in developing understandings regarding the factors which may contribute to the rehabilitation or relapse of the patient and the support required to strengthen positive family interactions.

Keywords

Family functioning expressed emotion family burden psychological distress socio-demographic determinants illness-related features

Introduction

The study of family in relation to schizophrenia has changed dramatically over the last five decades. In the 1960s and 1970s, the family was implicated in the aetiology of schizophrenia (Goldstein, 1988), and certain family structures and communication patterns were considered to be contributing factors to the development and the maintenance of the disease (Lam, 1991). However, with the trend of deinstitutionalisation of psychiatric patients and the simultaneous growth of community mental health services, the onus for the patient’s care has been transferred to family members (Bloch, Szmukler, Herrman, Benson, & Colussa, 1995), and the focus of both research and in service provision moved towards the study of intrafamilial relationships and how the family context can best support and care for patients with psychosis.

Most of the research on family functioning of patients with psychosis examined certain aspects of intrafamilial transactions. Two attributes that have been studied extensively are the following: (1) the affective attitudes and behaviours expressed to the patient from his or her family members, usually characterised as expressed emotion (EE), and (2) the burden of care. EE has been extensively researched across psychiatric disorders as the strongest family factor that influences the course of an illness (Butzlaff & Hooley, 1998; Cechnicki, Bielańskab, Hanuszkiewiczb, & Darenb, 2013; Hooley, 2007). EE constitutes the amount of relatives’ critical comments (CCs), hostility (H) and/or emotional over-involvement (EOI) towards the patient (Brown, 1985). CCs express dislike or disapproval of the patient’s behaviour, H reflects disapproval or rejection of the patient and EOI includes an exaggerated or overprotective attitude towards the patient, as demonstrated by an intrusive style of relating and the carer’s evident emotional distress. Family burden (FB) refers to the negative impact of the individual’s mental disorder on the entire family (Biegel & Schultz, 1999; Schene, 1990), and it is often the result of the addition of the caregiving role to already existing family roles (Schene, 1990). Burden is distinguished into two types: objective burden, which involves the disruption to the family/household due to the individual’s illness (Szmukler, 1996), and subjective burden, which involves the psychological consequences of the individual’s illness for the family (Schene, Tessler, & Gamache, 1994). It has long been established that families of patients with psychosis experience a great deal of FB (Bulger, Wandersman, & Goldman, 1993) in different life domains, including reduction of subjective health (Vaddadi, Soosai, Gilleard, & Adlard, 1997; Wittmund, Wilms, Mory, & Angermeyer, 2002), restrictions in leisure time, daily routine and social contacts, problems in working life, coping with the patients’ symptoms and emotional problems (L. Kuipers, 1993; Provencher, 1996).

The contribution of social and clinical factors on family functioning of patients with psychosis has been the focus of limited research. Recent investigations suggest that a variety of both patients’ and caregivers’ socio-demographic characteristics, as well as clinical variables, may affect family environment. Specifically, socio-demographic characteristics, such as the caregiver’s not being a spouse or being a father, unemployed and the patient’s being young and unmarried, were found to be related to family EE (Heikkila et al., 2002; Mo, Chung, Wong, Chun, & Wong, 2008). Additionaly, caregivers’ female gender, age and educational level, financial status and hours spent per day on providing care, as well as patients’ young age, male gender and marital status, were recognised as important predictors of FB (Caqueo-Urizar & Gutierrez-Maldonado, 2006; Li, Lambert, & Lambert, 2007; Roick et al., 2007; Schneider, Steele, Cadell, & Hemsworth, 2011). In a similar vein, the high level of caregivers’ EE status was associated with increased number of patients’ previous hospitalisations (Bertrando et al., 1992; Mavreas, Tomaras, Karydi, Economou, & Stefanis, 1992; Vaughan et al., 1992), whereas clinical characteristics, such as symptom type and severity, impaired functioning and higher frequency of relapses, were found to predict FB (Grandon, Jenaro, & Lemos, 2008).

However, family functioning is a multifaceted concept which includes numerous constructs, such as family cohesion, flexibility and communication, as well as family members’ satisfaction from the way their family functions. Within the Circumplex Model of Marital and Family Systems (Olson, Sprenkle, & Russell, 1979), family cohesion represents the balance between closeness and individuation (Olson, 1993), flexibility represents the balance between stability and change (Olson & Gorall, 2006), whereas communication is defined as the positive communication skills used by the family members (Olson & Gorall, 2006), and it is viewed as a facilitating dimension that helps families negotiate cohesion and flexibility issues (Olson, Gorall, & Tiesel, 2007). Research has shown that families of patients with psychosis may have deficits in family cohesion and adaptability as compared to non-clinical families (Friedmann et al., 1997; Sun & Cheung, 1997).

To the best of our knowledge, no study has explored in depth the effect of both caregivers’ and patients’ socio-demographic characteristics, as well as clinical features of the illness, in family functioning of patients with psychosis, as reflected through family cohesion and adaptability, whereas such research with regard to EE and FB is limited. Furthermore, studies on determinants affecting family functioning are limited in Greece, where family members are the major source of caregiving in psychosis. Recent studies which have been carried out in similar Southern European countries, for example, Italy (Carra, Cazzullo, & Clerici, 2012), have shown that clinical characteristics, such as the number of patient’s previous hospitalisations or the duration of illness, were not associated with high levels of caregiver’s EE; however, patient’s previous psychosocial functioning, as measured by educational level, seemed to protect the caregivers from high-EE status.

The aim of this study is to describe the socio-demographic and illness-related characteristics associated with family functioning in psychosis and to identify the determinants of unhealthy family functioning in a sample of patients and their families in Heraklion, Crete, Greece. On the basis of the existing literature, it is predicted that both socio-demographic and illness-related factors would have an effect on unhealthy family functioning in psychosis. Understanding the nature and development of unhealthy family functioning is critical in formulating effective interventions. Instead of focusing on a single dimension, such as EE or FB, an approach which embraces a broader view of family functioning encompassing constructs, such as cohesion and adaptability, would appear more useful. Such knowledge may increase our understanding of the intrafamilial relationships of patients with psychosis, thereby making it easier to identify patients and relatives who need intervention.

Methods

Participants

A total of 104 patients consecutively admitted to the Psychiatric Clinic of the University Hospital of Heraklion, Crete, Greece, and their family caregivers were contacted and informed about the purpose of this study during a 12-month period (October 2011–October 2012). Finally, 100 (response rate = 96.1%) patients and their family caregivers agreed to participate (50 patients with a first episode psychosis and 50 patients with chronic psychosis). The key caregiver was defined as the person who provides the most support devoting a substantial number of hours each day in taking care of the patient. To be eligible for inclusion in the study, the patients had to meet the following criteria: (1) to be between 17 and 40 years old, (2) to have a good understanding of the Greek language, (3) to have been out of hospital for at least 6 weeks and considered stabilised by their treating psychiatrist, (4) to be living with a close relative and (5) to have a diagnosis of schizophrenia or bipolar disorder according to Diagnostic and Statistical Manual of Mental Disorders (4th ed., text rev.; DSM-IV-TR) or International Classification of Disease (ICD-10) and with no evidence of organicity, significant intellectual handicap or primary diagnosis of substance abuse. Inclusion criteria for the caregivers were as follows: (1) to be between 18 and 75 years old, (2) to have a good understanding of the Greek language, (3) to have no diagnosed psychiatric illness and (4) to be either living with or directly involved in the care of the patient.

Procedure

Caregivers were interviewed by the first author (K.K.) in individual sessions in the Psychiatric Clinic of the University Hospital of Heraklion, Crete, Greece, where participants were asked to take part in a study assessing family functioning of patients with schizophrenia and bipolar disorder. Caregivers were given an information sheet describing the aims of the study. The time needed to complete the interview was approximately 75–90 minutes. Patients’ socio-demographic and clinical data were extracted from medical records and confirmed during the interview by the caregivers, whereas patients’ symptoms and functioning were assessed by their treating psychiatrist within 2 weeks from the caregivers’ assessment. All participants involved in this study were informed about the scope and the purpose of the study and provided written informed consent. The study was approved by the Ethical Committee of the University Hospital in Heraklion, Crete, Greece.

Measures

Socio-demographic characteristics

Socio-demographic characteristics, such as caregiver’s gender, age, education, marital status, employment status, origin and current residence, financial status, family structure, relation to the patient, contact with the patient and so on were collected through structured questionnaires administered by the researchers. The socio-demographic indicators of the patient included the characteristics mentioned previously and the following: clinical diagnosis, illness’s onset, patient’s age at illness’s onset, hospitalisation to psychiatric clinic, longer and last hospitalisation and therapeutic interventions.

Family Adaptability and Cohesion Evaluation Scales IV Package

Family functioning was assessed by means of the Family Adaptability and Cohesion Evaluation Scales IV Package (FACES IV Package) (Olson et al., 2007). The FACES IV Package contains the six scales from FACES IV, as well as the Family Communication Scale (FCS) and the Family Satisfaction Scale (FSS), and includes 62 items in total. The scales are self-reports, and they can be completed by all family members over the age of 12 years.

FACES IV (Olson et al., 2007) measures family functioning in terms of family cohesion and flexibility. The instrument contains six scales comprised of seven items each, with a total of 42 items. Each family member rates his or her agreement or disagreement with how well each item describes his or her family by selecting among the five alternative responses: 1 = strongly disagree, 2 = generally disagree, 3 = undecided, 4 = generally agree and 5 = strongly agree. FACES IV displays a six-factor structure of family functioning. There are two balanced scales that assess Balanced Cohesion and Balanced Flexibility and four unbalanced scales assessing the high and low extremes of Cohesion and Flexibility – Disengaged and Enmeshed for cohesion, and Rigid and Chaotic for flexibility. These scales display high levels of reliability and validity (Gorall, Tiesel, & Olson, 2006). Higher scores on the balanced scales are indicative of healthier functioning, and the converse holds truth for the unbalanced scales. In addition, three additional ratio scores can be calculated with FACES IV (Cohesion, Flexibility and Total Circumplex). When each score of the Cohesion and Flexibility ratios is at one and higher, the family system has more balanced levels of Cohesion and Flexibility. When the Total Circumplex ratio is one or higher, the family system is viewed as more balanced and functional.

FCS (Olson & Barnes, 1996) is a 10-item scale which addresses many of the most important aspects of communication in a family system. The respondents are asked to state the degree of their agreement or disagreement with how well each item describes their families by selecting among the five alternative responses (1 = strongly disagree, 2 = generally disagree, 3 = undecided, 4 = generally agree and 5 = strongly agree). A higher score on the scale indicates more positive communication in family system.

FSS is, also, a 10-item scale that assesses the satisfaction of family members in regard to family cohesion, flexibility and communication (Olson, 1995). The respondents are asked to state how satisfied or dissatisfied are they for each item describing their families by selecting among the five alternative responses (1 = very dissatisfied, 2 = somewhat dissatisfied, 3 = generally satisfied, 4 = very satisfied, 5 = extremely satisfied). A higher score on the scale indicates greater satisfaction in family system.

The FACES IV Package has been translated and validated for the Greek population by Koutra, Triliva, Roumeliotaki, Lionis, and Vgontzas (2012) and has demonstrated good psychometric properties.

Family Questionnaire

EE was measured via the Family Questionnaire (FQ) (Wiedemann, Rayki, Feinstein, & Hahlweg, 2002). The FQ is a 20-item self-report questionnaire measuring the EE status of relatives of patients with schizophrenia in terms of EOI and CC. EOI includes unusually over-intrusive, self-sacrificing, overprotective or devoted behaviour, exaggerated emotional response and over-identification with the patient, whereas CC is defined as an unfavourable comment on the behaviour or the personality of the person to whom it refers (Leff & Vaughn, 1985). The measure consists of 10 items for each subscale. Responses range from 1 = ‘never/very rarely to 4 = very often and a higher total score indicates higher EE. The developers provide a cut-off point of 23 as an indication of high CC and 27 for EOI. The FQ has excellent psychometric properties, including a clear factor structure, good internal consistency of subscales and inter-rater reliability in relation to the Camberwell Family Interview (CFI) (Leff & Vaughn, 1985) of EE. The FQ has been translated and validated for the Greek population by Koutra et al. (2014) and has demonstrated good psychometric properties.

Family Burden Scale

The Family Burden Scale (FBS; Madianos et al., 2004) was used to measure FB. The FBS consists of 23 items. The four FBS dimensions are defined as follows: (A) impact on daily activities/social life (eight items) – defined in terms of burden experienced regarding disruption of daily/social activities; (B) aggressiveness (four items) – captures the presence of episodes of hostility, violence and serious damages at home; (C) impact on health (six items) – shows signs and psychopathological symptoms reported by the family caregiver; and (D) economic burden (five items) – defined in terms of financial problems created by the patient’s illness. Factors A, B and D items tap objective burden, whereas C items underlie subjective burden. The scale has been originally developed and standardised for the Greek population and has demonstrated good psychometric properties. The developers provide a cut-off point of 24 (for the total scale score) to indicate high levels of burden.

General Health Questionnaire

The General Health Questionnaire-28 (GHQ-28) item version (Goldberg et al., 1997), a self-administered instrument that screens for non-psychotic psychopathology in clinical and non-clinical settings, was used to assess relatives’ psychological distress. Its four subscales measure somatic symptoms, anxiety/insomnia, social dysfunction and severe depression. In this study, the Likert scoring procedure (0, 1, 2, 3) is applied providing a more acceptable distribution of scores and the total scale score ranges from 0 to 84. Higher scores on the scale are indicative of poorer psychological well-being. The cut-off score for identifying cases of psychiatric disorder is 23/24 for Likert scoring. The 28-item version of this instrument has been adapted for the Greek population by Garyfallos et al. (1991) and has demonstrated good psychometric properties.

Statistical analysis

We decided to conduct first univariate analysis to examine crude associations between each one of the exposures and the family outcomes. Additionally, we run multivariate analyses to investigate the relationship between exposures and outcomes after controlling for the potential effect of several confounders.

Bivariate associations between normally distributed continuous dependent variables (family variables) and categorical independent variables (socio-demographic and clinical characteristics) were studied using either Student’s t-test or analysis of variance (ANOVA). Bivariate associations between non-normally continuous dependent variables (family variables) and categorical independent variables (socio-demographic and clinical characteristics) were studied using non-parametric statistical methods (Mann–Whitney and Kruskal–Wallis tests). Pearson’s r or Spearman’s rho correlation coefficient was used to estimate the strength of the association between continuous dependent and independent variables. Multivariate linear regression models were implemented to examine the associations between family variables with socio-demographic and clinical characteristics, entered simultaneously in the models. Separate multivariate models were built having as an outcome each one of the subscales of FACES IV Package, FQ, FBS and GHQ-28. The determinants related to each specific scale with p < .05 in the univariate analysis were included in each model. Estimated associations are described in terms of β-coefficients (beta) and their 95% confidence intervals (CIs). All hypothesis testings were conducted assuming a .05 significance level and a two-sided alternative hypothesis. All statistical analyses were performed using SPSS Statistics 19 software (IBM, Armonk, NY, USA).

Results

Sample characteristics

As far as caregivers’ characteristics are concerned, our sample consisted of 15 males (15.0%) and 85 females (85.0%), ranging in age from 28 to 75 years with a mean age of 56.80 years (standard deviation (SD) = 9.98). The 64.0% had finished elementary or high school, and the vast majority of the sample (72.0%) were not working. The 82.0% were living in urban areas, and the 63.0% were married. Finally, 92.0% were parents, 81.0% were living with the patient and 95.0% had daily contact with the patient. In terms of family structure, the 64.0% of the families were two-parent families, while the 36.0% were one-parent families.

Regarding patients’ socio-demographic and clinical characteristics, the sample consisted of 66 males (66.0%) and 34 females (34.0%), ranging in age from 17 to 40 years with a mean age of 31.09 years (SD = 5.75). The vast majority of the patients were single (85.0%), they came from urban areas (91.0%) and they were living in urban areas (86.0%). Half of the sample had finished lyceum or had some years in university. The 86.0% were not working at the time of the assessment, whereas almost half of the sample had no income. As far as diagnosis is concerned, 82.0% of the patients had a diagnosis of schizophrenia spectrum disorders and 18.0% had a diagnosis of bipolar disorder. The patients had an onset of illness between 15 and 39 years of age with a mean age of 24.03 years (SD = 5.48). Half of the patients had an onset of illness at 4 years or longer, 50.0% had one hospitalisation and 40.0% had 2–4 hospitalisations. The length of longer hospitalisation was up to 20 days for 65.0% of the sample, and more than 20 days for 35.0% of the sample.

Socio-demographic and clinical characteristics and family cohesion and adaptability, multivariate analysis

Table 1 presents the results of the multivariate analysis. Spouses or siblings of patients tended to score lower in Balanced Cohesion and Cohesion Ratio, and higher in the Chaotic scale than parents. Caregivers who had contact with the patients 1–2 times/week scored lower in Balanced Cohesion, higher in the Disengaged and lower in the Enmeshed scales as compared to relatives who had daily contact with the patients. Single-parent families scored significantly higher in the Disengaged scale than two-parent families. Caregivers’ financial status was related to the Enmeshed scale indicating that caregivers with income <€10.000 tended to score lower in the Enmeshed scale as compared to those with no individual income.

Table 1.

Socio-demographic and clinical determinants of family cohesion and flexibility, multivariate analysis.

	Family Adaptability and Cohesion Evaluation Scale IV
	Balanced Cohesion^a	Balanced Flexibility^a	Disengaged^a	Enmeshed^a	Rigid^a	Chaotic^a	Cohesion Ratio^a	Flexibility Ratio^a	Total ratio^a
	β (95% CI)	β (95% CI)	β (95% CI)	β (95% CI)	β (95% CI)	β (95% CI)	β (95% CI)	β (95% CI)	β (95% CI)
Caregivers’ characteristics
Residence
Rural vs urban	–	–	−3.42 [−8.48, 1.63]	–	–	–	–	–	–
Marital status
Married vs single	–	1.61 [−6.12, 9.33]	1.95 [−6.30, 10.19]	–	–	–	–	–	–
Divorced/widow vs single	–	3.07 [−3.43, 9.58]	−6.15 [−14.22, 1.92]	–	–	–	–	–	–
Financial status
<€10.000 vs no income	–	–	–	−2.58 [−4.73, −0.42]b	–	–	–	–	–
Relation to patient
Other vs parent	−5.91 [−10.38, −1.44]b	–	3.02 [−1.80, 7.85]	–	–	8.83 [4.29, 13.37]b	−0.82 [−1.33, −0.30]b	–	–
Contact with the patient
1–2 times/week vs daily	−6.08 [−11.55, −0.61]b	–	8.40 [2.73, 14.08]b	−4.64 [−9.10, −0.18]b	–	–	–	–	–
Family structure
One-parent vs two-parent family	−2.12 [−4.70, 0.45]	−3.47 [−10.28, 3.35]	9.36 [1.66, 17.06]b	–	–	–	−0.25 [−0.55, 0.05]	−0.19 [−0.42, 0.04]	−0.23 [−0.48, 0.03]
Patients’ characteristics
Age
Years	−0.05 [−0.31, 0.20]	−0.07 [−0.29, 0.15]	–	–	–	–	−0.00 [−0.03, 0.03]	−0.00 [−0.03, 0.02]	−0.00 [−0.03, 0.02]
Education
Medium vs low	–	2.13 [−0.14, 4.39]	–	–	−2.20 [−4.23, −0.17]b	−0.98 [−3.58, 1.62]	–	0.27 [0.04, 0.50]b	0.24 [−0.02, 0.50]
High vs low	–	3.91 [−0.19, 8.01]	–	–	−3.81 [−7.18, −0.44]b	−5.13 [−9.42, −0.84]b	–	0.61 [0.23, 1.00]b	0.42 [−0.01, 0.85]
Origin
Rural vs urban	–	–	–	3.32 [−0.19, 6.83]b	–	–	–	–	–
Residence
Rural vs urban	–	–	−0.43 [−5.99, 5.13]	–	–	–	–	–	–
Diagnosis
Bipolar disorder vs schizophrenia	–	–	–	2.78 [0.23, 5.34]	3.88 [1.39, 6.36]b	–	–	–	–
Onset of mental illness
1–4 years vs ≤12 months	−6.08 [−11.55, −0.61]b	−0.25 [−3.28, 2.78]	–	–	–	2.59 [−0.88, 6.08]	−0.26 [−0.66, 0.14]	−0.09 [−0.40, 0.22]	−0.14 [−0.49, 0.21]
>4 years vs ≤12 months	−2.12 [−4.70, 0.45]	−2.26 [−5.44, 0.92]	–	–	–	4.61 [1.31, 7.92]b	−0.46 [−0.88, −0.05]b	−0.33 [−0.65, −0.01]b	–0.35 [−0.71, 0.01]
Last hospitalisation
7–12 months vs up to 6 months	–	–	–	–	–	–	–0.29 [−0.60, 0.02]	–	–
>1 year vs up to 6 months	–	–	–	–	–	–	1.01 [0.19, 1.83]b	–	–
Number of hospitalisations
Hospitalisations	−0.29 [−1.22, −0.07]b	−0.71 [−1.23, −0.20]b	0.23 [−0.27, 0.72]	–	–	0.26 [−0.32, 0.85]	−0.06 [−0.13, 0.01]	−0.33 [−0.11, −0.00]b	−0.04 [−0.09, 0.02]

CI: confidence interval.

β-coefficients and 95%CIs are retained from linear regression models and adjusted for all the variables that were significant in the univariate analysis. Variables indicated with a long dash (–) were not included in the respective model.

p < .05.

Regarding patients’ characteristics, caregivers of medium-educated patients scored significantly lower in the Rigid scale and higher in Flexibility Ratio, while caregivers of highly educated patients had lower scores in the Rigid and Chaotic scales and higher in Flexibility Ratio as compared with caregivers of low educated patients. Caregivers of patients who came from rural areas had higher scores in the Enmeshed scale. Increased number of patient’s hospitalisations was related to lower levels of Balanced Cohesion, Balanced Flexibility and Flexibility Ratio. Caregivers of patients with bipolar disorder tended to have higher scores in the Rigid scale than caregivers of patients with schizophrenia. Caregivers of patients with an onset of mental illness of 1–4 years scored lower in Balanced Cohesion, whereas caregivers of patients with an onset of mental illness of 4 years or more scored higher in the Chaotic scale and lower in Cohesion Ratio and Flexibility Ratio than those of patients with an onset of illness of 12 months or less. Finally, caregivers of patients whose last hospitalisation was >1 year scored higher in Cohesion Ratio than those whose last hospitalisation was up to 6 months.

Socio-demographic and clinical characteristics and family communication and satisfaction, multivariate analysis

As far as the two additional scales of FACES IV Package – communication and satisfaction – are concerned (Table 2), a negative association between relation to patient and satisfaction was observed contributing to almost an 11-point decrease in spouses or siblings’ satisfaction as compared to parents. Caregivers of highly educated patients had significantly higher scores in both communication and satisfaction scales as compared to caregivers of patients with lower levels of education.

Table 2.

Socio-demographic and clinical determinants of family communication and satisfaction, multivariate analysis.

	Family Adaptability and Cohesion Evaluation Scale IV
	Communication^a		Satisfaction^a
	β	95% CI	β	95% CI
Caregivers’ characteristics
Relation to patient
Other vs parent	–	–	−10.84	[−17.83, −3.85]^b
Family structure
One-parent family vs two-parent family	−3.58	[−7.41, 0.25]	–	–
Patients’ characteristics
Age
Years	−0.07	[−0.45, 0.31]	−0.13	[−0.52, 0.26]
Education
Medium vs low	3.80	[−0.05, 7.64]	2.69	[−1.34, 6.71]
High vs low	8.35	[1.98, 14.72]^b	7.32	[0.76, 13.87]^b
Onset of mental illness
1–4 years vs ≤12 months	−1.02	[−6.16, 4.11]	−5.61	[−10.94, −0.28]^b
>4 years vs ≤12 months	−2.81	[−8.11, 2.47]	−7.32	[−12.73, −1.91]^b
Number of hospitalisations
Hospitalisations	−0.86	[−1.73, 0.01]	−0.90	[−1.80, −0.01]^b

CI: confidence interval.

β-coefficients and 95% CIs are retained from linear regression models and adjusted for all the variables that were significant in the univariate analysis. Variables indicated with a long dash (–) were not included in the respective model.

p < .05.

Caregivers of patients with an onset of mental illness of 1–4 years or 4 years or more scored significantly lower in satisfaction in comparison with those of patients with an onset of illness of 12 months or less. Finally, increased number of patient’s hospitalisations was related to lower levels of satisfaction.

Socio-demographic and clinical characteristics and family EE, multivariate analysis

Higher levels of EOI were reported by female caregivers, those who were living in rural areas and those who were not working. Increased number of patient’s hospitalisations was found to be significantly correlated with higher levels of caregivers’ EOI. One-parent families reported higher levels of CC as compared to two-parent families, while caregivers of patients who were living in rural areas were less critical as compared to those who were living in urban areas (Table 3).

Table 3.

Socio-demographic and clinical determinants of family expressed emotion, multivariate analysis.

	Family questionnaire
	Emotional over-involvement^a		Critical comments^a
	β	95% CI	β	95% CI
Caregivers’ characteristics
Age
Years	0.01	[−0.12, 0.13]	–	–
Gender
Female vs male	4.34	[1.22, 7.46]^b	–	–
Origin
Rural vs urban	2.25	[0.03, 4.47]^b	–	–
Employment status
Not working vs working	2.51	[0.01, 5.03]^b	–	–
Contact with the patient
1–2 times/week vs daily	–	–	5.48	[−1.25, 12.20]
Family structure
One-parent vs two-parent family	–	–	2.95	[0.28, 6.17]^b
Number of family members
Family members	–	–	−0.71	[−1.90, 0.47]
Patients’ characteristics
Age
Years	–	–	0.01	[−0.30, 0.31]
Residence
Rural vs urban	–	–	−4.44	[−8.62, −0.26]^b
Number of children
Children	−1.34	[−2.85, 0.16]	–	–
Onset of mental illness
1–4 years vs ≤12 months	–	–	1.61	[–2.55, 5.77]
>4 years vs ≤12 months	–	–	2.61	[−1.65, 6.88]
Number of hospitalisations
Hospitalisations	0.62	[0.11, 1.12]^b	0.53	[−0.17, 1.24]

CI: confidence interval.

β-coefficients and 95% CIs are retained from linear regression models and adjusted for all the variables that were significant in the univariate analysis. Variables indicated with a long dash (–) were not included in the respective model.

p < .05.

Socio-demographic and clinical characteristics and FB, multivariate analysis

Relatives caring for patients who were not working reported higher economic burden than those who were working. Patient’s onset of mental illness of 4 years or more was found to be significantly correlated with family’s higher economic burden. Increased number of patient’s hospitalisations was related to an adverse impact in family’s daily activities and social life and higher levels of patients’ aggressiveness. Patients’ hospitalisation of 20 days and longer was related to an adverse impact in family’s daily activities and social life. Only clinical variables, such as patient’s onset of mental illness of 4 years or more and increased number of patient’s hospitalisations, were related to caregivers’ objective burden, whereas patient’s onset of mental illness of 4 years or more was also related to total burden (Table 4).

Table 4.

Socio-demographic and clinical determinants of family burden, multivariate analysis.

	Family Burden Scale
	Economic burden^a		Impact on daily activities/social life^a		Aggressiveness^a		Impact on health^a		Objective burden^a		Total burden^a
	β	95% CI	β	95% CI	β	95% CI	β	95% CI	β	95% CI	β	95% CI
Caregivers’ characteristics
Origin
Rural vs urban	–	–	–	–	−0.80	[−1.92, 0.33]	–	–	–	–	–	–
Living with the patient
Yes vs no	0.79	[−0.29, 1.87]	–	–	–	–	–	–	–	–	–	–
Patients’ characteristics
Age
Years	–	–	−0.01	[−0.21, 0.20]	–	–	0.11	[−0.06, 0.29]	–	–	0.04	[−0.43, 0.51]
Gender
Female vs male	–	–	–	–	–	–	−1.64	[−3.42, 0.14]	–	–	−4.25	[−9.01, 0.51]
Residence
Rural vs urban	–	–	–	–	−1.00	[−2.61, 0.60]	–	–	–	–	–	–
Education
Medium vs low	–	–	−1.43	[−3.57, 0.72]	–	–	–	–	−2.82	[−6.29, 6.65]	−3.55	[−8.31, 1.20]
High vs low	–	–	−3.16	[−6.69, 0.36]	–	–	–	–	−3.71	[−9.95, 2.53]	−1.79	[−10.45, 6.87]
Employment status
Not working vs working	3.53	[2.30, 4.76]^b	–	–	–	–	–	–	3.81	[−1.11, 8.74]	5.49	[−1.26, 12.24]
Number of children
Children	−0.23	[–0.81, 0.34]	–	–	–	–	–	–	–	–	–	–
Diagnosis
Bipolar disorder vs schizophrenia	–	–	2.71	[0.10, 5.32]^b	–	–	–	–	–	–	–	–
Onset of mental illness
1–4 years vs ≤12 months	0.98	[−0.24, 2.19]	1.53	[−1.31, 4.37]	0.26	[−1.26, 1.79]	1.72	[−0.69, 4.13]	3.50	[−1.18, 8.17]	4.66	[−1.61, 10.95]
>4 years vs ≤12 months	1.43	[0.28, 2.58]^b	2.65	[−0.28, 5.59]	0.61	[−0.86, 2.08]	2.24	[−0.17, 4.65]	5.72	[1.26, 10.18]^b	7.92	[1.62, 14.22]^b
Number of hospitalisations
Hospitalisations	0.06	[−0.14, 0.27]	0.71	[0.22, 1.19]^b	0.27	[0.01, 0.52]^b	0.06	[−0.34, 0.47]	1.00	[0.21, 1.79]^b	1.05	[0.00, 2.11]
Longer hospitalisation
20+ days vs up to 20 days	–	–	2.23	[0.16, 4.31]^b	–	–	–	–	–	–	–	–

CI: confidence interval.

β-coefficients and 95% CIs are retained from linear regression models and adjusted for all the variables that were significant in the univariate analysis. Variables indicated with a long dash (–) were not included in the respective model.

p < .05.

Socio-demographic and clinical characteristics and caregivers’ psychological distress, multivariate analysis

Caregivers living in rural areas reported less anxiety and insomnia symptoms than those living in urban areas. Increased levels of social dysfunction were reported by caregivers who were not working, as well as caregivers of older patients, and caregivers of patients who had been hospitalised for 20 days and longer. On the contrary, caregivers of patients who came from rural areas reported less social dysfunction (Table 5).

Table 5.

Socio-demographic and clinical determinants of caregivers’ psychological distress, multivariate analysis.

	General Health Questionnaire
	Somatic symptoms^a		Anxiety and insomnia^a		Social dysfunction^a		Severe depression^a		General health index^a
	β	95% CI	β	95% CI	β	95% CI	β	95% CI	β	95% CI
Caregivers’ characteristics
Age
Years	–	–	–	–	0.01	[−0.05, 0.07]	–	–	0.13	[−0.25, 0.50]
Gender
Female vs male	2.27	[−0.16, 4.71]	–	–	–	–	–	–	5.96	[−1.65, 13.58]
Origin
Rural vs urban	–	–	–	–	1.36	[0.26, 2.45]	–	–	–	–
Residence
Rural vs urban	–	–	−2.97	[−5.86, −0.08]^b	–	–	–	–	–	–
Marital status
Married vs single	4.59	[−0.72, 9.91]	–	–	–	–	–	–	–	–
Divorced/widow vs single	4.72	[−0.77, 10.22]	–	–	–	–	–	–	–	–
Employment status
Not working vs working	–	–	–	–	1.16	[0.04, 2.36]^b	–	–	–	–
Relation to patient
Other vs parent	−1.40	[−4.81, 2.01]	–	–	–	–	–	–	–2.35	[−15.31, 10.61]
Patients’ characteristics
Age
Years	–	–	0.14	[−0.09, 0.37]	0.13	[0.03, 0.23]^b	–	–	0.31	[−0.32, 0.94]
Origin
Rural vs urban	–	–	–	–	−2.36	[−4.29, −0.43]^b	−2.58	[−5.62, 0.45]	−6.77	[−16.48, 2.94]
Employment status
Not working vs working	–	–	–	–	–	–	1.93	[−0.57, 4.43]	–	–
Onset of mental illness
1–4 years vs ≤12 months	1.19	[−1.32, 3.70]	2.56	[−0.63, 5.75]	–	–	–	–	4.24	[−3.74, 12.21]
> 4 years vs ≤12 months	1.93	[−0.53, 4.39]	3.16	[−0.09, 6.40]	–	–	–	–	6.22	[−1.84, 14.28]
Number of hospitalisations
Hospitalisations	0.14	[−0.26, 0.55]	0.34	[−0.20, 0.89]	0.17	[−0.08, 0.42]	–	–	0.33	[−0.98, 1.65]
Longer hospitalisation
20+ days vs up to 20 days	–	–	–	–	1.47	[0.36, 2.58]^b	–	–	3.94	[−1.72, 9.61]

CI: confidence interval.

β-coefficients and 95% CIs are retained from linear regression models and adjusted for all the variables that were significant in the univariate analysis. Variables indicated with a long dash (–) were not included in the respective model.

p < .05.

Discussion

This study was carried out to provide insight into the family assessment of patients with psychosis by analysing the socio-demographic and clinical determinants affecting family functioning in Crete, Greece. This is the first study evaluating simultaneously the effect of socio-demographic and illness-related characteristics on family cohesion and adaptability and is among the few similar studies evaluating several aspects of family functioning of mental health patients in Greece. Finally, our study replicated previous findings and provided some new evidence as well, regarding the effect of several socio-demographic and clinical factors on caregivers’ EE, FB and psychological distress.

The role of socio-demographic characteristics on family relationships

Caregiver’s relationship to and contact with the patient were found to significantly impact family cohesion and adaptability. Specifically, we found that spouses or siblings had lower scores on Balanced Cohesion and higher scores on the Chaotic scale (very high levels of flexibility) as compared to parents. These families presented less cohesion and lower levels of satisfaction. Caregivers who had contact with the patients 1–2 times/week reported lower levels of Balanced Cohesion and significantly higher and lower scores in the two unbalanced scales of cohesion – Disengaged (very low levels of cohesion) and Enmeshed (very high levels of cohesion) – respectively. One-parent families, most of whom were female-headed in our study, had also higher scores in the Disengaged scale. Also, caregivers with no income had higher levels in Enmeshed scale (very high levels of cohesion). The higher the educational level of the patient, lower the scores in the two unbalanced scales of flexibility (Rigid and Chaotic) and the more balanced the levels of family flexibility, also, higher the scores in communication and satisfaction scales. Finally, patients’ rural origin was found to be related to higher scores in Enmeshed scale indicating very high levels of cohesion.

Although research has demonstrated a variety of caregivers’ and patients’ socio-demographic characteristics which are related to EE, such as the caregiver’s not being a spouse or being a father, being unemployed and the patient’s being young and unmarried (Heikkila et al., 2002; Mo et al., 2008), as well as having low levels of education (Carra et al., 2012), our study did not support these results. We suggest that the two components of EE – EOI and CC – may be influenced by separate factors. Our findings showed that female caregivers presented higher EOI than males. Given the fact that in our sample the vast majority of female caregivers were mothers, our findings are consistent with the results of previous research showing that mothers tend to be more over-involved and intrusive as compared to fathers (Bentsen et al., 1996). EOI has been found to be a dominant cultural feature of the behaviour of Greek families (Mavreas et al., 1992). This may be related to the mothers’ parenting role, which tends to place them generally in a more nurturing role. Mothers might be more alarmed by their child’s distress and be more willing to communicate EOI, whereas fathers might be more likely to support the normative values of the family and possibly more likely to express their concern by CC (King & Dixon, 1999). Furthermore, caregivers who came from rural areas and those who were not working tended to have higher EOI scores. The first might be explained by the fact that living in a rural area in Greece makes the already difficult task of caregiving even more of a struggle. One of the biggest challenges caregivers face in rural areas is the complete absence of mental health services. The lack of professional help may sometimes increase caregivers’ worries and often places the onus of care and monitoring solely on them. This may lead to a more intrusive manner of engaging with the patient. Regarding the second finding, we assume that non-working caregivers, who spend almost all day at home, are those who spend a great deal of time with the patient and who exclusively take care of him or her, thus presenting overprotective behaviour towards the patient.

Our results indicated that higher levels of CC were reported by one-parent families, most of whom were female-headed in our study. A possible explanation with regard to this finding would be that in one-parent families, the full responsibility of earning a living, coping with the patient’s illness and caregiving that falls on one parent contributes to a more critical attitude towards the patient. In one-parent families, there is often no one to share responsibilities and decision-making rendering parenting a stressful experience. This can have negative consequences for both the parent and the patient. Interestingly, our findings indicated that caregivers of patients who were living in rural areas were less critical towards the patient as compared to those of patients who were living in urban areas. This finding could be explained as patients who are unable to work or who do not participate in household tasks are criticised for being lazy and selfish; unfortunately, in this context, the caregivers fail to understand that these could be potential manifestations of negative symptoms. Patients living in rural areas are more likely to spend hours outside of house farming the family land and participating in village activities. Hence, they are more functional within the household and community when compared to patients living in urban areas.

Although research has demonstrated that male and female caregivers may experience burden differently with women having higher scores in caregiving, depression and burden as compared to men (Schneider et al., 2011), our results did not support these findings. The same applies to a variety of caregivers’ characteristics, such as age and educational level (Li et al., 2007), financial status (Tsang, Tam, Chan, & Cheung, 2003), and hours spent/day on providing care (Li et al., 2007; Roick et al., 2007), which research have shown to be important predictors of FB. In addition, our results did not support the correlation between patients’ socio-demographic characteristics, such as young age (Caqueo-Urizar & Gutierrez-Maldonado, 2006; Li et al., 2007), male gender, marital status (Roick et al., 2007) and FB, although a positive relationship was confirmed between patients’ unemployment and caregivers’ FB. Previous research has shown that there is less burden for caregivers when patients are employed (Pickett, Greenley & Greenberg, 1995; Roick et al., 2007). We found that caring for patients who are not working, and who spend most of the day and nearly every day confined at home is related to higher economic burden for the family. Since most patients with major psychiatric disorders are not able to work on a regular basis, and consequently earn income that is substantially below that of the general population, they are often supported for a lifetime by their family. Furthermore, due to the difficult economic conditions in Greece, there are limited opportunities for mental health patients to work in subsidised employment, while few of the patients receive public welfare benefits.

Caregivers’ and patients’ socio-demographic characteristics were not associated with caregivers’ psychological distress, with the exception of caregivers’ residence and employment status and patients’ age and origin. Specifically, our results indicated that non-working caregivers reported higher levels of social dysfunction as compared to working ones, which is in line with the results of a previous study (Provencher, Perreault, St-Onge, & Rousseau, 2003). Caregivers who do not work and probably have lower incomes may experience more distress because they may have fewer resources to meet care demands. A second finding of our study was that caregivers living in rural areas reported less anxiety and insomnia symptoms than those living in urban areas. Furthermore, although research has shown that increased levels of caregivers’ psychological distress were associated with caring for younger patients (Addington, Coldham, Jones, Ko, & Addington, 2003), we found that increased levels of social dysfunction were reported by those taking care of older patients. Finally, we found that decreased levels of social dysfunction were reported by caregivers of patients who came from rural areas.

The role of illness-related features on family relationships

The longer duration of illness was found to be a risk factor for impaired family functioning, indicating unbalanced levels of cohesion, flexibility and satisfaction for families of patients with an onset of mental illness of 4 years or more compared to those of patients with an onset of illness of 12 months or less. Lower scores in Balanced Cohesion for families of patients with an onset of illness of 1–4 years and higher in chaotic (very high levels of flexibility) scale for families of patients with an onset of mental illness of 4 years or more were indicative of the way these families function. Higher number of patient’s hospitalisations was negatively associated with Balanced Cohesion, Balanced Flexibility and satisfaction scales, as well as the Flexibility Ratio indicating that families of patients with frequent admissions to hospital presented unbalanced levels of flexibility. Families of patients with bipolar disorder were more rigid than those of patients with schizophrenia. Finally, the longer the time since patient’s last hospitalisation (over a year vs up to 6 months), more balanced the levels of family cohesion.

A number of illness-related factors, including clinical diagnosis, mental illness onset, patient’s age at illness onset and number of hospitalisations, were examined in relation to family EE, and contrary to our hypothesis, no associations were found. This finding is consistent with the results of previous studies, which indicated that high EE was not associated with pre-morbid characteristics, symptom type and severity, age of onset, illness length and diagnosis (Carra et al., 2012; Raune, Kuipers, & Bebbington, 2004). However, we found that increased number of patient’s hospitalisations was significantly correlated with higher levels of caregivers’ EOI. According to the results of previous studies, a high level of caregivers’ EE status was associated with a larger number of previous hospitalisations (Bertrando et al., 1992; Mavreas et al., 1992; Vaughan et al., 1992). Although these results do not pertain specifically to EOI, most of these studies lend support to our findings. Increases in the number of hospitalisations have been previously found to be related to hostility (Bentsen et al., 1998), indicating that numerous hospitalisations, usually under constraint, strain family relationships, bring shame to the family and make the relatives feel helpless, perhaps leading them to adopt a hostile attitude towards the patient. We found exactly the opposite in our sample. A possible explanation for this is that caregivers present more intrusive, overprotective, excessively self-sacrificing or devoted behaviour to patients with impaired functioning who are often in need of hospitalisation.

Data also indicated that the longer duration of illness caused greater burden for the family even though previous research has not confirmed this association (McDonell, Short, Berry, & Dyck, 2003). Our results indicated that FB was more extensive if patients had an onset of mental illness of 4 years or more, a clinical characteristic which was, also, found to be significantly correlated with caregivers’ objective burden, and more specifically with higher economic burden for the family. In addition, the number of previous hospitalisations was found to adversely impact family’s daily activities and social life, increase caregiver’s perception about patient’s aggressiveness and thus cause greater objective burden for the family. Also, the diagnosis of bipolar disorder as compared to schizophrenia, as well as patient’s hospitalisation of 20 days and longer seemed to cause an adverse impact on the family’s daily activities and social life. Several stressors, such as patient’s negative symptoms, disruptive symptoms (i.e. psychotic and aggressive symptoms), frequent psychiatric hospitalisations and short illness duration, have been linked to increased burden in caregivers of persons with schizophrenia (Lowyck, de Hert, Peeters, Gilis, & Peuskens, 2001). Although there is little consensus on which factors are associated with the extent and/or pattern of FB (Lowyck et al., 2001), patients who are symptomatic, have many hospitalisations and have a short illness duration may be the most burdensome for caregivers (McDonell et al., 2003).

Clinical features of patient’s illness were not associated with caregiver’s psychological distress, with the exception of patient’s longer hospitalisation. Specifically, we found that patients’ hospitalisations for 20 days and longer were found to increase caregivers’ levels of social dysfunction. This finding is in line with the results of a very recent study indicating that the psychological burden experienced by primary caregivers is partly explained by the cumulative effects of chronic psychotic symptomatology, which over the years poses challenges to caregivers’ coping abilities (Mitsonis et al., 2012).

Strengths and limitations

The strengths of this study include its large sample size, the assessment of various aspects of family functioning by using standardised tools and the high participation rate (96.1%). Additionally, it is one of the very few studies to explore a variety of socio-demographic and clinical determinants of family functioning in psychosis, involving at the same time family outcomes such as family cohesion, adaptability, communication and satisfaction dimensions, apart from EE, FB and caregivers’ psychological distress. Nevertheless, there are some methodological limitations in our study. A possible limitation is that the population of patients and caregivers were from one catchment area, and hence, generalisability may be limited. Future research should include larger and representative samples and should collect data from different diagnostic groups. A second limitation is that due to the cross-sectional data, no causal relationships between variables could be drawn. Finally, even though patients’ symptom severity and psychosocial functioning were assessed by means of the Brief Psychiatric Rating Scale (BPRS; Overall & Gorham, 1962) and the Global Assessment Scale (GAS; Endicott, Spitzer, Fleiss, & Cohen, 1976), respectively, these measures were not included in this analysis, and these findings will be published in an upcoming report.

Conclusion

The findings of our study are in line with those of other studies, which suggest that both social and illness-related risk factors tend to become important determinants of family functioning in psychosis. With regard to the caregivers’ socio-demographic characteristics, gender, employment status, origin, residence, financial status, relation to the patient, contact with the patient and family structure were among the most significant determinants of family functioning. Concerning patients’ socio-demographic characteristics, age, education, origin, residence and employment status were found to significantly affect family functioning. In addition, in this first report from a Greek population, illness-related factors, such as onset of mental illness, number of hospitalisations, last hospitalisation, longer hospitalisation and clinical diagnosis, were found to impact on intrafamilial relationships. Our study provides two important contributions to the literature on family functioning of patients with psychosis. First, it documents both caregivers’ and patients’ socio-demographic determinants of family functioning, and second, it explores more closely how they are related to specific domains of family functioning.

Based on the concept of psychosocial rehabilitation, families of individuals with psychosis are actively participating in the care of their relatives nowadays. The family has thus become an important agent tending to the patients’ mental functioning and aiding the course of their recovery (Heikkila et al., 2006). Identifying the aspects of family functioning in psychosis helps in developing understandings regarding the factors which may contribute to the rehabilitation or relapse of the patients and the support required to strengthen positive family interactions as they take on the burdens of care. In this respect, the results of our study underline the need for more research aiming at explaining the interactions between social and clinical factors and their contribution to family functioning.

Furthermore, family psychoeducational interventions should be implemented from the early stages of psychosis in order to promote caregivers’ knowledge about the patient’s illness and improve dysfunctional interactions and communication patterns within the family (Carra, Montomoli, Clerici, & Cazzullo, 2007; E. Kuipers, Lam, & Leff, 2002; Pharoah, Mari, Rathbone, & Wong, 2010). Special attention should be given to the specific cultural and socio-demographic characteristics of the caregivers, such as, female gender, employment status, rural residence and the patients, such as, young age, low educational level, unemployment and other, which were found to adversely impact intrafamilial relationships in psychosis. Additionally, psychoeducational interventions should be primarily targeted to the caregivers of patients with increased number and length of hospitalisations, as well as longer duration of mental illness to reduce relapse rates.

Footnotes

Acknowledgements

We are grateful to all patients and their caregivers who participated in this study. Special thanks to the psychiatrists and the whole staff of the Psychiatric Clinic of the University Hospital of Heraklion, Crete, Greece, for their contribution and understanding.

Funding

This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.

References

Addington

Coldham

E. L.

Jones

Addington

(2003). The first episode of psychosis: The experience of relatives. Acta Psychiatrica Scandinavica, 108, 285–289.

Bentsen

Boye

Munkvold

O. G.

Notland

T. H.

Lersbryggen

A. B.

Oskarsson

K. H.

… Malt

U. F.

(1996). Emotional overinvolvement in parents of patients with schizophrenia or related psychosis: Demographic and clinical predictors. British Journal of Psychiatry, 169, 622–630.

Bentsen

Notland

T. H.

Boye

Munkvold

O. G.

Bjorge

Lersbryggen

A. B.

… Malt

U. F.

(1998). Criticism and hostility in relatives of patients with schizophrenia or related psychoses: Demographic and clinical predictors. Acta Psychiatrica Scandinavica, 97, 76–85.

Bertrando

Beltz

Bressi

Clerici

Farma

Invernizzi

Cazzullo

C. L.

(1992). Expressed emotion and schizophrenia in Italy. A study of an urban population. British Journal of Psychiatry, 161, 223–229.

Biegel

Schultz

(1999). Caregiving and caregiver interventions in aging and mental illness. Family Relations, 48, 345–335.

Bloch

Szmukler

G. I.

Herrman

Benson

Colussa

(1995). Counseling caregivers of relatives with schizophrenia: Themes, interventions, and caveats. Family Process, 34, 413–425.

Brown

G. W.

(1985). The discovery of expressed emotion: Induction or deduction? In Leff

Vaughn

(Eds.), Expressed emotion in families (pp. 7–25). New York, NY: Guilford Press.

Bulger

M. W.

Wandersman

Goldman

C. R.

(1993). Burdens and gratifications of caregiving: Appraisal of parental care of adults with schizophrenia. American Journal of Orthopsychiatry Mental Health & Social Justice, 63, 255–265.

Butzlaff

R. L.

Hooley

J. M.

(1998). Expressed emotion and psychiatric relapse: A meta-analysis. Archives of General Psychiatry, 55, 547–552.

10.

Caqueo-Urizar

Gutierrez-Maldonado

(2006). Burden of care in families of patients with schizophrenia. Quality of Life Research, 15, 719–724.

11.

Carra

Cazzullo

C. L.

Clerici

(2012). The association between expressed emotion, illness severity and subjective burden of care in relatives of patients with schizophrenia. Findings from an Italian population. BMC Psychiatry, 12, 140.

12.

Carra

Montomoli

Clerici

Cazzullo

C. L.

(2007). Family interventions for schizophrenia in Italy: Randomized controlled trial. European Archives of Psychiatry and Clinical Neuroscience, 257, 23–30.

13.

Cechnicki

Bielańskab

Hanuszkiewiczb

Darenb

(2013). The predictive validity of expressed emotions (EE) in schizophrenia. A 20-year prospective study. Journal of Psychiatric Research, 47, 208–214.

14.

Endicott

Spitzer

R. L.

Fleiss

J. L.

Cohen

(1976). The global assessment scale. A procedure for measuring overall severity of psychiatric disturbance. Archives of General Psychiatry, 33, 766–771.

15.

Friedmann

M. S.

McDermut

W. H.

Solomon

D. A.

Ryan

C. E.

Keitner

G. I.

Miller

I. W.

(1997). Family functioning and mental illness: A comparison of psychiatric and nonclinical families. Family Process, 36, 357–367.

16.

Garyfallos

Karastergiou

Adamopoulou

Moutzoukis

Alagiozidou

Mala

Garyfallos

(1991). Greek version of the General Health Questionnaire: Accuracy of translation and validity. Acta Psychiatrica Scandinavica, 84, 371–378.

17.

Goldberg

D. P.

Gater

Sartorius

Ustun

T. B.

Piccinelli

Gureje

Rutter

(1997). The validity of two versions of the GHQ in the WHO study of mental illness in general health care. Psychological Medicine, 27, 191–197.

18.

Goldstein

M. J.

(1988). The family and psychopathology. Annual Review of Psychology, 39, 283–299.

19.

Gorall

D. M.

Tiesel

Olson

D. H.

(2006). FACES IV: Development and validation. Roseville, MN: Life Innovations.

20.

Grandon

Jenaro

Lemos

(2008). Primary caregivers of schizophrenia outpatients: Burden and predictor variables. Psychiatry Research, 158, 335–343.

21.

Heikkila

Ilonen

Karlsson

Taiminen

Lauerma

Leinonen

K. M.

… Salakangas

R. K.

(2006). Cognitive functioning and expressed emotion among patients with first-episode severe psychiatric disorders. Comprehensive Psychiatry, 47, 152–158.

22.

Heikkila

Karlsson

Taiminen

Lauerma

Ilonen

Leinonen

K. M.

… Salakangas

R. K.

(2002). Expressed emotion is not associated with disorder severity in first-episode mental disorder. Psychiatry Research, 111, 155–165.

23.

Hooley

J. M.

(2007). Expressed emotion and relapse of psychopathology. Annual Review of Clinical Psychology, 3, 329–352.

24.

King

Dixon

M. J.

(1999). Expressed emotion and relapse in young schizophrenia outpatients. Schizophrenia Bulletin, 25, 377–386.

25.

Koutra

Economou

Triliva

Roumeliotaki

Lionis

Vgontzas

A. N.

(2014). Cross-cultural adaptation and validation of the Greek version of the Family Questionnaire for assessing expressed emotion. Comprehensive Psychiatry, 55, 1038–1049.

26.

Koutra

Triliva

Roumeliotaki

Lionis

Vgontzas

(2012). Cross-cultural adaptation and validation of the Greek version of the family adaptability and Cohesion Evaluation Scales IV Package (FACES IV Package). Journal of Family Issues, 34, 1647–1672.

27.

Kuipers

Lam

Leff

(2002). Family work for schizophrenia: A practical guide. London, England: Gaskell Press.

28.

Kuipers

(1993). Family burden in schizophrenia: Implications for services. Social Psychiatry and Psychiatric Epidemiology, 28, 207–210.

29.

Lam

D. H.

(1991). Psychosocial family intervention in schizophrenia: A review of empirical studies. Psychological Medicine, 21, 423–441.

30.

Leff

Vaughn

(1985). Expressed emotion in families: Its significance for mental illness. New York, NY: Guilford Press.

31.

Lambert

C. E.

Lambert

V. A.

(2007). Predictors of family caregivers’ burden and quality of life when providing care for a family member with schizophrenia in the People’s Republic of China. Nursing & Health Sciences, 9, 192–198.

32.

Lowyck

de Hert

Peeters

Gilis

Peuskens

(2001). Can we identify the factors influencing the burden on family members of patients with schizophrenia? International Journal of Psychiatry in Clinical Practice, 5, 89–96.

33.

Madianos

Economou

Dafni

Koukia

Palli

Rogakou

(2004). Family disruption, economic hardship and psychological distress in schizophrenia: Can they be measured? European Psychiatry, 19, 408–414.

34.

Mavreas

V. G.

Tomaras

Karydi

Economou

Stefanis

C. N.

(1992). Expressed emotion in families of chronic schizophrenics and its association with clinical measures. Social Psychiatry and Psychiatric Epidemiology, 27, 4–9.

35.

McDonell

M. G.

Short

R. A.

Berry

C. M.

Dyck

D. G.

(2003). Burden in schizophrenia caregivers: Impact of family psychoeducation and awareness of patient suicidality. Family Process, 42, 91–103.

36.

Mitsonis

Voussoura

Dimopoulos

Psarra

Kararizou

Latzouraki

… Katsanou

M. N.

(2012). Factors associated with caregiver psychological distress in chronic schizophrenia. Social Psychiatry and Psychiatric Epidemiology, 47, 331–337.

37.

F. Y. M.

Chung

W. S.

Wong

S. W.

Chun

D. Y. Y.

Wong

K. S.

(2008). Expressed emotion in relatives of Chinese patients with first-episode psychosis in Hong Kong. Hong Kong Journal of Psychiatry, 17, 38–44.

38.

Olson

D. H.

(1993). Circumplex Model of marital and family systems: Assessing family functioning. In Walsh

(Ed.), Normal family processes (2nd ed., pp. 104–134). New York, NY: Guilford Press.

39.

Olson

D. H.

(1995). Family Satisfaction Scale. Minneapolis, MN: Life Innovations.

40.

Olson

D. H.

Barnes

(1996). Family Communication Scale. St. Paul: University of Minnesota.

41.

Olson

D. H.

Gorall

D. M.

(2006). FACES IV and the circumplex model, Retrieved from www.facesiv.com/pdf/3.innovations.pdf

42.

Olson

D. H.

Gorall

D. M.

Tiesel

J. W.

(2007). FACES IV manual. Minneapolis, MN: Life Innovations.

43.

Olson

D. H.

Sprenkle

D. H.

Russell

(1979). Circumplex model of marital and family systems: I. Cohesion and adaptability dimensions, family types, and clinical applications. Family Process, 18, 3–28.

44.

Overall

J. E.

Gorham

D. R.

(1962). The Brief Psychiatric Rating Scale. Psychological Report, 10, 799–812.

45.

Pharoah

Mari

Rathbone

Wong

(2010). Family intervention for schizophrenia. Cochrane Database System Review, 12, CD000088.

46.

Pickett

S. A.

Greenley

J. R.

Greenberg

J. S.

(1995). Off-timedness as a contributor to subjective burdens for parents of offspring with severe mental illness. Family Relations, 44, 185–201.

47.

Provencher

H. L.

(1996). Objective burden among primary caregivers of persons with chronic schizophrenia. Journal of Psychiatric Mental Health Nursing, 3, 181–187.

48.

Provencher

H. L.

Perreault

St-Onge

Rousseau

(2003). Predictors of psychological distress in family caregivers of persons with psychiatric disabilities. Journal of Psychiatric and Mental Health Nursing, 10, 592–607.

49.

Raune

Kuipers

Bebbington

P. E.

(2004). Expressed emotion at first-episode psychosis: Investigating a carer appraisal model. British Journal of Psychiatry, 184, 321–326.

50.

Roick

Heider

Bebbington

Angermeyer

Azorin

Brugha

, … the EuroSC Research Group. (2007). Burden on caregivers of people with schizophrenia: Comparison between Germany and Britain. British Journal of Psychiatry, 190, 333–338.

51.

Schene

A. H.

(1990). Objective and subjective dimensions of family burden. Towards an integrative framework for research. Social Psychiatry and Psychiatric Epidemiology, 25, 289–297.

52.

Schene

A. H.

Tessler

R. C.

Gamache

G. M.

(1994). Instruments measuring family or caregiver burden in severe mental illness. Social Psychiatry and Psychiatric Epidemiology, 29, 228–240.

53.

Schneider

Steele

Cadell

Hemsworth

(2011). Differences on psychosocial outcomes between male and female caregivers of children with life-limiting illnesses. Journal of Pediatric Nursing, 26, 186–199.

54.

Sun

S. Y.

Cheung

S. K.

(1997). Family functioning, social support to families, and symptom remittance of schizophrenia. Hong Kong Journal of Psychiatry, 7, 19–25.

55.

Szmukler

(1996). From family ‘burden’ to caregiving. Psychiatric Bulletin, 20, 449–451.

56.

Tsang

H. W.

Tam

P. K.

Chan

Cheung

W. M.

(2003). Sources of burdens on families of individuals with mental illness. International Journal of Rehabilitation Research, 26, 123–130.

57.

Vaddadi

K. S.

Soosai

Gilleard

C. J.

Adlard

(1997). Mental illness, physical abuse and burden of care on relatives: A study of acute psychiatric admission patients. Acta Psychiatrica Scandinavica, 95, 313–317.

58.

Vaughan

Doyle

McConaghy

Blaszczynski

Fox

Tarrier

(1992). The relationship between relative’s Expressed Emotion and schizophrenic relapse: An Australian replication. Social Psychiatry and Psychiatric Epidemiology, 27, 10–15.

59.

Wiedemann

Rayki

Feinstein

Hahlweg

(2002). The Family Questionnaire: Development and validation of a new self-report scale for assessing expressed emotion. Psychiatry Research, 109, 265–279.

60.

Wittmund

Wilms

H. U.

Mory

Angermeyer

M. C.

(2002). Depressive disorders in spouses of mentally ill patients. Social Psychiatry and Psychiatric Epidemiology, 37, 177–182.