National and Kapodistrian University of Athens
Διερεύνηση της σχέσης ψυχοπαθολογίας και επιθετικής συμπεριφοράς σχιζοφρενικών ασθενών με το εκφραζόμενο συναίσθημα και ο διαμεσολαβητικός ρόλος της απάντησης της κορτιζόλης σε πειραματικό κοινωνικό stress
Abstract
dc:descriptionIntroduction: Schizophrenia is a chronic and severe mental disorder. Factors such as stressful life events, the family environment, reduced compliance to treatment and substance misuse contribute to the emergence of relapses. In particular, the family environment is associated with relapses through "expressed emotion" (EE). Various characteristics of patients and family relatives have been studied with regard to EE and its two main components, Criticism and emotional over involvement (EOI). Recently, research in schizophrenia-related EE has extended to the staff of psychiatric facilities. EE toward patients with schizophrenia is typically reported to be lower in psychiatric halfway houses than in families, though based on indirect comparisons prone to biases. Moreover, it is unknown whether patient and caregiver characteristics are equally correlated with EE in families and halfway houses. Criticism and EOI have been found relatively independent in families of individuals with schizophrenia and often have selective (differential) correlates, though not yet systematically studied. Regarding the HPA (Hypothalamus Pituitary Adrenal) axis in schizophrenia, higher basic cortisol levels and reduced cortisol reactivity to various physical and psychosocial stress tests have been reported compared to controls. Simulated family EE conditions have been used as a psychosocial stressor in tests but not studied in regard to cortisol responses in individuals with schizophrenia. Purpose: This study aimed to: (a) directly compare patients with schizophrenia living in psychiatric halfway houses or with their families on staff’s or parents’ EE, respectively, adjusting for patient- and caregiver-related characteristics as confounders (b) identify staff- or parent-related and patient-related EE predictors in the two settings and also differential predictors of EE among staff and parents (c) search for differential correlates of criticism vs. EOI in each setting (families and halfway houses) among various patient- and caregiver-related characteristics and (d) investigate the effect of EE on patients’ cortisol response to experimental stress induced by simulated EE conditions and finally assess the mediatory role of cortisol response in the path from EE to patients’ psychopathology. Methods: We included 40 outpatients with schizophrenia living with their families and 40 inpatients in halfway houses for at least 3 months and recorded 56 parents’ and 22 psychiatric nurses’ EE, respectively, through Five Minutes Speech Samples (FMSS) interviews. Each outpatient was rated by 1–2 parents and each inpatient by 2-5 nurses while each nurse rated 1-12 inpatients. Four FMSS-derived outcomes were used in the analyses: Criticism, EOI (primary outcomes: borderline / low vs. high Criticism or EOI), the number of Critical comments and the number of Positive attitude statements (secondary outcomes of Criticism and EOI, respectively). We also recorded patients’ psychopathology (Brief Psychiatric Rating Scale, BPRS), Perceived criticism (PC) and clinicodemographic characteristics, caregivers’ (parents’ – nurses’) demographics, parents’ family burden (Family Burden Scale) and nurses’ burnout (Maslach Burnout Inventory). In addition, 36 inpatients and 34 outpatients were subjected to an experimental procedure simulating low and high EΕ conditions (based on FMSS), during which 4 saliva samples were taken (per EE condition) within an hour, in order to measure cortisol as an index of HPA axis reactivity. Statistical analyses were study aim-specific: (a) As EE ratings had a multilevel structure, multivariate generalized linear mixed models (GLMMs) were fitted in the total sample to investigate the effect of patient group on EE, adjusting first for patient-related confounders only and then also for basic caregiver demographics. Mediatory effects of the aforementioned characteristics were investigated in multilevel structural equation models. (b) EE predictors were investigated in backward stepwise GLMMs using “buildmer” R package. We first fitted models including either caregiver- or patient-related predictors in each setting and finally included both types of predictors in combined models. Setting-specific patient-related effects were investigated in interaction analyses. (c) to identify differential correlates of criticism vs. EOI, weighted Spearman correlations of Criticism and EOI with various patient- or caregiver-related characteristics were calculated and compared with Meng’s z-test. In all the above (a-c), adjustment for multiple tests identified the most robust associations. (d) The percentage of ratings with high Criticism or high EOI that each patient had received from his/her caregivers was initially calculated. Multivariate analyses for cortisol were implemented using two methods: (1) In the sample of patients having complete cortisol measurements across all four timepoints (complete tetrads) for at least one EE condition (low or high), we initially calculated, separately for each EE condition, the Area Under the Curve (AUC) of the 4 cortisol measurements along time. Subsequently, linear regressions of cortisol AUC were applied, including as predictors patient group and percentage of patients’ ratings with high Criticism or high EOI, while adjusting for confounders. For mediation analyses of cortisol AUC in the path from EE to patients’ psychopathology (BPRS), structural equation models were applied using MPlus 7.1. The 95% CI of indirect effects were calculated by bootstrapping 1000 samples. (2) To make full use of all available data, we applied linear mixed models of cortisol measurements, including as predictors time (T1-T4), patient group, percentage of patients’ ratings with high Criticism or high EOI and the interactions of the last two with time, while adjusting for confounders. For mediation analyses of cortisol in the path from EE to patients’ psychopathology (BPRS), multilevel structural equation models were applied using MPlus 7.1. The 95% CI of indirect effects were calculated with Bayes method.Results: Outpatients were younger, less chronic, and better educated, with higher negative symptoms and PC than inpatients. Nurses were younger and better educated than parents. Before adjustment, rates of high EE were similar in the two settings (EE 81.3% vs. 78.6%, Criticism 54.2% vs. 58.9%, EOI 46.5% vs. 48.2%, for nurses and parents, respectively). Study aim-specific analyses showed: (a) After adjusting for patient-related confounders, EOI was significantly higher in parents. However, after also adjusting for caregiver demographics, only Criticism was significantly higher in nurses. Patients’ age, negative symptoms, and PC and caregivers’ age and sex significantly mediated EE differences between the two groups. (b) After adjustment for multiple tests, in backward stepwise GLMMs including either caregiver- or patient-related predictors in each setting, nurses’ higher age and shorter work experience and lower inpatients’ negative symptoms robustly predicted higher EOI, while in the final combined models including both caregiver- and patient-related predictors, nurses robustly displayed lower EOI (i.e., reduced concern and disengagement) toward inpatients with higher negative symptoms. Several other features had nominal associations with Criticism and EOI in each setting, which did not survive adjustment for multiple tests. In interaction analyses, higher negative symptoms differentially predicted lower EOI in nurses only. (c) Criticism was weakly negatively correlated with EOI in nurses but negligibly in parents. Distinct patterns of significant differential correlates of Criticism vs. EOI arose across settings. Outpatients’ aggressive behavior and parents’ related burden were mainly associated with higher Criticism. Inpatients’ symptoms (agitation/aggression, negative and other psychotic symptoms) and nurses’ burnout (Depersonalization) were mainly associated with lower EOI. Inpatients’ PC and outpatients’ previous suicide attempts were equally associated with higher Criticism and lower EOI (mirror correlations). Finally, various inpatient attributes (older age, chronicity, unemployment and smoking) were associated with higher EOI only. Inpatients’ age, psychopathology (esp. agitation/aggression and negative symptoms) and PC survived adjustment for multiple tests. (d) The percentage of patients’ ratings with high Criticism or high EOI was not significantly different among the two patient groups. The type of stimulus (high Criticism / high EOI) that patients were exposed to during the high EE condition of the experiment strongly reflected the ratings they had mainly received from their carers. Out of 560 saliva samples initially collected, cortisol was finally measured in 497 after discarding 63 samples with blood admixture. Outpatients had higher cortisol only at T1 (before the experiment started) and only at the low EE condition. Cortisol AUC of any EE condition was not significantly different across patient groups. Cortisol was also not different among the two EE conditions irrespective of the sequence of their administration. In linear regressions of cortisol AUC, the percentage of patients’ ratings with high EOI had significant negative effects on low EE condition cortisol AUC in the total sample (p=0.022), in outpatients (p=0.014) and at trend level in inpatients (p=0.068) as well as on high EE condition cortisol AUC in inpatients only (p=0.017). The percentage of patients’ ratings with high Criticism had no significant effects in any sample and at no EE condition. Patient group had also no significant effect on cortisol AUC at any EE condition. In linear mixed models of cortisol measurements, significant effects of time, patient group and their interaction were recorded at both EE conditions, with outpatients having higher cortisol values at both EE conditions. We also found that patients with >50% of their ratings with high EOI had significantly lower cortisol values at both EE conditions compared to patients with <50% of their ratings with high EOI. Conversely, the percentage of patients’ ratings with high Criticism had no significant impact on cortisol. Mediation analyses using either of the two methods showed that indirect effects via cortisol (AUC or individual measurements) were non-significant for Criticism while they were always positive (either significant or most often non-significant) for EOI, i.e., non-compatible with a mediatory role of cortisol in the negative path from EOI to patients’ BPRS. Conclusions: Our findings highlight distinct and setting-specific pathogenetic pathways of Criticism and EOI in families of patients with schizophrenia and halfway houses, which might help customize psychoeducational interventions to families and staff of halfway houses. High carers’ Criticism had no impact on cortisol reactivity in a simulated EE conditions test but high EOI (possibly related to high Warmth) correlated to reduced cortisol reactivity.
Degree
thesis:*- Grantor dc:publisher
- National and Kapodistrian University of Athens
- Year dc:date
- 2026
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Douki, Stamatina
- Δούκη, Σταματίνα
Subjects
dc:subject × 33- Αντιλαμβανόμενη κριτική
- Διαφορικοί προβλεπτικοί παράγοντες
- Διαφορικές συσχετίσεις
- Εκφραζόμενο συναίσθημα
- Επαγγελματική εξουθένωση
- Κριτική
- Νοσηλευτές
- Οικογένειες
- Οικογενειακή επιβάρυνση
- Συναισθηματική υπερεμπλοκή
- Σχιζοφρένεια
- Ψυχιατρικοί ξενώνες
- Ψυχοεκπαίδευση
- Κορτιζόλη
- Criticism
- Differential correlates
- Differential predictors
- Emotional overinvolvement
- Expressed emotion
- Families
- Family burden
- Halfway houses
- Nurses
- Perceived criticism
- Professional burnout
- Psychoeducation
- Schizophrenia
- Επιστήμες Υγείας
- Ιατρική και Επιστήμες Υγείας
- Επιστήμες υγείας, γενικά
- Health Sciences
- Medical and Health Sciences
- Health sciences, general
Rights
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/61336