Applying Systems Theory and Therapy to Eating Disorders. Implications for Etiology, Prevention and Systemic Multifamily Therapy Across Cultures

BACKGROUND. Eating disorders are closely related to micro-system variables such as family, friends, school and work. Western systemic family therapy models have long postulated the positive impact of improving the relationship between the child and the family to predict treatment outcomes. Despite their strong evidence base, family therapy models such as multifamily therapy (MFT) for eating disorders have not been implemented or tested in routine care in Eastern European countries, like the Czech Republic. Additionally, it is questionable whether cultural variables from the macro-system impede with the implementation and effectiveness of western models across cultures. In contrast, current etiological theories for eating disorders acknowledge the risk of sociocultural variables (e.g. thin beauty ideal) for eating disorder onset, but underestimate the putative predictive effects of micro-system domains (family, friends, school and work) on onset of Diagnostic and Statistical Manual of Mental Disorders (DSM-V) eating disorders. Although a single longitudinal risk factor study proved that impairments of the individual with its micro-systems (operationalized as combined psychosocial outcome) predicted all four eating disorder categories, it remains unclear which domain and individual items show predictive effects on future onset of subthreshold and full syndrome DSM-V eating disorder in young women. OBJECTIVE. The first objective of this dissertation entitled ‘Applying Systems Theory and Therapy for Eating Disorders. Implications for Etiology, Prevention and Systemic Multifamily Therapy Across Cultures’ was to identify cultural barriers that impede with the implementation of the first MFT model for eating disorders in the Czech Republic (Study Ia). The second objective was to explore its effects on patients’ self-esteem and life quality (Study Ib). DSM-IV eating pathology and body weight of the same patient cohort was examined in addition to life quality and depressive symptoms of the entire sample (with parents) (Study Ic). The third objective was to investigate the putative predictive effects of the micro-system domains (family, friends, school and work) and single items on future onset of all subthreshold and full syndrome DSM-V eating disorders in a high-risk sample of young women (Study II). METHODS. Study Ia used a narrative approach to retrospectively identify cultural barriers of MFT implementation for the inclusion into a cultural competent checklist. In addition, qualitative feedback from patients with DSM-IV eating disorders (n = 15, aged 14 – 23) and their parents (n = 26, aged 40 – 54) were obtained at baseline and after 12-months treatment (T2). A patient pilot study (Study Ib), tested the effects of the Czech MFT on patients’ self-esteem and life quality by self-report, using an Analysis of Variance (ANOVA) with repeated measures and a Wilcoxon Rank Test. In Study Ic, the first Czech MFT case series examined the pre-post effects on eating pathology (interview-assessed and by self-report) and body weight including families’ self-reports on life quality and depression. A General Linear Model with fixed (GLM) and mixed effects (GLMM) analyzed the effects. Study II used studies with prospective longitudinal, interview-based research designs, to examine the univariate and multivariate predictive effects of family, friends, school and work impairments in samples of high-risk young females (n = 1,153; mean age = 18.5 years, SD = 4.2) over a three year period, using Cox Regression Hazard Models (CRM) and Classification Tree Analysis (CTA). Impairments in the micro-system domains were assessed with a psychosocial impairment self-report measure at baseline. Subthreshold- and full syndrome DSM-V eating disorder onset was monthly assessed, via diagnostic interviews. RESULTS. Study Ia summarized the barriers that complicated Czech MFT implementation to include into a cultural competent checklist. Barriers were1) underfinanced health-care, 2) stigmatization of eating disorders, 3) challenging pathways into therapy, 4) distrust in psychiatric care, 5) corporate climate and 6) internalized family roles. Despite cultural challenges, empirical results showed significant improved life quality scores (not self-esteem) for patients with F (1, 14) = 13.03, p < 0.001) at T2. Czech MFT for eating disorders was endorsed well by families, with ambivalent feedback from patients (Study Ib). In addition, MFT significantly improved eating pathology with F (1, 30) = 13.32 and body weight with F (1, 30) = 9.19 (both p .40). Post-hoc contrasts indicated better responsiveness for patients with Anorexia Nervosa (AN) and Bulimia Nervosa (BN) compared to patients with eating disorders not otherwise specified (EDNOS) at T2 (all p < 0.05). Significantly improved time effects for life quality and depressive scores were obtained for the entire sample with F (1, 82) = 12.83 and F (1, 82) = 9.64 (all p < 0.001). Patients showed the largest seizes of improvement (p < 0.05). For Study II, univariate CRM revealed increased hazard ratios of 107, 22 and 43% for the micro-system domains friends, family and school associated with one-unit increase in each domain raw score. In the multivariate CRM, friends emerged as the strongest predictor with 92% for eating disorder onset. The CTA suggested loneliness in the friends’ domain as most potent risk factors for subthreshold and full syndrome DSM-V eating disorder onset. CONCLUSION. Study I showed that multilevel system variables (such as macro-system variables) can make cross-cultural MFT implementation into Eastern European countries challenging. However, despite cultural barriers, Czech MFT for eating disorders was effective in reducing DSM-IV eating disorder pathology (including body weight) from pre- to post-treatment. In addition, families showed improved life quality and depressive symptoms after MFT. Study II found that micro-system impairments of young women within the domains: family, friends and school predicted the onset of all four DSM-V eating disorder categories, with friend impairments being the strongest transdiagnostic risk factor. Impairment with friends (e.g. loneliness) should be integrated into etiological theories and inform preventive strategies for DSM-V eating disorders in young women. CLINICAL IMPLICATIONS. When implementing MFT for eating disorders across cultures, a cultural competent checklist should be used to alleviate barriers. MFT for eating disorders showed transcultural robustness and thus should be implemented internationally (Study I). The new transdiagnostic risk factors (peer-impairment) for subthreshold and full syndrome DSM-V eating disorders should be incorporated into etiological models. Future studies should test whether impairments with friends is precedent to other risk factors (such as body image concerns). In addition, preventive programs should use impairments with friends as screening variables to identify young women at risk for DSM-V eating disorders. Preventive programs should be tailored to improve functioning with friends, led by same-age peer-advocates, and conducted in the teenager’s ecological context