Clinical and cost-effectiveness of group schema therapy for
complex eating disorders: the GST-EAT study

EDs are a major health care problem that interfere with daily social, physical,
and psychological functioning [3], and are associated with high mortality rates
[4]. Compared to healthy individuals without EDs, people with EDs have lower
employment participation and higher health-care costs [27]. The World Health
Organization estimates that 70 million people worldwide have an ED. Global ED
prevalence rates increased from 3.4% to 7.8% between 2000 and 2018 [18], and
health care professionals see a *tsunami* of EDs due to the Covid-19 pandemic
[34].

Comorbidity has been shown to interfere with ED treatment engagement and
effectiveness [20]. Approximately 60% of individuals with EDs also meet a
diagnosis for personality disorder, and various personality traits have been
found to be associated with ED symptom severity and negative treatment outcomes
[30], in particular perfectionism [8]. The majority of research suggests that
standard treatments are not consistently effective in addressing personality
pathology, such as perfectionism, in EDs [28].

Although several treatments for EDs are described in the *Zorgstandaard
Eetstoornissen*, most of these treatments focus on a specific ED (e.g. MANTRA
and SSCM for anorexia nervosa, CBT and DGT for bulimia nervosa), while CBT-E is
considered a transdiagnostic treatment for all eating disorders. For this
reason, CBT-E is often used as the treatment of choice. Although CBT-E has
shown positive treatment effects with a 30-50% remission rate, CBT-E is not
effective for a significant part of people with an ED [5]. Further, CBT-E shows
high drop-out rates (22-50%), especially for anorexia nervosa [15]. During
treatment, people with an ED are experts in identifying thinking errors and
challenging their own negative thoughts and beliefs, but generally describe
very limited changes to their deeply held core beliefs. This is likely related
to the fact that CBT-E does not tap into underlying core beliefs nor problems
that may originate from childhood experiences such as abuse, neglect or trauma,
which are relevant to EDs. The presence of personality pathology in people with
EDs who improve on eating symptoms after CBT-E, increases the risk of relapse
[32].
For many people, EDs therefore reflect a complex intertwining of behavioural
eating pathology and personality pathology. There is a clear need for
innovative treatment models that can address ED pathology alongside comorbid
personality pathology from a transdiagnostic point of view. The current project
aims to overcome the limitations of standard CBT-E treatment for EDs by
evaluating the clinical and cost-effectiveness of Group Schema Therapy (GST)
for individuals with EDs. Schema Therapy (ST) is an integrative therapeutic
approach that is primarily aimed at treating people with entrenched
interpersonal and self-identity difficulties [39]. Central to ST are
maladaptive schemas and schema modes. Maladaptive schemas refer to a broad,
pervasive theme or pattern, comprised of memories, emotions, cognitions, and
bodily sensations regarding oneself and one*s relationships with others, which
are developed during childhood and elaborated throughout one*s lifetime. Schema
modes refer to the moment-to-moment thoughts, feelings, and behaviors of a
person when schemas are activated and interact with coping strategies. The main
goal of ST is to strengthen the *healthy adult* mode of a person by combining
aspects of cognitive, behavioural, psychodynamic, attachment and Gestalt
models. Regulating emotions and behaviors from the healthy adult mode keeps a
person within their own window of tolerance. Similar to CBT-E, ST focuses on
cognitive and behavioural interventions, yet gives equal weight to emotional
needs by means of experiential interventions. For example, imagery rescripting
is an experiential interventions used for processing memories of early adverse
events, such as childhood abuse and neglect. Furthermore, ST emphasizes the
importance of the therapeutic relationship, which is used to provide corrective
experiences. GST may therefore buffer against the relatively high drop-out
rates of ED patients in standard CBT-E care. ST has proven to be an effective
treatment for personality disorders [7], including treatment in a group format
[16]. A pilot study of GST for treatment-resistant EDs has yielded promising
results [29]. More specifically, a 20-session GST protocol showed a clinically
significant improvement in four out of six ED patients and yet unpublished
pilot data showed that five out of 12 (41.7%) ED patients that did not improve
following CBT-E scored in the healthy range with regard to ED symptomatology
after a 26-week GST protocol. Given these promising results, combined with the
limited effectiveness of current standardized ED treatments, the present
project will strengthen the body of evidence for GST as a treatment for
patients that do not benefit from current standard treatments for EDs.

The aim of the current project is to test the clinical and cost-effectiveness
of GST for patients with an ED (i.e., anorexia nervosa, bulimia nervosa, other
specified), that are not likely to benefit from CBT-E. The study will focus on
those patients who do not show a clinical response to CBT-E in the first phase
(8 sessions) of CBT-E treatment, which is the most robust indicator of poor
outcome [2]. GST is a promising yet so far insufficiently tested treatment for
patients that do not benefit from current evidence-based and standardized
treatments [29]. Specifically, results of our own pilot work showed that
patients that were
offered GST after failing to benefit from CBT-E showed a significantly greater
improvement in disordered eating behaviours as assessed with the Eating
Disorder Examination Questionnaire (EDE-Q) compared to patients continuing in
CBT-E, with a medium-large effect size (eta-squared =0.13). In the current
project, we build on this encouraging pilot data and will compare the clinical
and cost-effectiveness of GST for EDs in a full Randomized Controlled Trial.

A Randomized Controlled Trial (RCT) with an active control group is used.
Patients that fail to show an early response at the end of phase 1 of CBT-E (as
measured with the EDE-Q [13]) will be randomized to either a GST arm, or
continue with CBT-E as usual. An RCT may provide the strongest empirical
evidence for the effects of GST compared to CBT-E.
Figure 1 presents a flow chart of this project providing an overview of the
study design and the different phases. Participants for the project will be
recruited at seven eating disorder centers in the Netherlands (GGNet Amarum,
Youz Maastricht; MUMC+ Maastricht; GGZ Breburg; GGZ Friesland; Emergis Goes,
Accare Groningen/Assen).

Investigational treatment - GST
This treatment consists of 26 weekly group sessions of ST for EDs, supplemented
with 8 individual ST sessions, and a psycho-education webinar for parents,
family members, and/or partners. The individual sessions can be used for
imagery rescripting of adverse childhood experiences. The first 5 sessions of
the GST focus on explaining the ST model, placing the ED symptoms and
behaviours in the context of coping modes, and organizing these in a mode map
conceptualization. The following sessions focus on recognising and changing
personal coping modes and underlying early maladaptive schema*s, and developing
and strengthening the healthy adult mode. GST combines interpersonal,
experiential, cognitive and behavioural elements in a unified ST approach [16].
Although not at the core of GST, addressing the physiological aspects of the ED
(weight care and (restrictive) eating) is necessary and therefore also
incorporated in the protocol.

Active control intervention - CBT-E
This transdiagnostic ED treatment (current standard of care/treatment of
choice) consists of 20-40 individual therapy sessions, based on the
transdiagnostic CBT model of EDs. CBT-E consists of four phases. In phase one,
the patient creates a personal case formulation, and the focus is on
psycho-education on maintaining factors and at starting well with monitoring
eating behaviours and establishing a regular eating pattern. In phase two, the
first phase is evaluated and a treatment plan is made. In phase three, the main
mechanisms that are thought to maintain the patient*s ED (over-evaluation of
shape, weight, and eating, dietary restraint or restriction, being underweight,
and event- or mood-triggered changes in eating behaviour), are targeted, and a
relapse plan is created. Phase four focuses on evaluating the progress so far
and maintaining the changes that have been obtained [15].

Participants will complete a set of questionnaires at the start of treatment
phase 1 (pre-measurement) and at the end of this phase (post-measurement), to
determine who is eligible for participating in this study. Assessments will
also be done at the start of randomisation (T0), at the end of treatment (T1),
and 6 (T2) and 12 (T3) months after end of treatment. Participants will
complete the questionnaires online through Qualtrics. They can do this after
one of the therapy sessions at the treatment center. It will take approximately
120 minutes for completing the set of questionnaires. Completing the
questionnaires may make participants aware of their complaints and problems,
which may affect their well-being making them feel a bit inconvenient. However,
to the investigators best knowledge, all questionnaires used in this project
have been used extensively in previous research with patients causing no risk.