Self-Esteem Treatment in Anxiety - positive versus negative memory representations

Anxiety disorders are reinforced by anticipated danger, activated in the
patient in certain situations. The association that a patient makes between
stimulus and expected danger is in short called the CS-US association in
learning-theoretical literature (Joos, Vansteenwegen, & Hermans, 2012). When
anticipating danger, patients will typically avoid the stimulus concerned (e.g.
do not engage with social activities, or do not stand in de cue in the
supermarket). Cognitive behavioral therapy aims to build safety associations as
opposed to the negative CS-US, by cognitive restructuring and by learning new
associations about the stimulus concerned through behavior change (Craske,
Liao, & Vervliet, 2012). Thus, avoidance behaviors should be reduced and
patients should be exposed to real life confirmations that the stimulus does
not imply danger. It has been shown that a patient does not unlearn old memory
representations. Rather, new safety associations will get stronger en inhibit
the old memory representations concerning the response a stimulus will elicit
(Craske et al., 2008). Thus, the association of social contact with a feeling
of safety gets stronger than its association with danger. Based on theory on
how cognitive behavior therapy works, it is said that negative and positive
memory representations compete with each other for determining a person*s
responses in terms of behavior, thoughts and feelings that a stimulus elicits
(competitive memory retrieval account) (Brewin 2006). The idea is that CBT does
not change the content of negative information, but rather changes the relative
activation of positive and negative memory representations, in order to support
the positive representations in being recollected from memory (Brewin 2006).

Cognitive behavioral therapy (CBT) is an effective treatment for anxiety
disorders (Balkom van et al., 2013; Hofmann, Wu, & Boettcher, 2014). However,
despite good effectiveness, there is still room for improvement (Hofmann, Fang,
& Gutner, 2014). About 40% is cured by CBT, but 30% will keep suffering from
severe symptoms despite intensive CBT (Durham, Higgins, Chambers, Swan, & Dow,
2012). Little is known about what makes these patients more vulnerable for poor
treatment response. One possible obstacle is low self-esteem. Recently, based
om 18 longitudinal studies, a meta-analysis has been conducted, examining
causal relationships through time with sophisticated analyses. It turned out
that low self-esteem is a cause of anxiety, as well as a consequence of it
(Sowislo & Orth, 2013). This means that, even though the effects are small,
both psychopathological phenomena reinforce each other. So there is evidence
for the scar-model (anxiety disorders damage one*s self-esteem) as well as for
the vulnerability model (low self-esteem renders someone more vulnerable for
anxiety symptoms). An even more recent study (with 5.607 adolescents) found
that the effects of family disadvantage and family functioning on social
anxiety symptoms was explained in large part by lower elf-esteem (Yen, Yang,
Wu, & Cheng, 2013). Another longitudinal study (1.641 high school students)
found that low self-esteem predicted later anxiety symptoms, and not the other
way around; anxiety symptoms did not predict a later low self-esteem (van
Tuijl, de Jong, Sportel, de Hullu, & Nauta, 2014). The vulnerability model is
therefore winning ground. Presumably, people with low self-esteem feel more
insecure and less able to cope with stressful situations, they may experience
less personal control and more selective pay attention to negative and fearful
information instead of reassuring information that signals safety. Low
self-esteem is presumably associated with vulnerability for stress, while
people with a positive and stable self-image have more of a buffer for
stressful situations (Zeigler-Hill 2011). Patients with a weak and inferior
self-image interpret situations and people as more threatening (Kesting &
Lincoln, 2013).
The treatment of low self-esteem may exert a positive effect on anxiety
symptoms, also when these symptoms do not respond to CBT. But, looking back at
the CBT mechanisms outlined above, it is unclear what works best: strengthening
positive memory representations or reducing negative memory representations. It
is also not clear whether both approaches will have an additive effect for low
self-esteem, and what the secondary effect on anxiety symptoms will be. Not
much research has been conducted in this area.

In CBT terms, then, there are two treatment strategies possible for better
self-esteem: to reduce or desensitize negative memory representations or
enhancing or strengthening positive memory representations. There exist two
therapy manuals in clinical practice that match these strategies:

- *EMDR second method* aims to reduce negative memory representations, by
desensitizing three to five memories from someone*s past that for the patient
still *prove* that he/she is worthless (or other negative self-belief).
- Competitive Memory Training (COMET) aims to strengthen positive memory
representations by (1) enhancing the salience of the positive memory
representations, (2) repeated activation, and (3) associating the positive
representation to the negatively laden stimulus.

While *EMDR second method* has thus far not been investigated for it*s
effectiveness on self-esteem, COMET has been proven effective in various
patient groups, yet not specific for people with an anxiety disorder. The
current research project will assess the effectiveness of both approaches in
this patient group. Furthermore, this project will look at whether and when a
combination is preferable over applying just one of these two treatment
approaches.

We want to answer the following research questions:

- Which of the mentioned treatment approaches is more effective for improving
self-esteem?
- Do both approaches have additive value (does a patient benefit more after
receiving both treatments, in comparison to receiving just one?)
- Does the order of conducting these two treatment approaches matter for the
total effect?
- How do the treatment approaches exert their effect on positive and negative
self-esteem as separate constructs? Are the expected specific effects confirmed
by the data?
- Do subjective positive personal characteristics and experiences predict the
success of COMET? Or if these factors are not present, does EMDR work better
than COMET?
- Will anxiety symptoms go down as self-esteem goes up? (it is only possible to
perform an explorative analysis with the current study design)

A Randomized Controlled Trial (RCT) with a crossover design and two
allocations. The two study allocations will receive this order of treatment
modules:

1. First EMDR second method (6 sessions in 6-8 weeks) and then COMET (6
sessions in 6-8 weeks)
2. First COMET ((6 sessions in 6-8 weeks) and then EMDR second method (6
sessions in 6-8 weeks)

Patients will be informed about the study by their treating doctor of
psychologist, or by professionals involved in the regular intake procedure. If
the patient is interested, the Rosenberg Self-esteem Scale will be filled out
in order to assess if the patient meets the basic inclusion criteria. The
doctor/therapist will inform the patient about the study, verbally as well as
on paper. After two weeks of time to consider, and if the patient agrees to
have contact with there researcher, the coordinating researcher of that
particular institution will contact the patient and ask if he/she wants to
participate. If the patient agrees, an *informed consent* will be signed and
the baseline measurement will be conducted. After that, the patient will be
randomized to allocation 1 or 2.

Randomization
The randomization will be executed using a digital system (www.randomised.com),
executed by the randomization bureau of Parnassia; a person that is not
involved in the research project and works at a different location than the
researcher. The randomization will be conducted with small batches of 4,
stratified across the two institutions. This way, it is prevented that the
mental health institution may constitute a confounding factor in the eventual
data interpretation.

Measurements
At baseline, halfway, and after end of the treatment, measurements will take
place using self-administered questionnaires. These questionnaires will be
given and taken back in by an independent research assistant. This way, the
patient can see that his/her therapist will not see the actual answers, and the
risk of bias due to socially desirable answering is reduced.

Therapists, training and fidelity checks
The executing therapists are at least a psychologist and schooled in EMDR.
Furthermore, they have experience with active CBT. They will be trained in the
two treatment approaches of this study; two days in total, followed by
supervision of the ongoing therapies. All sessions will be recorded, unless a
patient objects. A random sample of these recordings will be scored to assess
fidelity to the treatment manual. Furthermore, the treatment preference of the
therapist will be assessed, and included as a covariate in the analyses.
Possible, preference makes a difference executing the treatment and it*s
effectiveness, so it is useful to correct for this.

*EMDR second method* is a well defined therapy, manuals existing in books and
tought by the Dutch EMDR Association. EMDR stands for Eye Movement and
Desensitization Reprocessing. It is an effective treatment for Posttraumatic
Stress Disorder (Balkom van et al., 2013; Engelhard 2012; van den Hout,
Rijkeboer, Engelhard et al., 2012). EMDR desensitizes vivid mental
representations with negative emotionality (Shapiro n.d.). By using EMDR, these
representations get reduced in their vividness and emotionality, and the memory
content gets less accessible (van den Hout, Bartelski, & Engelhard, 2012). An
underlying principle is that negative events leave their tracks in the memory
of an individual in such a way that it causes symptoms, including dysfunctional
beliefs about oneself (e.g. *I am a bad person*) or the world (*I am in
danger*) (de Jongh, ten Broeke, & Meijer, 2010). By desensitizing negative
memory content, low self-esteem can probably be treated, and this approach is
known as *EMDR second method* (Broeke ten, Jongh de, & Oppenheim, 2012).

COMET stands for Competitive Memory Training and has in various studies proven
to be effective in reducing low self-esteem as well as depression (Korrelboom,
de Jong, Huijbrechts, & Daansen, 2009; Korrelboom, Maarsingh, & Huijbrechts,
2012; Korrelboom, Marissen, & van Assendelft, 2011; van der Gaag, van
Oosterhout, Daalman, Sommer, & Korrelboom, 2012). COMET uses positive memories
that encompass *counter-themes* of the current negative self-image. For
example: if someone thinks of himself *I am incompetent*, then the
counter-theme will be: *I am competent.* Representationts of this theme within
that person*s autobiographical memory will by selected and then repeatedly
relived as vividly as possible. This way the positive counter-theme becomes
more active in the actual memory, inhibiting the negative memory
representations, and improving self-esteem.

It will be expected of participating patients that they participate with the
three assessments (that will take up about 60 minutes per assessment), in which
they will answer questions about their psychological symptoms. Even though this
is not fun to do, we known that there are no risks involved in these
measurements. The scales and questionnaires have been used in research quite
often and are safe to administer. This burden is not thought to be high or very
taxing.

Next to the assessments, participants will receive two treatment modules, each
consisting of six sessions of 45 minutes (12 sessions in total). The treatment
will be provided by certified psychologists who have been trained to execute
these treatment modules. Within these treatments, CBT techniques will be used
to treat low self-esteem. In between the sessions, patients are asked to
practice with these techniques in their daily life situations. So this
comprises a larger burden than the assessments, yet this is regular within
psychotherapy.

We do not expect any risks for the patients, concerning these assessments and
treatment modules. The COMET therapy has been evaluated in six earlier
intervention studies, without negative effects and with positive effects. These
studies were concerned with various psychological conditions, e.g. depression
and even psychosis (van der Gaag et al., 2012). A COMET version has also been
applied to anxiety patients, without any negative effect (Korrelboom, Peeters,
Blom & Huijbrechts, 2014). The EMDR procedure has been executed in clinical
practive a lot, yet has not yet been investigated for self-esteem within
anxiety.