Mental imagery in women with eating disorders and healthy controls.

Eating disorders are serious psychiatric disorders (APA, 2000) and patients are
impaired for years and need a long time to achieve remission if remission is
achieved at all (Keel & Brown, 2010).
Treatment outcome for anorexia nervosa (AN) is poor. Seven to nine years after
intake only 31% of the women with AN achieved remission and 36% of this
remitted group relapsed in the years after remission (Keel, Dorer, Franko,
Jackson & Herzog, 2005). Remission rates for women with bulimia nervosa (BN)
were 75% of which 35% relapsed in the years after remission (Keel, et al.,
2005). Overall 30% of the eating disorder patients recovers and 30% recovers
partially and 30% has a chronic course and 10% dies due to the consequences of
the disorder (Steinhausen, 2002, 2009).
A meta-analytic review (Stice, 2002) showed that the thin-ideal internalization
is a risk factor for body dissatisfaction, dieting, negative affect and bulimic
pathology. Thin ideal internalization and body dissatisfaction also appear to
be maintenance factors for bulimic pathology. Body dissatisfaction is one of
the most robust and consistent factors for maintaining eating pathology because
it leads to dieting and dieting increases the risk for bulimic behavior.
Predictors for relapse for AN as well as BN, even after remission, are:
misperception of body weight or shape, fear of gaining weight or becoming fat
and over-concern with weight or shape (Keel et al., 2005). For women with BN
worse psychosocial function and overconcern with weight or shape increased the
risk of relapse (Keel et al., 2005).

There is evidence that in all eating disorder subtypes the same
transdiagnostic mechanisms play a major role in the maintenance of eating
disorder psychopathology (Fairburn, 2003, 2008). For BN Cognitive behavioral
therapy (CBT) is the evidence-based treatment of first choice (Hay, 2013).
Evidence based treatments for anorexia nervosa are limited, partially due to
the relative rarity of the disorder. However, evidence is growing that the
cognitive behavioral model and CBT are also applicable to AN as well as family
based therapy (Fairburn, Cooper, Doll, O'Connor, Palmer & Dalle Grave, 2013;
Hay, 2013). Better understanding of factors maintaining eating disorders and
improving treatment is of great importance.

Mental imagery has received growing attention in the last two decades because
of the powerful effect on emotions. Due to this effect on emotions and behavior
imagery plays a critical role in psychopathology (Brewin, Gregory, Lipton &
Burgess, 2010; Holmes & Mathews, 2010). Imagery is a key feature of
post-traumatic stress disorder (APA, 2000). Images in the form of flashbacks
evoke great emotional distress. Research has shown that imagery is not solely
linked to PTSD, but also to other mental disorders (Hackmann & Holmes, 2004),
such as social phobia (Hackmann, Suraway & Clark, 1998; Hackmann, Clark &
McManus, 2000), hypochondriasis (Muse, McManus, Hackmann, Williams & Williams,
2010), bipolar disorder (Holmes et al. , 2011), major depressive disorder and
anxiety disorders (Morina, Deeprose, Pusowski, Schmid & Holmes, 2011). The
powerful impact on emotion can cause distress and can have a role in the
maintenance of psychopathology (Holmes & Mathews, 2010).

Holmes and Mathews (2010) have presented a cognitive model of imagery versus
verbal representations. Including the impact of imagery and verbal
representation on emotion, perceived reality and behavior. In this model two
ways are distinguished in which images can be initiated: 1. a bottom-up process
leading to intrusive involuntary imagery as a result of a sensory cue matching
with a representation in episodic memory associated emotion and 2. a top-down
controlled process leading to the construction of an image. Imagery that is
emotional in content is hypothesized to activate emotional information
processing in the brain similar to real perceived events which has an influence
on action readiness. Verbal representations are less likely to have an
emotional impact and to be treated as "real" compared to imagery. Verbal
therapies, such as CBT, do not seem to be able to address fully the
intrusiveness and emotionality of the images .
Compared to verbal representations imagery evoked greater emotional responses
in an experimental study (Holmes & Mathews, 2005). Therefore, it could be
hypothesized that in eating disorders imagery might play a role in the
persistence of feeling fat. This imagery is overruling the cognitive knowing of
not being fat.

As stated earlier by Hackmann & Holmes (2004) verbal thoughts have been the
major focus of cognitive therapy and the importance of imagery in
psychopathology and thereby in psychotherapy is not fully recognized or
acknowledged. Imagery itself can also be apositive factor in treatment (Holmes
& Mathews, 2010). The use of imagery as a therapeutic technique possibly could
enhance treatment outcome and offers a non-verbal approach as an addition to
the already existing verbal therapies. As in fact it is already for PTSD. Two
techniques which are based on mental imagery appear to be effective in
patients with PTSD (NICE, 2005): imagery and rescripting (Brewin et al., 2010)
and Eye Movement Desensitisation and Reprocessing (EMDR; Shapiro, 1989).

The emotional impact of images is influenced by the perspective from which an
image is viewed and the vividness (Holmes & Mathews, 2010) The observer
perspective images (as seeing one-self from a distance) are supposed to be more
negative and unrealistic than the field perspective (seeing the image as
through one's own eyes) (Hackmann, Suraway & Clark). But the opposite with
observer perspective leading to a reduction of emotional arousal is also stated
(Brewin, et al., 2010). Hackmann, Suraway and Clark (1998) found that people
with social phobia reported significantly more spontaneously occurring negative
images in social situations compared to a control group. The images reported by
the individuals with social phobia were dominated by the observer perspective
but also contained field perspective. Observer perspective was related to more
negative emotions. Later research by Hackmann and colleagues (2000) showed that
in fear evoking social situations patients experience spontaneous recurrent
images with negative content involving several sensory modalities such as
vision and sound. Most recurrent images were linked to memories of unpleasant
experiences at the time of the onset of the disorder.

Research has shown that people with agoraphobia experience recurrent
distressing visual images often accompanied by body sensations in so called
agoraphobic situations, such as traveling by train or being in a supermarket
(Day, Holmes & Hackmann, 2004). Two third of the participants reported that the
images are related to memories of unpleasant events in childhood, such as an
abusive event at home or an threat or attack by a non-family member, and
negative self-beliefs. Interestingly these participants were not diagnosed with
PTSD. Osman, Cooper, Hackmann and Veale (2004) investigated the role of
spontaneously occurring images and their linkage to early memories in people
with body dysmorphic disorder (BDD). Interestingly, compared to the control
group no difference was found in the frequency of spontaneous appearance
related images. But the quality of the images differed significantly. People
with BDD had significantly more negative spontaneous occurring images or
impressions related to their appearance. This images had a negative content,
were viewed from an observer perspective and were more vivid and with greater
detail and involved more sensory modalities.

There are only a few studies on imagery in eating disorders. Somerville, Cooper
and colleagues(2007) used a semi-structured interview to investigate the
presence, content and characteristics of spontaneously occurring images in
women with bulimia nervosa compared to control participants. The control
participants were divided in dieting and non-dieting participants. Compared to
non-dieting controls, women with bulimia nervosa and the dieting controls
reported more spontaneous images when worrying about eating, weight or shape.
The emotional tone of the images of the women with bulimia nervosa was more
negative and anxiety provoking. Another intriguing finding was that the
vividness of the images increased with dietary restraint with women with BN and
dieting controls significantly differing from non-dieting controls. The authors
suggest that the meaning of the images might be more catastrophic and linked to
underlying assumptions and negative core beliefs. Such as was found by Cooper,
Todd & Wells (1998): the beliefs were linked to images of unpleasant
experiences in childhood or young adolescence of which some explicit related to
situations about weight and shape. All patients believed that restricting their
food intake helped with this negative self-beliefs. However, this study focused
more on exploring the core beliefs than on the images self.
In a sample of 30 patients with bulimia nervosa 57% reported to have images
prior to vomiting. These images were often recurrent and linked to past events
in which they were humiliated, abused or abandoned (Hinrichsen, Morrison,
Waller & Schmidt, 2007).

It is not surprising that images found in eating disorder patients are linked
to past negative events knowing that PTSD common is a comorbid disorder. Since
the comorbidity of PTSD in eating disorders is high: ranging from 21% up to
62% (Brewerton, 2007). A history of trauma is a non-specific risk factor for
the development of eating disorders (Swinbourne & Touyz, 2007) The nature of
the reported images (Somerville et al., 2007; Cooper et al., 1998) might be
symptoms of a comorbid PTSD and not a specific feature of the eating disorder.
Distinguishing between images related to PTSD or images related to eating
disorders might be of importance in understanding and treating eating disorders.

Not only images of past events are associated with psychopathology, also images
of the future have a role in the maintenance of psychopathology.
Holmes, Crane, Fennell and Williams (2007) found in a small sample of people
recovered of depression and suicidality that all of them reported mental images
of acting out future suicidal plans of being dead during crisis. These
prospective imagery is described as vivid and detailed and intrusive. There is
evidence that people with major depressive disorder or an anxiety disorder had
more problems with imagining positive events happening in the future and that
people with an anxiety disorder had a greater ability to imagine vividly
negative events in the future than patients with a depression (Morina,
Deeprose, Pusowski, Schmid & Holmes, 2011).
Patients with a bipolar disorder reported more general and prospective imagery
than the control group. The patients with an unstable mood pattern (i.e. higher
levels of anxiety and depression) reported the highest levels of involuntary
intrusive prospective imagery (Holmes et al., 2011).
This research shows that prospective mental imagery can play an important role
in the maintenance of psychopathology. It might be hypothesized that the fear
of becoming fat or gaining weight can be seen as future oriented. Patients with
eating disorders might experience prospective negative mental imagery similar
to patients with an anxiety disorder or depressive disorder.

Summarizing the above, eating disorders are complex disorders with long
duration of treatment, limited positive treatment outcome and high relapse
rates. Treatments are characterized by a verbal and behavioral approach.
Imagery is a factor of importance in a variety of psychiatric disorders.
If imagery also plays a role in the maintenance of eating disorder pathology,
adding these imagery based techniques maybe could enhance the effectiveness and
outcome of treatment. In eating disorders there are three domains that are
potentially valuable with respect to imagery based techniques: negative core
beliefs, emotional regulation difficulties and disturbed body image (Tatham,
2011).

Primary Objective:
First we need to establish that patients with an eating disorder and healthy
controls differ in mental imagery. So, the first aim of our research is to
compare the nature and characteristics (emotionality, vividness and
perspective) of spontaneously occurring mental imagery in women with eating
disorders and healthy controls. Do women with eating disorders differ from
healthy controls with respect to recurrent spontaneously occurring mental
imagery? We expect that patients with an eating disorder, have a higher
frequency of occurring mental imagery which are more emotional and vivid
compared to the healthy control group.

The second aim is to investigate the characteristics and impact of intrusive
prospective mental imagery in women with eating disorders and healthy controls.
Do women with eating disorders differ from healthy controls with respect to
intrusive prospective mental imagery? We hypothesize that patients with an
eating disorder report a higher impact of intrusive prospective and personally
relevant imagery and more negative prospective events than healthy controls.

Secondary Objective:
The third aim of our research is to explore whether there is an association
between differences in mental imagery and PTSD-symptoms, BMI and severity of
eating disorder. Are there underlying factors that explain for differences in
intrusive prospective imagery? We expect that a higher impact of intrusive
prospective imagery is associated with more PTSD symptoms, a lower BMI and more
severe eating disorder psychopathology.

This study is a cross-sectional observational study with a semi structured
interview and three questionnaires.

Participants with eating disorder diagnosis will be recruited from Centre for
Eating Disorders Ursula. All eligible participants in this treatment centre
will be asked to participate at the beginning of their treatment. The
assessment will be part of the Routine Outcome Monitoring (ROM) at the begin of
treatment.
Participants for the control group will be asked to participate through
advertisements in the local newspaper, women magazines and at Leiden University
and Hogeschool Leiden.
The participants have to fill out three questionnaires (paper and pencil) and
will be interviewed. The total duration is approximately one hour and will take
place at Centre for Eating Disorders Ursula.

The risk and the burden of this research proposal can be considered relatively
low and involves assessment by means of a semi structured interview and
questionnaires. The semi structured interview is already used in patients as
well in healthy controls and questionnaires are already used in patients.
Previous studies using these measurements did not report any negative side
effects or risks. For patients the assessment will be additional to the Routine
Outcome Monitoring so they do not need to travel an extra time. Patients are in
treatment at the centre and if necessary a psychologist or psychiatrist is
available. People in the healthy control group need to travel once to the
research location.