This study was possible thanks to the collaboration with the Eating Disorders Unit of the Mental Health Area of the Sant Joan de Déu Hospital in Barcelona. Thanks to all the center staff for their valuable collaboration in the project.
Introduction
Anorexia nervosa (AN) is an eating behavior disorder (ED) characterized by low weight (less than 85% of what is expected considering age and height), body image disturbances and extreme fear of gaining weight (APA, 2013). In addition to the disorder's own symptomatology, AN presents a multitude of medical complications derived from the state of malnutrition, high mortality rates and high comorbidity with other disorders, especially anxiety, depressive and personality disorders. Despite the existence of different types of intervention, there is consensus that the priority in the treatment of AN is the recovery of a healthy weight and addressing the physical complications associated with underweight. Among the goals of intervention, therefore, are weight restoration, normalization of hunger and satiety perceptions of hunger and satiety, normalization of eating patterns and overcoming malnutrition through refeeding programs. However, the extreme fear of weight gain and becoming obese, characteristic of people with AN, often produces great resistance towards normalization of intake and makes it difficult for patients to reach a healthy weight. Patients with with AN often show anxiety towards certain stimuli (such as food or the body itself) and engage in avoidance behaviors (such as food restriction). Therefore, treatment of AN can be enhanced by incorporating components that target the anxiety experienced by patients while eating and face to weight gain. This is why exposure techniques have been proposed as an effective treatment in this disorder (Hildebrandt, Bacow, Greif, & Flores, 2014; Hildebrandt, Bacow, Markella, & Loeb, 2012; Steinglass et al., 2012).
Also similar to what occurs in anxiety disorders, there is a large number of studies showing the presence of attentional biases in patients with EDs and, in particular, in patients with AN. Among this diversity of research, our group has been the first (Porras-Garcia et al., 2018; Porras-Garcia et al., 2019), and to date the only one, to use virtual reality (VR) techniques in combination with eye-tracking (ET) techniques to investigate the presence of attentional biases towards the body. The synergy between these two technological resources offers opportunities that are of great interest for the study, by means of objective indicators (such as those offered by ET instruments) in highly controlled situations and with high ecological validity (such as those achievable by VR techniques), of basic processes altered in EDs, and also, relevantly, for the design of new treatment components that allow to improve the efficacy of currently available interventions for EDs; in particular, for AN.
Combined use of Eye Tracking and Virtual Reality techniques
The methods that have been used to evaluate attentional bias in patients with ED are very diverse, the most frequent being those based on modified Stroop tasks, dot-probe, visual search and eye-tracking (ET). Also, although to a lesser extent, other methods have been used, such as dichotic listening, lexical decision, and spatial pointing tasks. A complete review of these methods and the main results obtained can be found in Jiang and Vartanian (2018). Of all these techniques, ET is the only one that allows obtaining a continuous measure of the attention dedicated to different stimuli. On the other hand, with few exceptions, most cognitive tasks involve verbal stimuli, while ET techniques allow attentional behavior to be analyzed directly on visual stimuli, more directly related to the stimuli that in natural situations provoke dysfunctional responses of the patients with ED. With the ET technique, among other variables, the first fixation point, the proportion of time spent observing each stimulus, the orientation speed (through saccadic movements), or the frequency of fixations can be quantified; offering the possibility of obtaining objective measures of the location of attention. This technique therefore facilitates the objective and direct measurement of attentional biases, while cognitive tasks such as the others mentioned above only allow these biases to be inferred from response latencies. The ET technique, on the other hand, facilitates the decomposition of visual behavior into factors that are under voluntary control and factors less susceptible to that type of control, which allows us to differentiate, with greater validity than other methods, between orientation responses. automatic (vigilance), voluntary attention responses (maintenance of attention), and avoidance responses, as well as establishing their time course.
Although the ecological validity of the evaluation of attentional biases using ET is greater than with other methods, this validity can be increased even more if this technology is combined with another, such as virtual reality, which allows carrying out the evaluation of visual attention while the participant is immersed in virtual simulations of natural situations (Hans-Martin Lutz et al., 2017). With VR, in addition, one of the problems that appear in studies in which the evaluation of attentional biases towards the body is carried out using the image reflected in a mirror is solved, and that is that the patient must replace her usual clothing by tight-fitting garments or that allow observation of one's own body without distortions due to clothing. This can produce feelings of discomfort in the patient that contaminate the results, since she must show her body in these conditions to the therapists or researchers who are conducting the study. By using VR, the patient does not need to change her wardrobe during the treatment sessions, since the avatar's characteristics are defined by the program and are independent of the clothing she wears at all times.
Video: Example of eye tracking during a body exposure, processed by OGAMA.
Modification of attentional bias related to the body through Virtual Reality
Although research on the presence of attentional biases in EDs is extensive, very little has been done to date to apply the information obtained in these studies with the aim of improving the effectiveness of available treatments. Among the few investigations carried out in this line are those by Cardi et al. (2015) and Turton, Cardi, Treasure, and Hirsch (2017), on reducing negative biases in the interpretation of social stimuli in patients with AN; and those of Boutelle, Monreal, Strong, and Amir (2016), and Schmitz and Svaldi (2017) with patients with binge eating disorder. The modification of attentional biases toward the body, in particular, has not yet been explored in patients with AN or in patients with other EDs. Although mirror exposure is frequently used to intensify cognitive-behavioral therapy (CBT), its application is aimed at extinguishing negative cognitive, emotional, and behavioral responses to one's own body (Fairburn et al., 2008), rather than to directly modify attentional processes (Renwick et al., 2013).
Taking into account the current state of knowledge, described in the previous sections, the question arises: is it possible to improve the effectiveness of the treatment of exposure to one's own body, in patients with AN, through training aimed at modifying the attentional biases they show? If the results confirm the possibility of increasing the effect of exposure treatment to reduce anxiety and fear of gaining weight in patients with AN, as a result of the modification of attentional bias, there will be a new therapeutic resource to improve the available treatments for AN, and adapt them to a greater number of patient profiles. This objective is consistent with the growing interest in what has been called “stratified medicine”, whose purpose is to adapt treatments to the individual profiles of patients (Lester and Eley, 2013).
To reduce attentional bias, a training is applied based on an adaptation of the bias induction procedure of Smeets, Jansen and Roefs (2011). It is a procedure that has shown its effectiveness in producing changes in attentional biases in healthy participants with different levels of body dissatisfaction. This procedure was adapted to produce a reduction in bias, rather than an induction of bias toward some parts of the body or others. The result, therefore, is not greater attention to certain parts of the body (e.g. attractive or unattractive, weight-related or not), but a balanced distribution of attention between different parts of the body. Furthermore, the procedure was adapted for use with VR Head-Mounted Displays (HMD) and with ET devices integrated into them, so that the training is carried out while the patient observes, through immersive VR, an avatar that corresponds to the simulation of your own image reflected in a virtual mirror. Training is carried out by selecting geometric figures (e.g. square, rectangle, circle) that approximately fit specific parts of the body. Each of these figures can have different colors. The patient must detect and identify the figures that appear in different parts of the avatar's body. The image of the avatar is tinted using a filter that reduces its sharpness by 75%, so that the patient sees a blurred image. In half of the studies, participant must discriminate the shape of the figure and in the remaining 50% the discrimination is based on color.The change in the discrimination task (shape, color) is intended to maintain motivation to perform the task. Through the ET device, changes were programmed in the avatar so that, as soon as the participant focuses her gaze on the part of the body on which the test stimulus is located (the geometric figure), the filter that reduces the sharpness of that part of the body, making it stand out from the rest of the avatar's body. The highlight stays that way for four seconds. The patient is instructed to keep his gaze on the highlighted part until the start of the next study (four seconds). Throughout the training, the test stimulus appears on some of the body parts related to weight in 45% of the studies, and in another 45% of the studies it appears on some of the body parts not related to weight. In the remaining studies (10%), the test appears on one of the three neutral stimuli located next to the avatar, outside his body (in this case, some boxes).
Reference: Smeets, E., Jansen, A., y Roefs, A. (2011). Bias for the (un) attractive self: On the role of attention in causing body (dis)satisfaction. Health Psychology, 30(3), 360.
Video: Attentional Bias Modification Task (ABMT), adapted from Smeets, Jansen, and Roefs (2011) for the project's immersive virtual reality environment.
Hypotheses and objectives
The starting hypothesis of the project is the following:
If a component of virtual-reality-based exposure to the body image, intensified through the production of the illusion of ownership of the virtual image, is added to the usual treatment of anorexia nervosa, then the treatment will be more effective. If, in addition, a virtual-reality-based component of attentional bias reduction toward the body is added, then the treatment will further increase its efficacy.
The general objectives are:
1) To develop a procedure for the reduction of attentional biases towards the body by means of ET and VR techniques.
2) To integrate this procedure in a treatment of exposure to the body aimed at reducing the fear of gaining weight suffered by patients with Anorexia Nervosa.
3) To analyze the increase in efficacy of the usual treatment by intensifying it with the addition of a body-exposure component and by the addition of an attentional bias reduction component.
The specific objectives are:
-
1.- To develop virtual reality environments, which can be presented via HMD with ET to expose patients with AN to virtual representations of their own body shape that can be modified to simulate progressive increases in the body mass index.
-
2.- To adapt the attentional bias modification procedure of Smeets, Jansen and Roefs (2011) to be used to reduce biases towards the body in patients with anorexia, by means of ET and VR techniques.
-
3.- To test the functioning of the program developed in specific objective 2 by means of its application in people without ED with different degrees of body dissatisfaction.
-
4.- To carry out the necessary modifications in the program developed in specific objective 2, after the analysis of the results of the pilot study corresponding to objective 3.
-
5.- To test the functioning of the modified program in specific objective 4 through a pilot study conducted with patients with AN. Both its general functioning and its capacity to reduce the attentional bias towards the body will be examined.
-
6.- To make the necessary modifications to the program, after analyzing the results of the pilot study corresponding to objective 5.
-
7.- To carry out a randomized controlled clinical study to analyze the efficacy of an intensification of the usual treatment of AN by adding a component of exposure to one's own body and a component of attentional bias reduction.
-
8.- To elaborate an intervention protocol, including a clinical manual for patients with anorexia based on the results of the study carried out to achieve objective 7.
-
9.- To design and publish a web site for the free dissemination of the virtual environments and procedures developed in the project.
-
10.- To disseminate the results of the studies carried out through scientific publications with impact index and presentations at conferences.