Virtual food for real thought: how to improve extinction learning in anorexia nervosa

Lead Research Organisation: King's College London
Department Name: Psychological Medicine


People with anorexia nervosa experience intense anxiety at mealtimes. The anxiety is often related to the prospect of eating and to the consequences of eating, such as weight gain, unmanageable distress, and loss of control over food. Current psychological treatments of anorexia nervosa encourage patients to develop a more positive relationship with food, but rarely demonstrate the use of practical strategies to cope with food-related anxiety. This project will use exposure to virtual foods to establish which mechanisms underline successful reduction of food-related anxiety in anorexia nervosa. Exposure therapy consists of gradual exposure to a feared stimulus and is the most effective treatment of anxiety disorders. In anorexia nervosa, food exposure has been used with some degree of success to improve eating behaviour and favour weight gain. However, the mechanisms of food exposure remain largely untested with the result that there is limited understanding of the strategies that can maximise its efficacy. This study will use the novel technology of virtual reality to test the mechanisms of food exposure. Virtual reality enables researchers to create environments that resemble those experienced in real life, while ensuring high levels of control over the variables which are measured. Participants with anorexia nervosa will be exposed to a virtual kitchen and asked to explore as many virtual food items as possible. Their anxiety and behaviours in response to the virtual foods will be recorded (e.g. looking at, approaching, grabbing, holding virtual foods). Each participant will be assigned by chance to one of three conditions: exposure to virtual foods (1), exposure to virtual foods while interacting with a virtual pet showing affection (2) and exposure to virtual foods while interacting with a talking human-like character (AVATAR) that encourages to challenge the eating disorder voice (3). Following exposure to the virtual kitchen, participants will be invited to take part in a taste-test, during which anxiety and behaviours in response to a range of real foods will be measured. The hypothesis is that participants exposed to the virtual foods while also receiving positive mood induction or recovery-focused social support will experience a greater reduction in food-related anxiety and will approach food more, both in the virtual environment and during the taste-test, compared to participants who only receive exposure to virtual foods. This study will prove the efficacy of food exposure, positive mood induction and recovery-oriented social support to help people with anorexia nervosa to cope with eating-related anxiety. The clinical potential of this project is to develop a novel intervention that helps people creating a positive relationship with food. The use of virtual reality ensures that the intervention can be easily disseminated to, and used by, patients over time and in their own environments.

Technical Summary

Anorexia nervosa is a deadly psychiatric disorder primarily characterised by eating-related anxiety. No currently available psychological treatments include the provision of effective meal support for adults with anorexia nervosa. The aim of this study is to use virtual reality to test the mechanisms of extinction learning and help patients to develop a more positive relationship with food. The hypothesis is that exposure to virtual foods enhanced by positive mood induction or recovery-focused social support will be associated with greater reductions in food-related anxiety and greater caloric intake during a taste test, compared to exposure to virtual foods alone. This hypothesis is based on recent advances in the science of learning. These advances consist in the understanding that extinction of previously acquired negative associations can be enhanced by strategies that inhibit the reacquisition and return of fear and boost the development of safer associations (i.e. positive mood and social support). Adult patients with anorexia nervosa will be recruited (N=159) and randomised to one of three experimental conditions: (1) exposure to virtual foods, (2) exposure to virtual foods enhanced by positive mood induction and (3) exposure to virtual foods enhanced by recovery-focused social support. Anxiety and behaviours towards virtual foods will be measured during exposure to a virtual kitchen and caloric intake will be recorded during a taste-test. This project paves the way for the understanding of the mechanism of extinction learning in anorexia nervosa and other fear-related disorders. The use of virtual reality ensures high levels of experimental control to test causal relationships between variables, as well ecological validity and relevance. Clinical implications include the development of a novel intervention to shape the process of behaviour change around food and eating in anorexia nervosa.

Planned Impact

This project will create impact in at least three areas: 1) the understanding of the mechanism of extinction learning and exposure therapy, 2) the use of technology to scale up novel therapeutic strategies and 3) the implementation of acceptable and effective meal-support strategies for patients with eating disorders. In particular, the process and outcomes of this 24-month study will provide further insights as to whether theorised mechanisms to maximise inhibitory learning, such as positive affect and social support, are effective in achieving this goal (Craske et al., 2018). This study will also prove the concept that virtual reality enables advances in the understanding of causal mechanisms of psychopathology (e.g. the hypothesis that food cues elicit negative emotions in people with anorexia nervosa) and treatment (e.g. the hypothesis that positive mood triggers a more positive re-evaluation of food stimuli) (Riva et al., 2016). Patients' feedback will indicate whether this technology is safe and acceptable to them. Virtual reality technology enables broad dissemination and long-term use of contents. The employment of affordable strategies to scale up effective psychological support is particularly warranted in the field of mental health, where physical and psychological barriers prevent a large number of individuals from accessing evidence-based treatments (Holmes et al., 2018). This study will have a specific impact in the management of meal-related anxiety in anorexia nervosa. Meal support in eating disorders is often implemented by nurses or carers and associated with high levels of distress in those individuals, as well as patients, who describe the process of eating as a "battleground" (Long et al., 2012). This study will create knowledge as to whether a brief training of positive mood and social support can improve patients' attitudes towards foods and eating, while reducing the burden on professionals and carers.
Improvements in the understanding of extinction learning and exposure therapy, virtual reality and meal support will benefit scientists and clinicians working in the field of eating disorders and other fear-related conditions, and also service users and carers. Findings of this project will benefit the industrial sector. Our research group collaborates with Mindwave ( and Virtulware (, both of which focus on developing digital products for healthcare improvements and have expressed interest in supporting further developments of the proposed project.
Pathways to impact will involve a collaborative process of co-creation and evaluation of the study materials with professionals, service users and carers. The study materials and findings will be disseminated to members of the lay public, as well as eating disorder and mental health communities of clinicians and researchers. The strategies for dissemination will involve the use of social media (e.g. Twitter, Facebook, Instagram, online forums), presentations at conferences through posters or brief talks (e.g. conferences organised by eating disorders charities, scientific societies and academies) and open-access peer-reviewed publications. Findings will be translated in the development of meal-support protocols to test and implement in clinical services, such as the inpatient, day-care and outpatient services at South London and Maudsley NHS Foundation Trust. Findings will also inform the development of future grant proposals, with the overall aims of (1) testing a broad range of strategies to maximise extinction learning during exposure therapy, (2) include multimodal methods of assessments (e.g. neuroimaging, physiological indices) and (3) translate the experimental findings in novel clinical interventions to improve treatment efficacy in eating disorders (currently below 50%).


10 25 50