Anorexia nervosa (AN) has a lifetime prevalence of between 0.5 and 2%, with a peak age of onset between 13 and 18 years (1). AN is characterized by abnormally low body weight, an intense fear of gaining weight, and a disturbed perception of the own body (2). AN is associated with poor outcomes: Mortality rates are high, and after treatment, less than one half of patients recover, whereas 20% remain chronically ill (3). Specifically, adolescent AN patients show high hospital readmission rates [45% after inpatient treatment (4)].

Body image (BI) is a multifaceted construct including perceptual, affective, cognitive, and behavioral components referring to the own and others' bodies (5). BI disturbances are diagnostic features and stabilizing factors in eating disorders (EDs): patients overestimate their shape and weight (6). Regarding behavioral aspects of BI disturbances, typical features in AN are body checking (repeated measuring and checking of shape and weight) and body avoidance [e.g., covering up the body shape by wearing baggy clothes or avoiding looking at oneself in the mirror (7)]. Despite much fewer studies on adolescents than on adults, a recent systematic review reported more BI disturbances in adolescents with AN and bulimia nervosa (BN) compared to healthy controls (8), similar to findings in adults. Alongside, Legenbauer, Thiemann (5) point out age-dependent differences between children, adolescents, and adults and the need for more studies for the still underresearched group of children and adolescents with eating disorders.

Disturbances of BI in EDs are thought of as a type of cognitive bias, with attentional bias (AB) being a form of biased information processing (9, 10). Evidence suggests that individuals with EDs demonstrate an AB regarding body shape-related stimuli, as examined by, e.g., the Stroop and Dot Probe task (11). ED patients not only show an AB for body-related information (i.e., words), but also for specific body parts (12). Especially, a bias toward negatively rated or most unattractive body parts was found in adolescents with eating disorders (13). This is in contrast to women without an ED diagnosis who showed a balanced attentional pattern regarding self-rated positive and negative body parts of their own body (14). Inducing a temporary AB for self-defined unattractive body parts leads to a significant decrease in body satisfaction in young women without EDs (15). This was confirmed by Tuschen-Caffier and colleagues (16) who found AB to be correlated with body dissatisfaction in women with AN and BN. In a previous study comprising of adolescent ED patients and a healthy control group, we assessed visual attentional processes via eye tracking while watching disease-specific pictures of female bodies (17). While all girls attended more to specific body parts (e.g., hips, upper legs), eating-disordered girls showed an AB toward unclothed body parts, which could be a behavioral marker for overestimation of shape and weight (17).

Based on the relevance of BI disturbances for the maintenance of EDs, it is crucial to target these explicitly (8). Particularly exposure interventions are highly recommended for ED treatment (8, 18, 19). Repeated body exposures in front of a full-length mirror (mirror exposure, ME), with patients wearing tight clothing or underwear to view the body shape and size as best possible, is a commonly used and recommended practice in ED treatment (20). This is applicable to both AN patients approaching a healthy weight and patients with BI disturbances without the diagnosis of AN (20). Former studies have shown ME as a one-time exposure or as part of a more comprehensive treatment program as being effective in the reduction of body dissatisfaction, body checking and body avoidance in adults (19, 21, 22) and adolescents (23) with AN or high body dissatisfaction. Besides ME, other body exposure interventions [e.g., exposure using video recording or virtual reality (24)] have been investigated and evaluated for treating patients with AN in a few studies, with cautiously positive results regarding improvement in BI facets (25). Altogether, evidence for body exposure interventions is still insufficient, especially for adolescent patients—results are mixed, and which patient groups benefit most from which form of body exposure, as well as the mode of action, are unclear (20, 25). Inter alia, modification of the AB and cognitive-affective habituation as along with habituation of physiological processes are discussed as potential underlying mechanisms (20).

Individuals with EDs (mainly investigated for BN) and highly body dissatisfied women show negative cognitive, emotional as well as altered physiological responses during the exposure to their own body [e.g., (26–28)]. For the assessment of stress reactivity, non-invasive measurements are typically used, e.g., heart rate (HR) and heart rate variability [HRV (29)] as sympathetic stress markers. These are affected by emotional stimuli (30) or stress (31). Studies have shown that individuals with AN have a consistently elevated HRV compared to healthy controls, which seems to normalize during recovery (32). Furthermore, salivary alpha amylase (sAA) is a marker of acute stress (33), highly correlates with blood norepinephrine levels (34, 35), and is thereby a further marker of the central sympathetic nervous system [SNS (36, 37)]. Lastly, cortisol, which is a correlate of an activated hypothalamic-pituitary-adrenal (HPA) axis, has been extensively investigated in EDs, showing both altered basal levels and reactivity (38). A study by Monteleone and colleagues (39) showed that individuals with AN exhibit an asymmetry between the HPA axis and the SNS response to a stressor, with the underweight AN group showing a strong cortisol reactivity but an almost completely lacking increase of sAA (comparable to the control groups).

While investigating stress reactivity to body exposure in women with BN who obtained six therapeutic sessions of ME, a decrease of salivary cortisol levels from the initial to the final intervention session was observed; this could be interpreted as habituation of the neuroendocrine stress response (40). Furthermore, a study in women with binge eating disorder found interactions between induced stress and body dissatisfaction during ME by measuring the biological response to social stress via salivary cortisol and sAA (41). Additionally, ED patients showed higher cortisol levels than control participants during a single 40-min body exposure task in a study exploring neuroendocrine (salivary cortisol) and physiological (skin conductance, heart rate) responses in women with EDs, compared to healthy controls. However, no group differences regarding neuroendocrine or physiological reactivity occurred (42). Investigating the effect of a three-session mirror exposure on psychophysiological sympathetic reactivity (HR, HRV, skin conductance) in women with BN vs. healthy controls showed an initial response within the sessions in both groups, but no physiological habituation between the sessions; thereby, the initial response was an unexpected decreasing sympathetic activity which might be explained by the stronger feelings of sadness compared to fear (28). So, until now, exact physiological mechanisms which go along with repeated body exposure in patients with AN or high body dissatisfaction are not yet clarified, with no study, to the authors' knowledge, investigating these in adolescents.

Due to the high association of BI disturbances and the development and maintenance of EDs, including AN, psychotherapy with patients with AN in any setting should include modules treating body image disturbances explicitly (8). A need for new, emerging treatments with a focus on brain-based interventions as well as mechanism-based approaches, which are specifically addressing psychopathology and BI disturbances, have been proclaimed (43, 44). While there is data available on the efficacy of body exposure interventions for reducing BI disturbances in women with EDs and high body dissatisfaction (20, 25), there is a lack of available data regarding the efficacy of body exposures with children and adolescents. Furthermore, different EDs are frequently mixed up in the literature, making it difficult to investigate specific processes, which might differentiate between body dissatisfied individuals with EDs and body dissatisfied individuals without EDs. Also, underlying psychophysiological and neurobiological processes during body exposure are still poorly understood. Therefore, the aim of this study is to investigate the short-term efficacy of a computer-based body exposure intervention for adolescents with AN. This group is being compared to an AN-TAU group and a body dissatisfied clinical control group.

The primary research question is: Is body dissatisfaction in adolescents with AN reduced after completing a four-session computer based body exposure intervention, compared to TAU?

Secondary research questions include: (1) Body checking and avoidance: Can four sessions of a computer based body exposure intervention reduce body checking and body avoidance behavior, compared to TAU? (2) Viewing patterns / attentional bias (AB): Can the AB in terms of focusing more on negatively rated body parts, measured via eye tracking, be reduced by a four-session computer based intervention in adolescents with AN, compared to TAU? Are there AN-specific AB modifications in comparison to a body dissatisfied control group also obtaining the intervention? (3) Stress reactivity: Do four sessions of body exposure modify stress reactivity in terms of a physiological habituation, measured via salivary cortisol, sAA, HR and HRV? Are there AN-specific effects in physiological habituation in comparison to a body dissatisfied clinical control group?

The multi-level approach of this study allows for both an objective and a subjective evaluation of the short-term effects of the proposed computer based body exposure intervention. It will give new insights into the effects and underlying processes of the underresearched field of body exposure in adolescents with AN. The collected data will be assessing different aspects of body dissatisfaction and general psychopathology by employing questionnaires, measuring AB, utilizing emotion ratings, recording HR and HRV, as well as determining salivary cortisol and sAA levels, throughout the exposure sessions. Besides comparing the pre-post intervention effect with TAU within an RCT, this study will also investigate how the processes that underlie body exposure change throughout the intervention and if there are AN-specific modifications in comparison to adolescents with high body dissatisfaction (CG-INT). Corresponding with the current literature regarding body exposure interventions, especially for adolescents with AN, we expect psychopathology (body dissatisfaction, body image avoidance, body checking) to be reduced at the end of the intervention compared to the beginning of the intervention and with TAU and the AB to weaken (shown by spending less time looking at the negatively rated body parts). Furthermore, we expect a physiological habituation, both within and across sessions, with a decrease in emotion ratings, HR and neuroendocrine stress levels. To put it in a nutshell, results of this pilot study will give new insights into the effectiveness of and the processes underlying body exposure interventions in children, adolescents, and young women. Despite a relatively small sample size, a comprehensive in-depth analysis, implemented by multi-level assessment, will be carried out, thereby expanding the so far weak literature on mechanisms of BI disturbances and specific interventions for the abovementioned patient groups. The results of the current pilot study will be used for the design of a subsequent larger, multi-center intervention study with adolescents with AN, helping e.g., to identify modifiable BI and psychopathologic outcome measures, to select relevant and modifiable stress markers, and to provide orientation toward expectable effect sizes.

VS, OK, GM, and SH initiated and designed the study. VS, LS, and SH wrote and prepared the manuscript. All authors reviewed the manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.