Most people with an eating disorder (ED) do not receive good treatment. The investigators have developed a new brief group treatment that is supposed to work by reducing how much women with an ED value the impossible thinness standard promoted by the media and how much they value/crave binge foods. The investigators want to test whether the treatment actually changes those two mechanisms using brain scan data, which is more objective than completing questionnaires and even interviews.

In the first phase of the study (R61), the investigators will compare women in the treatment versus those on a wait-list. If the investigators can show that the treatment "works" (does what the investigators think it does) compared to no active treatment (women will be allowed to seek and receive outside help but investigators will not provide it until after the wait-list), investigators will conduct the second phase of study (R33),where they will randomly assign women with an ED to either the new treatment or to a group treatment that represents what many college mental health clinics provide to their clients with ED.

Only 3-20% of individuals with eating disorders (EDs) receive treatment, and those who do rarely receive evidence-based treatments because they are very intensive and costly, few clinicians deliver them, and they differ for the various EDs. These factors have hindered broad implementation of evidence-based treatments, implying that a brief frontline outpatient treatment for a range of EDs that could be easily, cheaply, and widely implemented would address a major public health problem. There is also limited experimental evidence regarding factors that maintain EDs and mechanisms of action for ED treatments. The investigators propose to conduct a project that translates basic persuasion research into clinical hypotheses and a novel treatment.

The investigators hypothesize that EDs are maintained by (1) excessive valuation of the thin beauty ideal, which prompts caloric restriction and other unhealthy weight control behaviors (vomiting, laxative use, excessive exercise) that increase binge eating risk, and (2) excessive valuation of high-calorie foods, which maintains binge eating. The investigators propose to evaluate a treatment that addresses these hypothesized maintenance processes. Two pilot trials evaluated a novel 8-session group dissonance-based treatment (Counter Attitudinal Therapy; CAT) wherein women with any ED appropriate for outpatient care completed activities in which they collectively discuss costs of pursuing the thin ideal and the ED behaviors in which they personally engage (unhealthy weight control behaviors and binge eating), which putatively creates dissonance about engaging in those particular behaviors that reduce valuation of the thin ideal and high-calorie binge foods.

During the COVID-19 shelter-at-home orders taking place at both research sites, the investigators will not measure in-person only outcomes including fMRI, ECG, computerized implicit associated tests, and height and weight measurements for all participants that have assessments during this order. The investigators will continue to conduct treatment groups that will be administered on a virtual platform.

Inclusion Criteria:

• DSM-5 eating disorder
• Must have a primary care doctor

Exclusion Criteria:

• Non-English speakers
• BMI <75% ideal body weight
• Current acute suicidal ideation (defined as thoughts of a specific method or plan)
• Comorbid psychiatric disorder that would disrupt groups (e.g., bipolar disorder, substance misuse)
• Serious medical problems (e.g., diabetes)

Study data will be shared via the National Database for Clinical Trials Related to Mental Illness (NDCT), which is part of the NIMH Data Archive (NDA). All data, with the exception of video recordings of the participants in treatment, will be provided.

Descriptive/raw data will be submitted while the study is ongoing (every 6 months). Analyzed data will be submitted either when results are published or after the completion of the award period, whichever comes first.