How Do Eating Disorders Develop and Persist?

Vulnerability, maintaining processes, social context and the meaning of recovery

Level: Intermediate
Reading load: 10 core readings in 5 units
Prerequisites: Familiarity with the distinction between describing a disorder and explaining it; What Is a Mental Disorder? provides a starting point. Some experience comparing observational and experimental evidence. How Do Psychological Disorders Develop and Persist? provides a useful foundation.
Last reviewed: September 2026

Introduction

An eating difficulty may persist despite distress and harm, but the appearance of persistence does not reveal its cause. Self-evaluation, learned choices, sensory experience, emotion, biological vulnerability, and social conditions suggest different processes. Their relevance may vary across presentations. An explanation must also allow that the difference it measures could be a consequence of the difficulty rather than its origin.

This path begins with two frameworks whose comparison prevents weight and shape concerns from becoming a universal explanation. It then moves through genetic and longitudinal biological evidence, choices and momentary emotion, developmental risk and cultural change. The final pair asks what recovery means and what a long follow-up can establish. Throughout, distinguish a diagnostic label, an observed pattern and a proposed mechanism.

The aim is to evaluate explanations, not to infer a diagnosis or future course from a person’s appearance. Body size, nutritional state, symptoms, functioning and recovery are related but distinct. The more technical papers introduce genetic and meta-analytic methods; their guides emphasize the comparisons and assumptions that carry the argument.

Unit 1 of 5

What needs explaining across eating difficulties?

A shared mechanism can explain why presentations overlap, but it can also obscure meaningful differences. Fairburn and colleagues propose a maintenance account centered on self-evaluation. Thomas and colleagues consider avoidant/restrictive presentations that require a different starting point. Compare the scope and predictions of the frameworks before treating either as a general explanation.

Reading
1 of 10
A Transdiagnostic Theory of Eating Disorders
Fairburn, C. G., Cooper, Z., & Shafran, R. (2003). Cognitive behaviour therapy for eating disorders: A “transdiagnostic” theory and treatment. Behaviour Research and Therapy, 41(5), 509–528. https://doi.org/10.1016/s0005-7967(02)00088-8
Foundational argument · Moderate · Full work: 20 pages
Follow the distinction between common maintaining processes and the additional processes proposed for particular people.
Why you’re reading it. Establish a testable maintenance framework before comparing it with a different account of restriction.
A question to consider. Choose one arrow in the model. What observation could distinguish that causal claim from two symptoms changing together?
Reading
2 of 10
A Dimensional Model of Avoidant/Restrictive Eating
Thomas, J. J., Lawson, E. A., Micali, N., Misra, M., Deckersbach, T., & Eddy, K. T. (2017). Avoidant/restrictive food intake disorder: A three-dimensional model of neurobiology with implications for etiology and treatment. Current Psychiatry Reports, 19(8), Article 54. https://doi.org/10.1007/s11920-017-0795-5
Foundational argument · Moderate
Compare these overlapping dimensions with the self-evaluation framework. Similar eating patterns may have different meanings.
Why you’re reading it. Make the limits of a single transdiagnostic explanation part of the syllabus from the outset.
A question to consider. For a hypothetical pattern centered on fear of an aversive eating consequence, which predictions would differ from a weight-and-shape account?

Unit 2 of 5

Which biological differences are vulnerabilities, consequences or both?

A biological association does not announce its place in a causal sequence. Genetic associations and repeated brain measurements offer different kinds of leverage. Use their contrast to ask what was measured before an illness could alter it, what changes with a person’s condition, and what neither design can settle.

Reading
3 of 10
Genetic Associations in Anorexia Nervosa
Watson, H. J., Yilmaz, Z., Thornton, L. M., Hübel, C., Coleman, J. R. I., Gaspar, H. A., Bryois, J., Hinney, A., Leppä, V. M., Mattheisen, M., Medland, S. E., Ripke, S., Yao, S., Giusti-Rodríguez, P., Anorexia Nervosa Genetics Initiative, Hanscombe, K. B., Purves, K. L., Eating Disorders Working Group of the Psychiatric Genomics Consortium, Adan, R. A. H., … Bulik, C. M. (2019). Genome-wide association study identifies eight risk loci and implicates metabo-psychiatric origins for anorexia nervosa. Nature Genetics, 51(8), 1207–1214. https://doi.org/10.1038/s41588-019-0439-2
Landmark empirical study · Methodological contribution · Challenging · Full work: 8 pages
Follow the move from associated variants to psychiatric and metabolic overlap, keeping those claims distinct from individual prediction.
Why you’re reading it. Examine vulnerability through evidence that is not simply a snapshot of current nutritional state.
A question to consider. Could a genetic correlation fit two different causal pathways? Specify what additional evidence would distinguish them.
Reading
4 of 10
Brain Structure During Recovery From Anorexia Nervosa
Kaufmann, L.-K., Hänggi, J., Jäncke, L., Baur, V., Piccirelli, M., Kollias, S., Schnyder, U., Martin-Soelch, C., & Milos, G. (2020). Age influences structural brain restoration during weight gain therapy in anorexia nervosa. Translational Psychiatry, 10(1), Article 126. https://doi.org/10.1038/s41398-020-0809-7
Methodological contribution · Moderate · Full work: 10 pages
Compare repeated structural measurements with the preceding genetic study. Treatment, time and nutritional change occur together here.
Why you’re reading it. Test how within-person change affects the interpretation of a biological difference.
A question to consider. Which inference about vulnerability would remain unresolved even if a brain measure returned to the control-group range?

Unit 3 of 5

What maintains restrictive choices and binge eating?

A proposed mechanism needs a comparison that could distinguish it from alternatives. Foerde and colleagues examine choices across clinical and nonclinical groups; Borm and colleagues compare emotional levels and trajectories around episodes. In both, attend to the gap between an informative measurement and a demonstrated cause.

Reading
5 of 10
Restrictive Eating Across Clinical and Nonclinical Groups
Foerde, K., Schebendach, J. E., Davis, L., Daw, N., Walsh, B. T., Shohamy, D., & Steinglass, J. E. (2022). Restrictive eating across a spectrum from healthy to unhealthy: Behavioral and neural mechanisms. Psychological Medicine, 52(9), 1755–1764. https://doi.org/10.1017/s0033291720003542
Methodological contribution · Contemporary update · Moderate · Full work: 10 pages
Track where the behavioral and neural measures agree about group differences and where they diverge.
Why you’re reading it. Examine a maintaining-process proposal without equating restrictive behavior, low weight and habit.
A question to consider. What result from an outcome-devaluation study would add information that this food-choice comparison cannot provide?
Reading
6 of 10
Binge Eating and Emotion Over Time
Borm, I. M., Hartmann, S., Barnow, S., & Pruessner, L. (2025). Binge eating as emotion regulation? A meta-analysis of ecological momentary assessment studies. Clinical Psychology Review, 121, Article 102625. https://doi.org/10.1016/j.cpr.2025.102625
Major synthesis · Contemporary update · Challenging · Full work: 19 pages
Name the reference point and time window behind each emotional comparison.
Why you’re reading it. Learn why different analyses can give apparently conflicting answers to an emotion-regulation account.
A question to consider. Draw a possible emotional trajectory that fits a worse post-episode level and a later decline. What causal question would still be unanswered?

Unit 5 of 5

What counts as recovery over time?

An explanatory account must leave room for change, but studies can count change differently. Bardone-Cone and colleagues make recovery criteria explicit; Eddy and colleagues examine course across decades using a different definition. Compare the outcomes before comparing percentages, and keep a population estimate separate from an individual future.

Reading
9 of 10
What Counts as Recovery From an Eating Disorder?
Bardone-Cone, A. M., Harney, M. B., Maldonado, C. R., Lawson, M. A., Robinson, D. P., Smith, R., & Tosh, A. (2010). Defining recovery from an eating disorder: Conceptualization, validation, and examination of psychosocial functioning and psychiatric comorbidity. Behaviour Research and Therapy, 48(3), 194–202. https://doi.org/10.1016/j.brat.2009.11.001
Methodological contribution · Moderate · Full work: 9 pages
Separate physical, behavioral and psychological criteria from the additional measures used to evaluate recovery.
Why you’re reading it. Define the endpoint before interpreting evidence about how often or how quickly it occurs.
A question to consider. Describe a hypothetical pattern of improvement that meets some criteria but not others. Which meaningful change would a binary label hide?
Reading
10 of 10
Eating-Disorder Recovery at Long-Term Follow-Up
Eddy, K. T., Tabri, N., Thomas, J. J., Murray, H. B., Keshaviah, A., Hastings, E., Edkins, K., Krishna, M., Herzog, D. B., Keel, P. K., & Franko, D. L. (2017). Recovery from anorexia nervosa and bulimia nervosa at 22-year follow-up. The Journal of Clinical Psychiatry, 78(2), 184–189. https://doi.org/10.4088/jcp.15m10393
Landmark empirical study · Moderate · Full work: 6 pages
Compare this study’s symptom-and-duration definition with the preceding recovery criteria, and trace who enters its denominator.
Why you’re reading it. Examine long-term change without assuming that recovery rates are interchangeable or individually predictive.
A question to consider. What information would be needed to estimate recovery under Bardone-Cone’s definition in this cohort, and which missing observations would remain a problem?

Scope and omissions

This path examines explanations of development, persistence and recovery across selected restrictive and binge-eating presentations. ARFID is an early comparison, not an exception added after a universal weight-and-shape theory. The selection connects psychological, biological and social questions without treating any one level as sufficient.

The core evidence is disproportionately drawn from girls and women, Western clinical or school samples, and European-ancestry genetic data. The Fiji study expands the contextual comparison but cannot stand for all cultures. The ARFID framework raises questions relevant beyond those samples; it does not remove the collection’s empirical gaps. Men, gender-diverse people, older adults, underrepresented communities and experiences of unequal access to care deserve further study.

This is not a diagnostic manual, a comprehensive account of every eating or feeding presentation, or a treatment guide. Detailed medical management and comparative treatment delivery lie outside its scope. The optional works extend measurement, historical evidence, habit theory and dimensional assessment without adding required steps to the core route.

Synthesis

Eating difficulties call for explanations that distinguish presentations and follow change through time. Weight and shape concerns cannot organize every case. Psychological, biological, and social observations may suggest maintaining processes, while consequences can be mistaken for prior vulnerabilities. Recovery adds another set of questions about symptoms, nutritional state, functioning, and valued life. These outcomes may improve at different rates and require distinct evidence.

Develop an explanation of a fictional pattern of eating difficulty without specifying harmful procedures. Compare two plausible maintaining processes. State one observation that both accounts predict and one that could distinguish them. Identify a biological or psychological consequence that might be mistaken for a vulnerability.

Then change the case so that weight and shape concerns are absent. Which parts of the explanation survive, and which need revision? Add a social condition without assuming that its association establishes its causal role. Propose a design that could clarify one link, explaining both its comparison and its limitations.

Finally, define two meaningful outcomes that could improve at different rates. Explain how your definition, follow-up duration, missing participants and denominator would shape an estimate of recovery. State what even a well-conducted group study could not establish about a particular person.

To revisit the distinction between diagnosis and explanation, return to What Is a Mental Disorder?. The Psychopathology area brings together the other pathways.

For a cross-diagnostic examination of explanation and forecasting, continue with Why Are Suicidal Thoughts and Actions Difficult to Explain and Predict?. It distinguishes outcomes, proposed mechanisms, individual predictions and evidence that prevention changes a population’s experience.