Is a 20-Session Treatment Sufficient for Adult Anorexia Nervosa? A Critical Comparison of CBT-AN-20 with CBT-E

 

Introduction

The search for effective, acceptable and scalable outpatient treatments for adults with anorexia nervosa remains a major clinical and service priority. Existing psychological treatments are relatively lengthy, non-completion is common, and full remission is achieved by only a minority of patients in many studies.

A recent study provided a preliminary evaluation of CBT-AN-20 (Duggan et al., 2026), a 20-session cognitive-behavioural treatment for medically stable adult outpatients with anorexia nervosa (Waller et al., 2025). The
intervention was developed from the logic underlying CBT-T, a brief treatment originally designed for non-underweight eating disorders (Waller et al., 2019). CBT-AN-20 aims to retain key cognitive-behavioural procedures while accelerating behavioural change and reducing treatment duration.

The current study addresses an important question: can adults with anorexia nervosa obtain clinically meaningful change in about half the number of sessions generally recommended in enhanced cognitive behaviour therapy (CBT-E) for underweight patients (Fairburn, 2008)? The answer cannot be derived from raw outcome similarity alone because the available studies differ in baseline severity, design, treatment setting, outcome definitions, statistical methods and follow-up.

This article therefore aims to: (1) clarify the conceptual differences between CBT-AN-20, CBT-T and CBT-E; (2) compare CBT-AN-20 outcomes with the main adult outpatient CBT-E studies; and (3) identify clinical and research implications.

CBT-AN-20: main clinical features

CBT-AN-20 is a 20-session outpatient treatment for adults with anorexia nervosa who are medically stable and have a BMI over 15 (Waller et al., 2025). It places strong emphasis on behavioural change from the beginning of treatment. Patients are expected to increase their food intake, regain weight, accept open weighing and monitor eating.

Progress is formally reviewed around Sessions 6-8. Treatment may be discontinued when patients are unwilling or unable to engage in central procedures. Later sessions include behavioural experiments, exposure to feared foods and weight-related situations, body-image work, work on emotional triggers and relapse prevention.

Conceptual differences between CBT-AN-20 and CBT-E

CBT-AN-20 and CBT-E differ in their conceptualisation of the core psychopathology of anorexia nervosa, which leads to differences in treatment duration and methods. CBT-AN-20 emphasises anxiety, avoidance, and fear-driven learning, viewing dietary restriction, avoidance of feared foods, and resistance to weight gain as strategies to limit exposure to feared outcomes such as weight gain, fatness, or loss of control. The treatment prioritises rapid reversal of avoidance through increased eating, weight restoration, open weighing, and exposure to food-, weight-, and body-related fears. Early behavioural change is considered essential for the development of new inhibitory learning.

CBT-E does not identify anxiety as the core psychopathology of anorexia nervosa. From this perspective, anxiety alone does not sufficiently account for the full phenomenology of the disorder. In contrast to simple phobic disorders, where the feared stimulus is experienced as aversive and avoidance reduces distress, individuals with anorexia nervosa may experience control over eating, shape, and weight as sources of achievement, mastery, superiority, pride, and self-control. Weight loss and dietary restraint can be positively valued, not solely as mechanisms to escape anxiety. Patients may interpret hunger, rigid control, and low weight as evidence of personal strength, and related preoccupations may be perceived as useful for maintaining control. These egosyntonic features are challenging to explain if anorexia nervosa is conceptualised primarily as a fear-based or avoidance disorder (Dalle Grave et al., 2019).

The transdiagnostic cognitive-behavioural theory underlying CBT-E posits that the core psychopathology is the over-evaluation of shape, weight, and control over eating. Individuals with anorexia nervosa judge their self-worth predominantly, and at times almost exclusively, on these factors. Behaviours such as strict dieting, excessive exercise, body checking, body avoidance, and other weight-control practices are understood within this framework, as successful control provides a significant foundation for self-evaluation. While anxiety about food and weight gain is clinically relevant, it is viewed as a secondary manifestation of this broader psychopathology rather than its primary cause. Fear intensifies because weight gain threatens an area on which personal value is excessively dependent.

This conceptual distinction has significant implications for treatment. In CBT-AN-20, exposure is a central component. Patients are encouraged to confront feared foods, increase food intake, gain weight, and engage in body-related situations while tolerating substantial anxiety, with the aim of reducing avoidance and facilitating new learning. The model assumes that reducing fear or increasing tolerance of fear is a primary pathway to recovery.

CBT-E also targets avoided foods, feared weight gain, and body-related situations, but does not structure treatment primarily around anxiety reduction. Behavioural interventions are selected based on an individualised formulation and are intended to modify the processes that maintain the eating disorder. Patients may gradually reintroduce avoided foods, decrease body checking, or regain weight; however, these changes are conceptualized as a strategy to test predictions, challenge rigid rules, reduce the preoccupations with shape, weight, and eating control, and diminish the dominance of shape, weight, and eating control in self-evaluation. The principal therapeutic goal is not merely to demonstrate anxiety tolerance, but to assist the patient in developing a more nuanced and adaptive system of self-evaluation.

The two treatments differ in their pace and approach to weight restoration. CBT-AN-20 consists of 20 sessions and aims to rapidly initiate nutritional and behavioural changes, with early emphasis on increased food intake and weight regain. Early progress is considered essential for corrective learning within this brief intervention. In contrast, CBT-E for underweight patients typically involves approximately 40 sessions over 40 weeks, providing more time to engage the patient, develop an individualised plan, assess the risks of remaining underweight, and collaboratively decide on weight restoration. Nutritional rehabilitation is integrated with the treatment of eating disorder psychopathology, followed by consolidation.

The role of patient control differs between the two approaches. Exposure-based methods such as CBT-AN-20 encourage patients to confront high levels of anxiety to maximise inhibitory learning (Craske et al., 2014). In contrast, CBT-E emphasises predictable, collaborative behavioural change that aligns with the patient's individualised formulation. For example, planning food intake and weight restoration enables patients to observe the relationship between eating and weight, maintaining their active involvement and sense of control throughout the process. This approach promotes cognitive change without external conflicts over food and weight.

Nevertheless, the contrast between the two models should not be overstated. CBT-E recognises that anxiety, avoidance and safety behaviours may contribute substantially to the maintenance of anorexia nervosa. Conversely, CBT-AN-20 includes cognitive restructuring, behavioural experiments and body-image interventions that extend beyond exposure alone. Both treatments address restrictive eating, low weight, feared foods, weighing, body-image difficulties and relapse prevention.

The principal difference concerns the mechanism considered to organise the disorder. CBT-AN-20 places fear, avoidance and inhibitory learning at the centre of treatment and seeks to produce rapid corrective learning through early behavioural change. CBT-E locates these phenomena within a broader and often egosyntonic system in which shape, weight, eating and their control have become central to personal value.

Comparison of CBT-AN-20 study with CBT-E studies

Although the study of Duggan and colleagues is the only effectiveness study of CBT-AN-20 and included a comparison with CBT-E, we consider any conclusions about the relative outcomes of the two treatments to be premature. CBT-AN-20 has so far been evaluated in this single preliminary effectiveness study involving adult outpatients (Duggan et al., 2026), whereas CBT-E has been examined in the same population in a randomised controlled trial (Byrne et al., 2017) and several cohort studies. Nevertheless, for completeness, Table 1 summarises the findings of all available outpatient studies of CBT-AN-20 and CBT-E in adults with anorexia nervosa.

CBT-AN-20 did not show a clear advantage over CBT-E in treatment completion, suggesting that a shorter intervention does not necessarily reduce dropout. Its short-term BMI increase was comparable with that observed in some CBT-E studies, indicating potential efficiency in patients with relatively higher baseline BMI. However, CBT-E produced greater weight restoration in more severely underweight samples and higher rates of clinically significant improvement in some cohorts. Most importantly, the limited and inconsistent follow-up data for CBT-AN-20 do not yet establish whether its short-term benefits are maintained over time.

 

Study	Design and adult sample	Treatment	Completion	BMI outcome	Psychopathology / categorical outcome	Follow-up
CBT-AN-20
Duggan et al. (2026), CBT-AN-20	Naturalistic outpatient case series; 61 adults; baseline ITT BMI 17.59	20 sessions	35/61 (57.4%); 28/51 (54.9%) among those starting below BMI 18.5	ITT 17.59 to 19.45 (+1.86); completers 17.67 to 19.62 (about +1.95)	EDE-Q ITT 3.40 to 1.30; completers 3.30 to 1.05. Among completers starting below BMI 18.5, 64.3% reached BMI >=18.5. EDE-Q <2.77: 56.5% among those initially clinical; 65.7% across all completers	Non-uniform 1-, 3- or 6-month follow-up; substantial missing data
CBT-E
Fairburn et al. (2013), total UK-Italy	Prospective two-site cohort; 99 adults; baseline BMI 16.1	Focused CBT-E; 40 sessions over 40 weeks	63/99 (63.6%)	ITT 16.1 to 17.9 (+1.8); completers 16.0 to 18.8 (+2.77)	Completer EDE-Q 3.19 to 1.30; 61.9% reached BMI >=18.5; 88.7% had EDE-Q <2.77	Closed 60-week follow-up
Fairburn et al. (2013), Italian sample	49 adults; markedly underweight; baseline BMI 15.7 ITT and 15.5 among completers	Focused CBT-E; 40 sessions over 40 weeks	32/49 (65.3%)	ITT 15.7 to 17.7 (+2.0); completers 15.5 to 18.6 (about +3.1)	Completer EDE-Q 2.74 to 1.15; 53.1% reached BMI >=18.5; 87.1% had EDE-Q <2.77	At 60 weeks: BMI 18.3; 40.7% BMI >=18.5; 80.8% EDE-Q <2.77
Byrne et al. (2017) CBT-E arm	Multicentre RCT; 39 adults; baseline BMI 16.59	CBT-E; 25-40 sessions over 10 months	26/39 (66.7%)	Estimated ITT increase +2.10 post-treatment and +2.35 at 12 months	Global EDE 3.52 to 2.13 post-treatment and 1.93 at 12 months; 59.0% BMI >18.5, 48.7% Global EDE <1.81, and 30.8% strict remission at 12 months	6- and 12-month follow-up; boosters and additional treatment permitted
Dalle Grave et al. (2018), DSM-5 severity study	128 adult women: 64 outpatient and 64 intensive CBT-E; mean BMI 15.5; 38.3% extreme AN	Outpatient or intensive CBT-E selected according to clinical need	93/128 (72.7%) overall	BMI improved across all severity categories; severe and extreme groups initially showed faster BMI change	No significant differences among mild, moderate, severe and extreme groups in weight recovery or good outcome	6- and 12-month follow-up; categorical outcomes combine settings
Calugi et al. (2021), severe and extreme AN	Naturalistic outpatient case series; 30 patients aged 17-48; BMI <16; mean BMI 15.1; 63.3% extreme AN	Outpatient CBT-E; 40 sessions over 40 weeks	20/30 (66.7%)	Completers 15.1 to 18.8 (+3.7); BMI 19.0 at both follow-ups	EDE-Q global 3.5 to 1.0; 55% full response (BMI >=18.5 and EDE-Q <2.77)	20- and 60-week follow-ups; changes maintained
Dalle Grave et al. (2023) adults aged 18-25	Adult subgroup n=54; mean age 20.5; baseline BMI 16.1	Outpatient CBT-E; 40 sessions over 40 weeks	28/54 (51.9%)	ITT 16.1 to 19.3 at EOT and 19.6 at 20-week follow-up	EDE-Q global 3.1 to 0.9 at EOT and 1.0 at follow-up. Completer full response 53.6% at EOT and 52.6% at follow-up; adult ITT full response 54.1% and 54.4%	20-week follow-up

 

Clinical and service implications

CBT-AN-20 addresses an important clinical and service need. A treatment delivered in 20 sessions could increase service capacity, reduce waiting times, and allow more patients to receive specialist outpatient care. Its preliminary findings, if confirmed in subsequent studies, suggest that clinically meaningful weight gain and substantial reductions in eating-disorder psychopathology can occur over a relatively short period in some medically stable adults with anorexia nervosa. However, the population in which CBT-AN-20 has been evaluated must be clearly defined. Although the study of Duggan et al. (2026) admitted patients with a BMI between 15 and 19, the mean baseline BMI was 17.59, indicating that the treatment was tested predominantly in patients with mild rather than moderate, severe, or extreme underweight, according to the DSM-5 severity specifier classification (American Psychiatric Association, 2013).

This distinction is clinically important. The findings cannot presently be generalised to many patients with anorexia nervosa and lower BMI. By comparison, outpatient CBT-E has also been evaluated in substantially more underweight samples, including patients with a mean baseline BMI of 15.1 and a high proportion meeting criteria for extreme anorexia nervosa. These patients achieved greater absolute increases in BMI during the 40-session treatment, with improvements maintained at 20- and 60-week follow-ups.

CBT-AN-20 may therefore prove most suitable for a selected subgroup of medically stable adults with mild underweight who are able and willing to engage rapidly in increased eating, weight regain, open weighing and dietary monitoring. Its strong emphasis on immediate behavioural change may be advantageous for patients who are ready to act and who respond positively to a focused, exposure-oriented intervention. Conversely, it may be less acceptable to patients who require more time to develop engagement, understand the personal functions of the eating disorder or make an autonomous decision to regain weight. Indeed, the early behavioural demands of CBT-AN-20 may accelerate change among those who remain engaged while contributing to withdrawal among patients who are not yet prepared to undertake these changes.

Clinical efficiency should therefore not be equated simply with brevity. A shorter intervention is genuinely efficient only if it produces comparable rates of treatment completion, adequate weight restoration, minimal residual eating-disorder psychopathology, functional recovery and sustained remission without increasing relapse or the subsequent use of additional treatment. The relevant service-level outcome is not the cost per session or per treatment episode, but the cost per patient who achieves and maintains a clinically meaningful recovery. A brief intervention that requires frequent extension, renewed treatment or later intensive care may not reduce costs across the full care pathway.

Our clinical experience in a service that routinely delivers CBT-E while allowing flexibility in treatment duration also suggests that a fixed number of sessions may not be optimal for every patient. Some patients progress more rapidly than anticipated and may achieve a full response in approximately 20-25 sessions, whereas others change more gradually and may require as many as 60 sessions to reach a comparable outcome. These differences appear to be influenced by several factors, including the patient’s awareness of the disorder and its consequences, readiness to engage actively in change, degree of low weight, severity and complexity of the eating-disorder psychopathology, and the extent of the difficulties encountered in restoring weight. Extending treatment beyond 40 sessions should not, however, become an open-ended response to a lack of progress. In our view, an extension is justified only when the patient remains actively engaged, and there is continuing evidence of meaningful improvement in weight, eating-disorder psychopathology, or both. The most clinically appropriate implementation strategy may therefore be neither to prescribe an identical fixed duration for all patients nor to regard the shortest possible treatment as inherently superior, but to individualise treatment length according to the patient’s presentation, rate of change and ongoing response.

The conceptual differences between CBT-AN-20 and CBT-E may eventually have implications for treatment selection. CBT-AN-20 may be particularly appropriate when food-, weight- and body-related fears and avoidance appear prominent and when the patient is prepared to confront them from the outset. CBT-E may be preferable when ambivalence is pronounced, when weight control is strongly egosyntonic, or when the clinical picture requires a more extensive, formulation-driven intervention addressing the over-evaluation of shape, weight and control over eating together with other maintaining mechanisms. At present, however, these possible indications are hypotheses rather than empirically established treatment-matching criteria

Research priorities

Future research on CBT-AN-20 should investigate clinical outcomes in cohort studies, with longer follow-up assessment and with patients with moderate or severe underweight. If the results are confirmed, the next step should be a sufficiently powered randomised comparison of CBT-AN-20 and CBT-E, potentially using a non-inferiority design. Equivalence should not be judged from short-term BMI change alone but from predefined outcomes combining weight recovery and eating-disorder psychopathology.

The trial should use identical eligibility criteria, stratify participants by baseline BMI and illness duration, and provide comparable therapist training, supervision, and independent assessment of competence and fidelity.

Outcome definitions should be standardised and include completion, full response, remission, relapse, adverse events, clinical impairment, quality of life, functioning, acceptability, medical status, and use of additional treatment. Assessment should include both the EDE and EDE-Q, with follow-up at 6, 12, and preferably 24 months.

Health-economic analyses should establish whether fewer sessions produce genuine savings across the full care pathway. Mechanism and qualitative studies should also examine how early weight change, anxiety, avoidance, engagement, and over-evaluation of shape and weight influence outcome, why some patients respond well to the immediate behavioural demands of CBT-AN-20 while others disengage, and whether these factors can guide treatment matching.

Conclusion

CBT-AN-20 is a promising brief intervention. Its preliminary findings indicate that clinically meaningful weight gain and substantial reductions in eating-disorder psychopathology can occur within 20 sessions in a subgroup of medically stable adults with anorexia nervosa and predominantly mild underweight. The short-term intention-to-treat increase in BMI is broadly comparable with that reported in some longer studies of CBT-E.

The available evidence does not, however, establish equivalence between CBT-AN-20 and CBT-E. CBT-AN-20 needs more robust effectiveness studies and efficacy studies before a comprehensive comparison can be made.

Conceptually, CBT-AN-20 appears closer to CBT-T than to a shortened version of CBT-E. It prioritises brevity, rapid behavioural change, exposure to feared situations and early evaluation of engagement. CBT-E, by contrast, is a transdiagnostic, formulation-driven treatment that places greater emphasis on collaborative engagement, the patient’s active decision to change, modification of the individual mechanisms maintaining the eating disorder, and the consolidation of progress over time.

The findings also raise a broader question about how treatment duration should be determined. Clinical efficiency should not be equated with delivering the fewest possible sessions. Some patients may respond fully to a relatively brief intervention, whereas others may require substantially more time because of limited awareness of the disorder, greater ambivalence about change, more severe or complex psychopathology, or difficulties with weight regain. Treatment duration may therefore be better guided by the patient’s clinical characteristics, rate of progress and continuing active engagement than by a predetermined maximum number of sessions. Extending treatment beyond the standard duration is most defensible when meaningful improvements in weight, eating-disorder psychopathology or both are still occurring.

The two approaches may eventually prove useful for different patients, or brief treatment may become an initial option within a more flexible stepped-care pathway. At present, however, CBT-AN-20 should be regarded as an intervention requiring adequately powered controlled evaluation, clearer identification of the patients for whom it is suitable, and longer-term follow-up. CBT-E remains supported by a broader evidence base across different eating-disorder presentations, levels of underweight and follow-up periods. Future research should therefore move beyond asking which treatment is shorter and determine which intervention, delivered for what duration, produces a full and sustained response for which patients.

References

American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders, (DSM-5). American Psychiatric Publishing.

Byrne, S., Wade, T., Hay, P., Touyz, S., Fairburn, C. G., Treasure, J., Schmidt, U., McIntosh, V., Allen, K., Fursland, A., & Crosby, R. D. (2017). A randomised controlled trial of three psychological treatments for anorexia nervosa. Psychological Medicine, 47(16), 1-11. https://doi.org/10.1017/s0033291717001349

Calugi, S., Sartirana, M., Frostad, S., & Dalle Grave, R. (2021). Enhanced cognitive behavior therapy for severe and extreme anorexia nervosa: An outpatient case series. Int J Eat Disord, 54(3), 305-312. https://doi.org/10.1002/eat.23428

Craske, M. G., Treanor, M., Conway, C. C., Zbozinek, T., & Vervliet, B. (2014). Maximizing exposure therapy: An inhibitory learning approach. Behaviour Research and Therapy, 58, 10-23. https://doi.org/10.1016/j.brat.2014.04.006

Dalle Grave, R., Sartirana, M., & Calugi, S. (2019). Weight phobia or overvaluation of shape and weight? A cognitive analysis of the core psychopathology of anorexia nervosa. IJEDO. https://doi.org/10.32044/ijedo.2019.08

Dalle Grave, R., Sartirana, M., Dalle Grave, A., & Calugi, S. (2023). Effectiveness of enhanced cognitive behaviour therapy for patients aged 14 to 25: a promising treatment for anorexia nervosa in transition‐age youth. European Eating Disorders Review. https://doi.org/10.1002/erv.3019

Dalle Grave, R., Sartirana, M., El Ghoch, M., & Calugi, S. (2018). DSM-5 severity specifiers for anorexia nervosa and treatment outcomes in adult females. Eating Behaviors, 31, 18-23. https://doi.org/10.1016/j.eatbeh.2018.07.006

Duggan, H. C., Hardy, G., & Waller, G. (2026). Cognitive-behavioural therapy (CBT) for outpatients with anorexia nervosa: a systematic review and meta-analysis of clinical effectiveness. Cognitive Behaviour Therapy, 55(2), 139-184. https://doi.org/10.1080/16506073.2025.2465745

Fairburn, C. G. (2008). Cognitive behavior therapy and eating disorders. Guilford Press.

Fairburn, C. G., Cooper, Z., Doll, H. A., O'Connor, M. E., Palmer, R. L., & Dalle Grave, R. (2013). Enhanced cognitive behaviour therapy for adults with anorexia nervosa: a UK-Italy study. Behav Res Ther, 51(1), R2-8. https://doi.org/10.1016/j.brat.2012.09.010

Waller, G., Eddy, K. T., Rose, C. L., Thomas, J. J., Turner, H. M., & Wade, T. D. (2025). Cognitive-behavioural therapy for anorexia nervosa: A clinician’s guide to CBT-AN-20. Taylor & Francis.

Waller, G., Turner, H., Tatham, M., Mountford, V., & Wade, T. (2019). Brief cognitive behavioural therapy for non-underweight patients: CBT-T for eating disorders. Routledge.