Why reaching the lower end of the normal BMI range may not be enough to prevent relapse.
Intensive treatment for anorexia nervosa often results in substantial weight gain and improvements in eating disorder psychopathology. However, some patients lose weight after discharge, and a significant proportion experience a relapse during the following year.
Identifying factors associated with relapse may help clinicians improve treatment outcomes. Among the variables investigated, body mass index (BMI) at discharge has emerged as one of the most consistent predictors of subsequent weight maintenance.
What does the research show?
A systematic scoping review published by Frostad and colleagues in the Journal of Personalized Medicine in 2022 examined 19 studies involving 1,398 adolescents and adults with anorexia nervosa. The aim was to identify factors associated with relapse following weight restoration.
Relapse rates varied considerably between studies, partly because of differences in how relapse was defined and the length of follow-up. Nevertheless, approximately half of the studies reporting 12-month outcomes found relapse rates close to 50%.
The main finding was that a lower BMI at the end of treatment, even within the normal range, was associated with an increased risk of relapse.
Based on these findings, the authors suggested that treatment in adults should aim for a BMI well above 20 kg/m². However, the available evidence does not establish a specific BMI threshold that reliably prevents relapse in all patients.
Other potential predictors included lower body fat percentage, lower leptin concentrations, and residual eating disorder psychopathology. The evidence supporting these variables was less consistent than that for discharge BMI.
Does a small difference in discharge BMI matter?
One of the studies included in the review was conducted at Villa Garda Hospital and published in the European Journal of Clinical Nutrition in 2016.
The study involved 54 adult women with anorexia nervosa who had received intensive enhanced cognitive behaviour therapy (CBT-E). All participants had achieved and maintained a BMI of at least 18.5 kg/m² for a minimum of four weeks before discharge.
At the one-year follow-up, 26 patients had maintained a BMI of at least 18.5 kg/m², whereas 28 had returned to an underweight BMI.
The main difference between the two groups was their BMI at discharge. Patients who maintained a BMI in the normal range had a mean discharge BMI of 20.2 kg/m², compared with 19.4 kg/m² among those who subsequently became underweight.
The study also assessed total body fat percentage and fat distribution using dual-energy X-ray absorptiometry (DXA). Neither measure was significantly associated with weight maintenance at follow-up.
These results suggest that reaching the lower boundary of the normal BMI range may not be sufficient to optimise the likelihood of maintaining weight restoration after intensive treatment.
Although the difference in mean discharge BMI between the two groups was relatively small, it was associated with significantly different weight outcomes one year later.
Does illness duration influence the importance of discharge BMI?
Another study included in the review was conducted by Redgrave and colleagues at Johns Hopkins University and published in Frontiers in Psychiatry in 2021.
The researchers examined 191 adult women hospitalised for anorexia nervosa or another underweight eating disorder. Patients were divided into two groups according to illness duration: less than seven years and seven years or longer.
Patients with a longer illness duration generally had a more complicated clinical history, including more previous hospital admissions. Nevertheless, both groups achieved similar rates of weight gain during treatment.
Six months after discharge, BMI at discharge was a significant predictor of maintaining a BMI of at least 19 kg/m², whereas illness duration was not.
In particular, patients who had achieved a BMI of at least 19 kg/m² at discharge had approximately five times the odds of maintaining a BMI of at least 19 kg/m² at follow-up compared with those discharged below this threshold.
These findings have an important clinical implication: a long duration of anorexia nervosa should not, in itself, be considered a reason to abandon the goal of full weight restoration.
Patients with a longstanding illness may still benefit from intensive treatment aimed at achieving a healthy weight.
What are the clinical implications?
Taken together, these findings support the importance of helping patients achieve full weight restoration before discharge from intensive treatment.
However, the BMI thresholds used in research should not be applied rigidly to individual patients. A healthy weight needs to be established by considering age, weight history, physical health and, in adolescents, expected growth and development.
It is also important to distinguish weight restoration from recovery from anorexia nervosa. Although achieving a healthy weight is essential, patients may continue to experience eating disorder psychopathology, including the overvaluation of shape and weight, dietary restraint, and other maintaining mechanisms.
In this respect, a prospective study conducted at Villa Garda Hospital and published in the International Journal of Eating Disorders in 2025 examined outcomes in 421 patients with anorexia nervosa receiving intensive CBT-E.
The study found that a higher baseline BMI and greater BMI increases during treatment were associated with more favourable weight outcomes at follow-up. Greater reductions in eating disorder psychopathology during treatment were also associated with better subsequent psychological outcomes.
These findings indicate that weight restoration and the reduction of eating disorder psychopathology should both be important treatment targets.
Another consideration is how patients are prepared for discharge. They need to develop the skills required to maintain regular eating, manage situations that may trigger dietary restriction, and recognise early signs of re-emerging eating disorder behaviours and attitudes.
Continuing treatment after discharge may also help patients consolidate the changes achieved during intensive treatment and address difficulties encountered when returning to everyday life.
Conclusions
The available evidence indicates that BMI at discharge is one of the most consistent predictors of weight maintenance after intensive treatment for anorexia nervosa. A relatively low BMI, even within the normal range, is associated with an increased risk of subsequent weight loss or relapse.
Although most studies are observational and use different definitions of relapse, their findings support pursuing full, individually defined weight restoration rather than considering the lower boundary of the normal BMI range a sufficient treatment target.
Weight restoration, however, is not sufficient on its own. It needs to be accompanied by a reduction in eating disorder psychopathology and the development of strategies to maintain the changes achieved during treatment.
The aim of intensive treatment should be not only to help patients restore a healthy weight, but also to enable them to maintain it after discharge.
References
- Frostad S, et al. (2022). BMI at Discharge from Treatment Predicts Relapse in Anorexia Nervosa: A Systematic Scoping Review. Journal of Personalized Medicine, 12, 836. https://doi.org/10.3390/jpm12050836
- El Ghoch M, Calugi S, Chignola E, Bazzani PV, Dalle Grave R. (2016). Body mass index, body fat and risk factor of relapse in anorexia nervosa. European Journal of Clinical Nutrition, 70, 194–198. https://doi.org/10.1038/ejcn.2015.164
- Redgrave GW, et al. (2021). Discharge Body Mass Index, Not Illness Chronicity, Predicts 6-Month Weight Outcome in Patients Hospitalized With Anorexia Nervosa. Frontiers in Psychiatry, 12, 641861. https://doi.org/10.3389/fpsyt.2021.641861
- Calugi S, Cattaneo G, Chimini M, Dalle Grave A, Conti M, Dalle Grave R. (2025). Predictors of Intensive Enhanced Cognitive Behavioral Therapy for Anorexia Nervosa. Prospective Cohort Study. International Journal of Eating Disorders. https://doi.org/10.1002/eat.24455