Umeå universitets logga

Background: Anorexia nervosa (AN) has high rates of enduring disease and mortality. Currently, there is insufficient knowledge on the predictors of relapse after weight normalization and this is why a systematic literature review was performed. Methods: PubMed, EMBASE, PsychInfo, and Cochrane databases were searched for literature published until 13 July 2021. All study designs were eligible for inclusion if they focused on predictors of relapse after weight normalization in AN. Individual study definitions of relapse were used, and in general, this was either a drop in BMI and/or reccurrence of AN symptoms. Results: The database search identified 11,507 publications, leaving 9511 publications after the removal of duplicates and after a review of abstracts and titles; 191 were selected for full-text review. Nineteen publications met the criteria and included 1398 AN patients and 39 healthy controls (HC) from adults and adolescents (ages range 11–73 years). The majority used a prospective observational study design (12 studies), a few used a retrospective observational design (6 studies), and only one was a non-randomized control trial (NRCT). Sample sizes ranged from 16 to 191 participants. BMI or measures of body fat and leptin levels at discharge were the strongest predictors of relapse with an approximate relapse rate of 50% at 12 months. Other predictors included signs of eating disorder psychopathology at discharge. Conclusions: BMI at the end of treatment is a predictor of relapse in AN, which is why treatment should target a BMI well above 20. Together with the time to relapse, these outcomes are important to include in the evaluation of current and novel treatments in AN and for benchmarking.