Treatment of eating disorders: an umbrella review
Systematic review
|Published
The objective was to conduct an umbrella review on the effect of treatments for eating disorders. The Norwegian Directorate of Health will provide an evidence synthesis on the prevention and treatment of eating disorders to the Ministry of Health and Care Services.
Download
Key message
There has been an increasing incidence of eating disorders in recent years. Many develop eating disorders already during childhood and adolescence. The most common eating disorders are anorexia, bulimia, binge eating disorder, and other specified/unspecified feeding or eating disorders (OSFED/USFED).
We conducted an umbrella review (a review of systematic reviews) on the effect of treatments for eating disorders. We searched seven databases and included systematic reviews with a literature search conducted in 2020 or later that examine the effect of treatments for eating disorders. Two researchers independently screened titles and abstracts, and then read relevant systematic reviews in full text.
We included 165 systematic reviews on treatment for eating disorders.
We assessed the methodological quality of the included reviews using AMSTAR-2 and based our conclusions on the most recent reviews with high or moderate methodological quality. We evaluated the certainty of the evidence using the GRADE approach (Grading of Recommendations, Assessment, Development and Evaluation).
Only four treatment outcomes were assessed as having moderate certainty in the results, all related to treatment for binge eating disorder. Two of the outcomes concerned cognitive behavioral therapy, and two concerned lisdexamfetamine.
Our main conclusion is that there is a lack of strong evidence due to a shortage of large, well-conducted controlled trials. For each intervention considered for treatment, more high-quality studies (on the same treatment and with the same comparison) are needed to draw conclusions with good confidence in the results.
Summary
Introduction
There has been an increasing prevalence of eating disorders in recent years. Many people develop eating disorders already during childhood and adolescence. The most common eating disorders are Other Specified/Unspecified Feeding or Eating Disorders (OSFED/USFED), binge eating disorder, bulimia, and anorexia.
Objective
To conduct an umbrella review on the effect of treatments for eating disorders. The Norwegian Directorate of Health will provide an evidence synthesis on the prevention and treatment of eating disorders to the Ministry of Health and Care Services. This deliverable forms part of the assignment ‘Prevention and Treatment of Eating Disorders’ (TB2023-36), as outlined in Proposition 1 S (2022–2023). The assignment is further anchored in the Escalation Plan for Mental Health (2023–2033).
Method
We searched seven databases and included systematic reviews with a literature search conducted in 2020 or later. We included reviews addressing the effect of treatment for eating disorders. Two researchers independently screened titles and abstracts, and then relevant systematic reviews in full text.
We identified so many systematic reviews that it was not possible for us to assess, grade, and summarize all within the available time. The Directorate of Health prioritized which reviews we proceeded with. We assessed the methodological quality of the prioritized reviews using AMSTAR-2 and, for each treatment form, based our work on the most recent reviews of high or moderate methodological quality. One researcher extracted results, and another checked the extraction.
We assessed our certainty in the results using the GRADE approach (Grading of Recommendations, Assessment, Development and Evaluation) for all except two systematic reviews, where grading had already been performed.
Results
We included 165 systematic reviews in this umbrella review. The 22 prioritized treatment forms covered 53 treatment comparisons, most of which were compared with placebo. The majority of comparisons (39 of 53) were examined in only one study. Among these primary studies, 25 had fewer than 50 participants, and only three had more than 100 participants (105, 108, and 154 participants). When something is studied in such small samples, the possibility that the results are due to chance is high, and findings should therefore be interpreted with caution.
Our main conclusion is that there is a lack of solid evidence due to a lack of large, well-conducted randomized controlled trials. For each treatment considered, more high-quality studies (on the same treatment and comparator) are needed before conclusions can be made with high confidence.
Only two treatment forms had direct comparisons based on multiple well-conducted studies (low risk of bias), with sufficiently large sample sizes, consistent results (low or no heterogeneity), and narrow confidence intervals — providing moderate confidence in the results.
We have moderate certainty in four outcomes for adults with binge eating disorder:
- It is likely that cognitive behavioral therapy (CBT) improves eating disorder symptoms measured as depression, and probably results in little or no change in BMI compared to a waiting list or “standard treatment” in adults with binge eating disorder.
- It is likely that lisdexamfetamine for 11–12 weeks reduces the number of binge eating days and compulsive symptoms compared with placebo in adults with binge eating disorder. However, amphetamines and related substances are known to have problematic long-term effects, so these results should be interpreted with caution given the short follow-up time.
We have low certainty in several outcomes for people with binge eating disorder:
- CBT may reduce the number of binge episodes, body image concerns, weight concerns, and eating concerns compared with a waiting list or standard treatment.
- CBT may have little or no effect on restrictive eating behaviors, quality of life, or self-esteem compared with waiting list, usual care, or other treatments.
- Lisdexamfetamine for 11–12 weeks may reduce binge frequency, eating disorder symptoms, increase weight loss, and reduce depressive symptoms compared with placebo. Again, these results should be interpreted with caution.
- Topiramate for 14–16 weeks may reduce binge frequency, increase remission rates, promote weight loss, and improve eating disorder symptoms compared with placebo.
We have low certainty in several outcomes for people with mixed eating disorders (where people with different eating disorder diagnoses were included in the same treatment groups):
- Exercise and mind-body practices may reduce harmful exercise habits and the frequency and severity of disordered eating behaviors compared with waiting list, usual care, or other treatments.
We have low confidence in several outcomes for adults with bulimia:
- CBT may reduce the number of binge episodes, compensatory behaviors, and vomiting compared with waiting list or standard treatment in adults.
- Fluoxetine may reduce binge frequency and have little or no effect on weight compared with placebo in adult women with bulimia.
We have low certainty in several outcomes for people with anorexia:
- CBT combined with family therapy may reduce eating disorder symptoms and increase weight in children and adolescents with anorexia compared with waiting list or standard treatment.
- Fluoxetine for 7–52 weeks may reduce the severity of depression and obsessive-compulsive symptoms compared with placebo in adult women with anorexia.
- Olanzapine may increase weight and have little or no effect on obsessive-compulsive symptoms in adults with anorexia.
Thirteen of the preselected treatments for eating disorders were not summarized or assessed in any recent systematic review, meaning that we do not know whether they are effective treatments. It is also unclear whether any studies have evaluated them. There is a clear need for systematic reviews on these treatments.
For six of the 22 prioritized treatments, the available systematic reviews were rated as low or critically low in methodological quality using AMSTAR-2. Thus, there is a need for systematic reviews of higher (high or moderate) methodological quality for these treatments.
Discussion
Our umbrella review examines the effect of treatments for eating disorders. A total of 165 systematic reviews were included, and 22 treatment types were assessed in detail. We based our synthesis on the most recent high- or moderate-quality reviews. It should be emphasized that even the best systematic review cannot provide stronger conclusions than the quality of the included primary studies allows. A high-quality systematic review can include low-quality studies (with high risk of bias), and conversely, a low-quality review can include high-quality primary studies.
Conclusion
Our main conclusion is that there is a general lack of strong evidence due to a shortage of large, well-conducted randomized controlled trials. Cognitive behavioral therapy was the most frequently studied treatment.