Key messages

• Cannabinoids might not increase weight or improve symptoms in anorexia nervosa. 
• Cannabinoids are probably associated to frequent adverse effects in patients with anorexia nervosa, so the risk / benefit balance is not favorable. 

What is the evidence.
See evidence matrix  in Epistemonikos later

We found four systematic reviews [1],[2],[3],[4] which included two primary studies that answered the question of interest, reported in three references [5],[6],[7]. Both studies correspond to randomized trials.

What types of patients were included*

One trial [5] included 11 women with a primary diagnosis of anorexia nervosa in a hospital setting. The average age was 23.6 years. The diagnosis of anorexia nervosa was made based on the Feighner criteria. All patients had amenorrhea and had lost at least 25% of their weight but were not below 15% of ideal weight **.

The other trial [6] included 25 women aged over 18 years, diagnosed with DSM-IV anorexia nervosa for at least 5 years. Patients were considered in ambulatory and hospitalized setting.

What types of interventions were included*

In one trial [5] THC (delta-9-tetrahydrocannabinol) was used orally in doses between 2.5 and 10 mg three times daily. The control group received an active placebo (oral diazepam between 3 and 15 mg three times daily).

In addition to the pharmacological treatment under study, patients received individual psychotherapy twice a week, group therapy once a week, and nasogastric tube feeding occasionally **.

In another trial [6] dronabinol 2.5 mg twice daily was compared orally versus placebo.In both groups, the regular baseline treatment was maintained.

What types of outcomes
were measured

The main outcome used by systematic reviews was weight gain measured in kilograms [5],[6].
In addition, the following psychiatric scales were used [5].

• HSCL-90: Hopkins Symptom Checklist-90
• GAAQ: Goldberg Anorectic Attitude Questionnaire
• SDS: Goldberg Situational Discomfort Scale
• PRS: Psychiatric Rating Scale

Other outcomes evaluated were the presence of adverse effects, such as sleep disturbance, paranoia and dysphoria.

Follow-up was performed in one trial for 4 weeks [5] and  for 8 weeks in the other trial [6].

To whom this evidence does and does not apply

• The evidence presented in this summary applies to women diagnosed with anorexia nervosa.
• The population studied includes inpatients and outpatients, so it is applicable to any setting.
• In one trial the diagnosis of anorexia nervosa was based on Feighner criteria, which are part of the DSM-III and in the other it was based on DSM-IV.

• The outcomes presented in the summary of findings table are those considered critical for health decision-making by the authors of this article.
• We searched for a core outcome set for anorexia nervosa in the COMET initiative database (Core Outcome Measures in Effectiveness Trials) but it is not yet available.

• It is an intervention that might have no benefit and that entails adverse effects, so the risk-benefit balance is not favorable.

• Commercial cannabinoid formulations are generally expensive.
• The cost-benefit balance is not favorable, since it is a costly intervention with no clear benefit and entails adverse effects.
• In addition, the use and marketing of these drugs is not authorized in many countries, so the cost associated with legalization, production and marketing is probably substantive.

• Faced with the evidence presented in this summary, most patients and clinicians should lean against the use of cannabinoids in anorexia nervosa.
• There is currently a positive perception of the therapeutic effects of cannabinoids, both in the public and health professionals, which puts additional difficulties for informed decision-making in this context.

Differences between this summary and other sources

• The conclusions of this summary coincide with the different systematic reviews identified, both in terms of the absence of effect and the presence of adverse effects.
• The conclusions of this summary agree with the main guidelines, such as The Health Effects of Cannabis and Cannabinoids: The Current State of Evidence and Recommendations for Research, which state there is insufficient evidence to support or refute the use of cannabinoids in anorexia nervosa [9].

• The likelihood of future evidence changing the conclusions of this summary regarding the benefits of cannabinoids in anorexia nervosa is high, due to the existing uncertainty.
• There are no ongoing trials addressing this question according to the International Clinical Trials Registry Platform of the World Health Organization.
• New systematic reviews could provide clearer conclusions, since the ones identified for this summary may not contain the totality of the evidence available for this topic, and have important limitations. We did not identify ongoing systematic reviews in PROSPERO database.