CBD for Cannabis Use Disorder: What an Adaptive Dose-Finding Trial Established
A randomized trial found signals that pharmaceutical CBD could reduce cannabis use during a supported quit attempt. It did not show that CBD reliably produces complete abstinence, that higher doses always work better or that ordinary retail products have the same effect.
A treatment study with two kinds of outcomes
The phase IIa trial randomized 82 participants across two stages to placebo or oral CBD at 200, 400 or 800 mg per day for four weeks. Participants had cannabis use disorder, wanted to quit and had previously made an unsuccessful attempt. All received motivational interviewing.
The primary outcomes combined self-reported days abstinent with urinary THC-COOH adjusted for creatinine. This urine metabolite measure helps assess change in use under the protocol; it is not a direct measurement of current intoxication or a count of exactly how much cannabis someone consumed.
The design allowed an ineffective dose to be dropped
The investigators used a prespecified adaptive Bayesian approach, evaluating doses against probability criteria rather than relying exclusively on conventional significance tests. The 200-mg arm was discontinued after the interim analysis. The 400- and 800-mg arms continued and met the trial’s final primary-endpoint criteria.
At 400 mg, estimated additional abstinence was 0.48 days per week compared with placebo, with a 95% interval of 0.15–0.82. At 800 mg, the estimate was 0.27 days, with an interval of −0.09–0.64. The latter interval includes no difference even though the dose met the study’s Bayesian decision criterion. Reporting only “both doses worked” conceals that uncertainty.
More CBD was not uniformly better
Both retained doses reduced the urinary metabolite outcome during treatment. Follow-up findings were more favorable for 400 mg than for 800 mg, and secondary outcomes differed by dose. For example, the 400-mg arm had poorer reported sleep quality, while some withdrawal and anxiety measures favored 800 mg.
The study reported no severe adverse events and no clear excess of mild or moderate events over placebo. A short trial with this sample cannot exclude uncommon harms or establish long-term safety.
What can be applied outside the trial?
The findings concern a specified high-purity preparation, substantial daily doses, a motivated clinical population and concurrent psychological support. They do not isolate the effect of a low-dose supplement bought without comparable quality control.
This was an early dose-finding study, and the planned subsequent efficacy stage did not proceed because of funding. Its strongest contribution is to identify doses and outcomes for further controlled investigation. It is not a validated self-treatment schedule or proof that a negative urine result should follow CBD use.
Related analysis: Adult Epilepsy CBD Outcomes: Treatment Retention and Confounding Matter.
Primary research
Tom P Freeman and colleagues. Cannabidiol for the treatment of cannabis use disorder: Phase IIa double-blind placebo-controlled randomised adaptive Bayesian dose-finding trial. The lancet. Psychiatry (2020). DOI: 10.1016/S2215-0366(20)30290-X.
