In a double-blind, placebo-controlled trial of more than 170 people with post-traumatic stress disorder, patients who took Dronabinol — a synthetic form of THC, the psychoactive compound in cannabis — before bed saw their nightmares drop sharply, and more than a third stopped having PTSD nightmares altogether after ten weeks. The trial, led by researchers at Charité – Universitätsmedizin Berlin together with the Psychiatric University Clinic of the Charité at St. Hedwig Hospital, University Medical Center Hamburg-Eppendorf, and the Central Institute for Mental Health in Mannheim, was published August 5, 2026 in Nature Medicine (DOI: 10.1038/s41591-026-04546-9).
What the trial measured
Participants — all diagnosed with PTSD and experiencing recurrent trauma-related nightmares — were randomly assigned to take either Dronabinol or a placebo each evening, self-administered at home, over ten weeks. Nightmare severity was tracked on a zero-to-eight scale. By the end of the trial, the Dronabinol group’s average nightmare burden had dropped by 3.7 points, compared with a 2.2-point drop in the placebo group. More than a third of patients treated with Dronabinol reported they no longer experienced any nightmares at all after ten weeks, and another 21% reported their nightmare burden had been cut at least in half.
A double-blind, placebo-controlled design — where neither participants nor researchers know who received the active drug until after the data are analyzed — is what the field considers the gold standard for establishing that an effect is attributable to the drug itself rather than expectation or natural symptom fluctuation, which matters especially for a subjective, self-reported outcome like nightmare frequency.
The funding disclosure that belongs in the headline, not the fine print
The trial was funded by Bionorica SE, a German pharmaceutical manufacturer with a commercial stake in Dronabinol and other THC-based products. That is not a disqualifying fact — industry funding of drug trials is common and, when properly disclosed, is exactly the kind of information a conflict-of-interest disclosure framework exists to surface rather than hide. But it is a genuine, name-it-directly angle: a manufacturer funding the pivotal trial of its own compound is a textbook case for why journals require funding-source and competing-interest statements, and why research-integrity offices ask “who paid for this, and does the answer change how a reader should weigh the result” before a study’s findings get repeated as settled fact. Readers evaluating this result should weigh the outcome alongside who financed the research that produced it, not instead of it.
A trial design with a real ethics-review angle
Beyond the funding question, this trial’s design carries its own protocol-review considerations. PTSD patients are, by definition, a trauma-exposed and psychologically vulnerable population, which typically triggers a higher level of scrutiny from an institutional review board (IRB) or research ethics committee than a trial in a general population would. Two design choices stand out: dosing was self-administered at home each evening rather than supervised in a clinical setting, and the intervention is a controlled psychoactive substance. Both raise practical protocol questions an ethics board has to work through before approval — how adverse events or intoxication-related risk get monitored without direct supervision, what safety instructions and contact protocols patients are given, and how informed consent communicates the risks of an at-home psychoactive medication to a population that may already be dealing with sleep disruption, hypervigilance, or impaired judgment as PTSD symptoms. None of that means the trial was designed improperly — multi-site trials across four major German academic medical centers went through formal ethics review as a condition of running at all — but it is a legitimate example of the kind of protocol-design trade-off that clinical-research ethics review exists to work through, not a footnote.
What this does and doesn’t establish
This is one randomized trial, not a regulatory approval or a standard-of-care recommendation. Dronabinol is an existing, already-approved medication (prescribed for other indications such as chemotherapy-induced nausea and appetite stimulation), which is what made a trial of this kind feasible at this scale, but repurposing an approved drug for a new indication — PTSD nightmares — still requires its own trial evidence and, typically, its own regulatory pathway before it becomes a standard prescribing option for the condition. The multi-site design (four independent academic medical centers) and sample size (170+) are real strengths over many single-site pilot studies, but a single trial result — however well-conducted — is not the same as multiple independent replications, which is generally what is needed before a treatment becomes standard practice.
Frequently asked questions
Who funded this trial, and does that affect the result?
The trial was funded by Bionorica SE, a pharmaceutical manufacturer with a commercial interest in Dronabinol and other THC-based products. Industry funding does not automatically invalidate a result, but it is a material fact readers and clinicians should weigh: funding-source disclosure exists precisely so a manufacturer-funded trial of a manufacturer’s own product can be evaluated with that context in view, rather than presented as if it were independently financed.
Is Dronabinol now an approved treatment for PTSD nightmares?
No. This trial demonstrates a statistically significant effect in a randomized, placebo-controlled setting, but a single trial result is not a regulatory approval or a clinical practice-guideline change. Dronabinol is already an approved drug for other indications; using it for PTSD nightmares specifically would require further evidence and, typically, its own regulatory pathway.
Why does at-home, self-administered dosing matter for this trial’s ethics review?
Dosing a psychoactive controlled substance without direct clinical supervision, in a population that has been through trauma, raises real protocol questions about adverse-event monitoring, safety instructions, and informed consent that a research ethics committee has to resolve before approving a trial — a genuinely different risk profile than an in-clinic-supervised dosing design.
Sources
Primary source: Charité – Universitätsmedizin Berlin et al., published in Nature Medicine, August 5, 2026. DOI: 10.1038/s41591-026-04546-9. Secondary coverage: ScienceDaily, August 5, 2026. This article was last checked against the cited sources on August 7, 2026.
