Higher THC Challenges Were Selected by Earlier Symptom Response

Higher THC doses were offered to participants without a sufficient symptom response at the standard dose. The primary analysis used the highest completed dose pair, while a separate sensitivity analysis restricted the comparison to 20 mg sessions.

The escalation was not random dose assignment

Thirty participants completed both 20 mg challenge sessions. Eleven qualified for higher-dose experiments based on their earlier response; six volunteered for 40 mg, and one proceeded to 60 mg. The eligibility rule used a defined increase in positive psychotic symptoms. These selected higher-dose participants were not a random sample assigned across THC doses.

Primary and sensitivity analyses answered different versions of the question

The per-protocol primary analysis retained each participant’s highest completed THC-dose pair. The 20 mg-only sensitivity analysis provided a common-dose comparison. Pooling the primary results does not establish an independent THC dose-response curve because dose escalation depended on prior response and participation.

Retain the actual conclusion

CBD pretreatment did not protect against the cannabis challenge and worsened the reported delayed-recall comparison. The adaptive procedure qualifies what that estimate represents; it does not erase the finding or demonstrate an effect at every possible dose. The participants had stable schizophrenia-spectrum illness and ongoing treatment, limiting transfer to other populations.

Related analysis: the schizophrenia cannabis-challenge findings.

Research cited

Chesney and colleagues (2025). Does cannabidiol reduce the adverse effects of cannabis in schizophrenia? A randomised, double-blind, cross-over trial. DOI: 10.1038/s41386-025-02175-3.