A drug’s effect need not increase steadily with dose. Demonstrating a curved dose-response relationship, however, requires more than finding a statistically significant result at one dose and nonsignificant results at two others. A small cannabidiol study illustrates both the pharmacodynamic question and the limits of the evidence.
Three doses, one laboratory stress task
Linares and colleagues studied 57 healthy men in a randomized, double-blind experiment published in the Brazilian Journal of Psychiatry in 2019. Participants received a single oral dose of cannabidiol (CBD) in corn-oil capsules or matching placebo. The 150 mg and 300 mg groups each contained 15 participants, the 600 mg group contained 12, and the placebo group contained 15.
The participants prepared and delivered a speech in front of a camera. Investigators measured subjective mood at several stages and recorded blood pressure. The pharmacodynamic question concerned changes in anxiety during an induced stress task, rather than drug absorption, clearance or concentration-time modeling.
What the anxiety result showed
The 300 mg group reported lower anxiety than placebo during the speech phase in a reported post hoc comparison, with p = 0.042. The 150 mg and 600 mg groups did not show significant differences from placebo on the reported mood-scale comparisons. The authors described the pattern as consistent with an inverted U-shaped dose-response curve.
That description needs qualification. The repeated-measures analysis reported no statistically significant group-by-phase interaction, with p = 0.1. The methods described phase-specific comparisons as following a significant interaction, yet the results reported those comparisons despite the nonsignificant interaction. This discrepancy limits confidence in a strong dose-response conclusion.
A significant placebo comparison at 300 mg and nonsignificant comparisons at other doses do not themselves establish that 300 mg was significantly better than both other doses. Nor do three active doses identify a universal optimum.
Different outcomes remained distinct
The study did not find significant group or group-by-phase effects for systolic or diastolic blood pressure. The mood-scale factors labeled sedation, cognitive impairment and distress also showed no significant effects. These were subjective scale outcomes, not a comprehensive objective assessment of cognition or driving performance.
Consequently, the anxiety signal cannot be generalized into a demonstrated improvement across all physiological or functional measures. A nonsignificant sedation score also does not certify that an individual can drive safely.
Limits on applying the findings
Only healthy men participated, and treatment involved one dose during one experimental session. The results do not directly establish long-term treatment efficacy in patients with anxiety disorders, effects in women, or equivalent effects from other preparations. The authors acknowledged the small sample and the need for larger studies.
The paper discussed receptor-based explanations for the proposed curve, drawing on other research. It did not directly establish a receptor mechanism in these participants. Several authors disclosed CBD-related patent or industry relationships, and the study drug was supplied without charge.
The defensible conclusion is narrow: one dose produced a reported anxiety signal under particular experimental conditions, with statistical and design limitations that prevent converting it into an individualized dosing rule.
Source: Linares IM and colleagues. Cannabidiol presents an inverted U-shaped dose-response curve in a simulated public speaking test. Brazilian Journal of Psychiatry. 2019;41(1):9-14. PubMed record.
Leave a Reply