Statistical Adjustment Did Not Isolate CBD’s Effect in a Naturalistic Memory Study

A naturalistic study found different recognition-memory changes in THC-dominant and THC-plus-CBD product groups, including after statistical adjustment for a THC-related exposure measure. That adjustment did not independently isolate CBD as the cause of the difference.

The product groups differed in more than CBD

Thirty-two regular cannabis users underwent a brief word-recognition task before and after home use of assigned cannabis products; 31 remained in the main analysis after an outlier exclusion. Products included flower and high-potency concentrates, with differing THC and CBD content.

Assignment was not completely random across the feasibility sample. Preferred product form and administration methods varied, and the THC-dominant group reported more cannabis use during the preceding month. The authors identified these factors as limitations.

What the adjusted analysis found

Recognition discrimination declined after use in the THC-dominant group but did not significantly decline in the THC-plus-CBD group. A group-by-session interaction remained significant when a log-transformed THC-related plasma measure was included as a covariate.

That measure combined parent THC, 11-hydroxy-THC and THC-COOH. The composite was an exposure index used in the analysis, not a chemically uniform amount of active THC. Its components should not be treated as equally psychoactive.

A covariate cannot recreate matched experimental exposure

Statistical adjustment addresses the included measurement under the model’s assumptions. It does not make the groups identical in product composition, prior use, route, timing or other unmeasured characteristics.

The separate regression found an association between the THC-related composite and poorer recognition accuracy, but neither the CBD term nor the THC-by-CBD interaction was significant. Those null findings should accompany the positive product-group comparison.

How far the interpretation can go

The results support further controlled investigation of how product composition relates to memory. They do not prove that adding a specified amount of CBD protects a particular person’s memory or neutralizes THC.

Blood was separated into plasma, and the brief word task occurred after the post-use blood draw. Neither the plasma composite nor performance on this recognition task supplies a direct measure of driving ability. A case-specific claim should preserve those differences rather than convert an observational comparison into a treatment guarantee.

Related: Interpreting THC concentration and impairment.

Primary research

Acute effects of naturalistic THC vs. CBD use on recognition memory: a preliminary study. DOI: 10.1186/s42238-020-00034-0.