A Later Null Comparison Was Not the Same as a Significant Recovery Interaction
The statistical evidence for recovery differed across outcomes in an alcohol–cannabinoid mouse study. A later nonsignificant group comparison did not establish the same treatment-by-time interaction for every response.
The study assessed two recovery intervals
Pava and colleagues compared cannabinoid responses one and ten days after repeated ethanol. Earlier differences were no longer significant in the later comparisons, supporting the authors’ account of reversibility within the observation window.
The interaction tests supplied an additional distinction
The treatment-by-day interaction was significant for locomotor activity and body temperature. It was not significant for antinociception, where the reported interaction p value was 0.1141, despite a significant overall treatment effect and a null later pairwise comparison. Those are different statistical questions.
Recovery claims should retain both analyses
A null comparison at the later visit does not establish exact equivalence or determine when the groups converged between visits. A nonsignificant interaction likewise does not prove that no recovery occurred. The defensible account describes the observed endpoint comparisons and the strength of evidence for differential change over time separately. It does not convert a ten-day animal observation into a guaranteed cannabis-tolerance reset schedule for people or a claim that every measured function recovered identically.
Related analysis: The tested agonist and direction of alcohol–cannabinoid cross-tolerance.
Research cited
Pava MJ, Blake EM, Green ST, Mizroch BJ, Mulholland PJ, Woodward JJ. Tolerance to cannabinoid-induced behaviors in mice treated chronically with ethanol. Psychopharmacology (2012).
