BACKGROUND: Cannabis warning labels can communicate risks, but there is little research on warning perceptions and differences by product type.
METHODS: In a 2019 online survey, 1,000 U.S. adults (500 cannabis users and 500 cannabis non-users who used tobacco) were randomly assigned to view no warning or one of four U.S. or Canadian warnings displayed on images of packaging for dried flower or edible cannabis. The warnings described cannabis risks related to psychosis, addiction, lack of FDA oversight, and impaired driving. We used linear regression to examine perceptions of warnings and product harm as a function of product type (dried or edible) and warning. We examined which warning participants selected as most effective for discouraging youth use and impaired driving.
RESULTS: Participants found the addiction warning (cannabis users: B = -1.04, p < 0.001; cannabis non-users: B = 1.17, p < 0.001) and psychosis warning (users: B = -0.65, p < 0.05; non-users: B = -0.71, p < 0.05) less believable than the driving warning but indicated that they learned more from the psychosis warning than the driving warning (users: B = 0.88, p < 0.01; non-users (B = 1.60, p < 0.001). Participants viewing any warning considered smoking cannabis to be more harmful than those viewing no warning (all p < 0.05). The psychosis warning was most frequently selected as the best warning for discouraging youth use.
CONCLUSIONS: Warnings have the potential to educate consumers and impact cannabis harm perceptions. Warnings have similar effects across product types, potentially eliminating the need for product type-specific warnings. The association of cannabis use with risk for psychosis, a topic addressed in Canadian warnings, could be a useful topic of focus in U.S. warnings.