65 IMPLICATIONS OF GENETIC TESTING FOR SUSCEPTIBILITY TO CANNABIS-ASSOCIATED SCHIZOPHRENIA

Background: Marijuana use, which has been identified as a risk factor for the development of psychosis (e.g., schizophrenia), is increasingly being legalized. The relationship between psychosis and cannabis use may depend on an individual's genetic makeup, and in the near future genetic testing could be used to identify people's level of risk for cannabis-associated psychosis (CAP), raising serious ethical and policy considerations. To date, it is unclear whether personalized information about genetic susceptibility to CAP might promote healthy behaviors (i.e., abstention from cannabis use among those with elevated risk), disincentivize healthy behaviors (e.g., by promoting a sense of invulnerability among those without heightened risk, which has been called a "genetic invincibility effect"), or both. Also unknown is how the effects of such genetic test results might depend on an individual's history of cannabis use. Methods: U.S. adults (N=306, 51.3% male, 47.4% female, 1.3% unknown; mean age 32.74 yrs, SD=10.56) were recruited online via Amazon.com's Mechanical Turk platform. They were randomly assigned to 1 of 3 conditions: a predisposition condition (in which they imagined learning that they were genetically predisposed to CAP), a no-predisposition condition (in which they imagined learning that they lacked a predisposition), or a control condition in which they were provided with no genetic testing scenario. Participants rated, on a scale from 1 (Not at all) to 7 (Very), how important it would be for them to avoid marijuana, as well as the likelihood that they would. These two ratings formed a reliable scale (Cronbach alpha =.87), so they were averaged to compute a marijuana-use avoidance (MUA) score for each participant. At the end of the study, participants were asked whether they had ever used marijuana at least three times per week (41.2% yes). We analyzed the data separately for those who answered yes to this question (i.e., had a significant history of marijuana use) and those who answered no. Results: Among participants without a significant history of marijuana use, a one-way ANOVA revealed a significant main effect of condition on MUA scores, F(2, 177)=19.73, p<.001. Dunnet t-tests revealed that compared to those in the control condition (N=51, M=5.23, SD=1.75), those in the predisposition condition had significantly higher scores (N=72, M=6.24, SD=1.15, p=.001), but those in the no-predisposition condition had significantly lower scores (N=57, M=4.51, SD=1.84, p=.035). Among those with a significant history of marijuana use, there was also a significant main effect of condition on MUA scores, F(2, 123)=12.06, p<.001. Dunnet t-tests revealed that compared to those in the control condition (N=51, M=3.08, SD=2.07), those in the predisposition condition had significantly higher scores (N=32, M=5.14, SD=1.72, p<.001), but scores in the no-predisposition condition were not significantly different (N=43, M=3.20, SD=2.11, p=.944). Discussion: Genetic testing that reveals a predisposition to CAP may promote intentions to avoid marijuana use, but among individuals with no significant history of marijuana use, genetic testing that reveals the absence of such a predisposition might also risk disincentivizing avoidance of marijuana use. If confirmed by additional studies, the hazards of this kind of "genetic invincibility effect" should be considered in future decision-making about ethical and policy guidelines regarding whether to offer genetic testing for CAP susceptibility, and if so, to whom. Disclosure: Nothing to disclose.