Self-reported polysubstance use is negatively associated with prospective memory in university students

Abstract The present study investigated self-reported patterns of substance use and prospective memory ability. Participants completed a self-report survey of their drug use history (ten substances and use-related behaviors), a self-report measure of prospective and retrospective memory failures, and a laboratory-based prospective memory task under one of two encoding conditions (standard or implementation intention). To better represent naturalistic patterns of substance use, a latent class analysis was applied to binary indicators of substance use. Class membership was then used as a categorical predictor of prospective memory. The latent class analysis revealed three classes: non-use, alcohol-cannabis use, and polysubstance use. Latent class was associated with both self-reported and laboratory-based prospective memory, and the polysubstance use class reported the most everyday memory failures. Implementation intention encoding improved task performance but did not interact with latent class. In sum, latent class analysis provided an ecological representation of how patterns of substance use can be classified and associated with prospective memory deficits.