Driving Under the Influence of Cannabis: Associations with Latent Profiles of Substance Use and Executive Cognitive Functioning

Abstract

Background: Driving under the influence of cannabis (DUIC) almost doubles car crash risk (odds ratios range: 1.28-2.49). Known DUIC correlates include male gender, low perceived danger of DUIC, and greater frequency of cannabis and other drug use. Less is known about the role of executive cognitive functioning (e.g., skills in planning, organization) as a correlate of DUIC. Deficits in executive cognitive functioning could precede, and be exacerbated by heavy cannabis use, potentially contributing to DUIC risk. Objectives: This cross-sectional survey study used a person-centered analysis (latent profile analysis) to (1) identify prototypical profiles representing aspects of executive functioning and substance use in young adults, and (2) determine which profiles were associated with self-report of DUIC. We hypothesized that at least two profiles would be identified: mainly or only cannabis use vs polysubstance use. We also predicted that the polysubstance use profile would be associated with worse executive functioning and self-report of DUIC. Method: Young adults (N=69; ages 18-25; mean age=20.0 [SD=1.9]; 62.3% female; 75.4% White, 13.0% Black, 11.6% Other race/ethnicity) who reported weekly cannabis use were recruited from the community in Pittsburgh, PA to participate in a study of cannabis effects on cognition. Baseline collected demographics, self-reported age of cannabis use onset (age <16 vs age >16), NIDA modified ASSIST, Marijuana Withdrawal Checklist, Alcohol Use Disorders Identification Test (AUDIT), Behavior Rating Inventory of Executive Functioning (BRIEF) (working memory, organization/planning scales), and Marijuana Consequences Questionnaire (item on “driven a car when high” in past 6 months). Latent profile analysis (LatentGold 5.1) was used to identify distinct classes, testing the fit of 1-5 classes. Each model included 10 indicators: age of cannabis use onset, frequency of cannabis and tobacco use, cannabis withdrawal severity, ASSIST scores for cannabis, cocaine and hallucinogens (the substances most often reported), AUDIT score, and BRIEF working memory, and organization/planning scores. For the best fitting model, covariates (i.e., self-report of DUIC, age, gender) were examined as profile correlates in a separate, final step. Results: A model with 3 latent profiles was selected (see Figure). The profiles represented “Polysubstance Use” (40.8%), “Primary Cannabis” (22.3%), and “Later Onset Cannabis” (36.9%). Polysubstance use profile reported more cannabis-related problems and other drug use, and more problems with executive functioning than the other profiles (p<.05). Later Onset (vs Polysubstance Use) profile had older onset age (p<.05), and had the lowest level of cannabis involvement. Primary Cannabis and Later Onset profiles did not differ in report of problems with executive functioning. DUIC in the past 6 months (reported by 50.7% of the total sample) was more likely to be reported by Polysubstance use than Later Onset profile (p<.01). Polysubstance use profile was younger than Primary Cannabis profile (p<.05). The profiles did not differ by gender. Conclusions: As hypothesized, Polysubstance Use profile (which reported early cannabis use onset; and worse executive functioning, including problems with memory, planning/ organization) was associated with self-report of DUIC. Results highlight the role of self-reported executive functioning difficulties in DUIC risk, and the importance of targeting polysubstance use in preventing DUIC.