Understanding cannabis use and car crashes: Insights from a randomized trial using a driving simulator on THC blood levels and subjective measures of sleepiness and performance (2025)

Introduction: To investigate whether THC inhalation increases driving collisions, and to explore the relationship between car crashes and self-assessed driving ability as measured by Visual Analog Scales (VAS) and the Karolinska Sleepiness Scale, 30 male volunteers aged 18-30 years old, with a BMI of 18.5-26.2 Kg/m², moderate tobacco and caffeine consumption, and at least one year of cannabis use either chronically (1-2 joints per day) or occasionally (1-2 joints per week) participated in a randomized, double-blind, crossover study comparing three doses of THC (placebo, 10 mg, and 30 mg).
Method: Participants received one cigarette with a different THC dose in three randomized sequences, separated by a seven-day washout period. Following each cigarette, participants were hospitalized for 24 h, during which blood samples were taken and driving performance was assessed using the York driving simulator. Additionally, self-assessment scales were completed to measure subjective feelings of awareness, calmness, confusion, adroitness, mental sharpness, and driving confidence.
Results: The number of car crashes on driving simulator increased from one hour, showing a significant increased and delayed effect observed 4 h post-inhalation of THC at 10 mg and 30 mg (Beta = 0.65, 95% CI = [0.45, 0.86], p < 0.001 and IRR = 1.9, p < 0.001), indicating this period as critical for assessing cannabis impairment and its implications for traffic safety. Although some crashes occur earlier, this delayed four hour-peak effect may suggests that the highest car crash risk on the simulator does not coincide with the period of maximum THC levels, which typically occur within 5 to 15 min of cigarette completion. Furthermore, blood THC-Cmax, was strongly associated with increased risk of car collisions at 4 h, (Beta = 3.63, 95% CI = [2.56, 4.70], p < 0.001 and IRR = 37.7, p < 0.001). The increase in sleepiness following cannabis consumption, measured by the Karolinska Sleepiness Scale, correlated with higher car crashes on simulator rates (Beta = 0.10, 95% CI = [0.05, 0.15], p < 0.001 and IRR = 1.1, p < 0.001).
Conclusions: This study also found that driving collisions were positively associated with VAS adroitness (Beta = 0.010, 95% CI = [7.89e-03, 0.01], p < 0.001) and IRR = 1.0, p < 0.001. Remarkably, no significant difference in car collisions between occasional and chronic users was found, opposing the notion that chronic users are less impaired. Both chronic and occasional cannabis users have a similar car crash risk on simulator testing, with younger participants showing a higher risk, highlighting the need for targeted road safety campaigns. Practical Applications: The direct link between subjective perceptions of driving proficiency and actual performance suggests that interventions aimed at improving self-awareness could be beneficial. [Author's abstract]
Highlights:
Randomised trial methodology showing the effect of inhaled cannabis on car crashes.
Demonstration that cannabis is dangerous until 4 h after inhalation.
No difference between chronic and occasional users.
Direct link between subjective perceptions of driving proficiency and actual performance.