Impact of cannabis on the driving performance of youth: A systematic literature review

Author(s): Colonna, Alvarez, Kim, Chen, Chippure, Grewal, Randell, Leung

Slidedeck Presentation:

Robert Colonna Presentation 6A



Driving under the influence of cannabis contributes to the high rates of motor vehicle collisions (MVCs) among young drivers ages 15-24. Understanding the specific impact of cannabis on youth’s driving performance outcomes can inform targeted injury prevention and policy efforts.


This systematic literature review applied the American Academy of Neurology quality appraisal guidelines to determine the class of evidence (I- highest to IV-lowest) and level of confidence (A- high, B-moderate, C- low, and U-insufficient evidence) on the effects of cannabis on youth’s driving performance.


This study was registered in PROSPERO (#CRD42020180541) and followed Pati and Lorusso’s (2018) methodology. Seven databases were systematically searched, and pairs of independent reviewers screened records. Studies were included if they: were published in English; were primary peer-reviewed quantitative studies; included an on-road assessment or a driving simulator task; primarily included young drivers under the age of 25; and had primary driving performance outcomes.


Searches produced 1701 unique records. 13 primary studies (12 on driving simulators and one on both on-road and simulator) met inclusion criteria and underwent appraisal. Class II evidence suggests that THC likely decreases mean speed and headway distance, and increases deviation of lane position, deviation of steering wheel position, and reaction times among young drivers (Level B, moderate confidence). There is insufficient evidence to determine whether THC increases the crash risk of young drivers.


Findings indicate a moderate to low level of confidence in the impact of cannabis on the specific driving performance outcomes of young drivers.


There remains a need for Class I and II studies that focus on the specific effects of cannabis on young drivers and include larger and more representative samples.