Dose-Dependent Effects of Cannabis Edibles on Simulated Driving Performance: A Randomized Clinical Trial.

Le Foll B.; Matheson J.; Antwi P.; Wright M.; Zaweel A.; Hasan O.; Kloiber S.; Hassan A.; Sproule B.; Wickens C. et al. · JAMA network open · 2026

**Background & Methods** This randomized, double-blind, placebo-controlled crossover trial enrolled 40 healthy adult cannabis users (mean age 30.4 years) at a Canadian academic hospital to evaluate dose-dependent effects of THC edibles on simulated driving performance. Participants received soft-chew gummies containing 0, 2, 10, or 20 mg of delta-9-tetrahydrocannabinol (THC), with primary outcome measures of standard deviation of lateral position (SDLP) assessed at 2 and 5 hours post-exposure. **Key Findings** • THC dose-dependently impaired simulated driving performance: SDLP significantly increased at 20 mg (3.3 cm; P<.001) and 10 mg (2.0 cm; P<.001) relative to placebo, but not at 2 mg (0.7 cm; P=.16). • Secondary driving measures were adversely affected by higher THC doses: reaction time increased significantly at 20 mg (0.034 seconds; P<.001) and standard deviation of speed increased at 20 mg (0.469 km/h; P=.004) compared with placebo. • Peak whole-blood THC concentrations (3.4 ng/mL at 20 mg and 1.6 ng/mL at 10 mg) were at or below commonly used per se legal thresholds for driving impairment (2-5 ng/mL), yet significant performance deficits were observed. • Self-reported measures showed dose-dependent decreases: willingness to drive and perceived driving ability were significantly reduced at 20 mg (P<.001), suggesting some user awareness of impairment. **Dosage & Administration** THC edibles administered as soft-chew gummies in doses of 0, 2, 10, and 20 mg. Assessment occurred at 2 and 5 hours post-ingestion. **Safety & Adverse Effects** Not reported. **Evidence Quality** This is a well-designed RCT with appropriate blinding and crossover methodology, representing moderate-to-high quality evidence. Limitations include small sample size (40 participants), single-site recruitment limiting generalizability, and use of simulated driving rather than real-world driving assessment. The population was restricted to regular cannabis users aged 19-45, which may not represent all drivers. The finding that impairment occurs below legal per se thresholds has important implications for cannabis-impaired driving legislation, though additional research in naturalistic driving settings is warranted.

More recent cannabis research