Marihuana: effects on simulated driving performance†☆

Abstract Performance of 23 male subjects was tested under smoked marihuana treatments on 4 occasions in a complex driving simulator. Doses were 0, 50, 100 and 200 micrograms delta-9 tetrahydrocannabinol per kilogram bodyweight. The simulator uses a car mounted on a chassis dynamometer facing a filmed scene subtending 160 degrees. Twenty-five performance measures are derived based on steering wheel, brake and accelerator pad usage as well as speed and tracking. The simulator also incorporates a visual search-and-recognition task based on random appearance of lights in the periphery. There is little evidence for a significant effect of marihuana upon car control and tracking. None of the 25 car control-tracking scores was significantly changed in either mean or variance by the treatments. However, there was a clear, statistically significant decrement in performance of the search-and-recognition task. Marihuana produced increased errors in recognition of the lights and delayed response times to their appearance. The results suggest the prime locus of marihuana impairment of driving performance as being in the interference with perceptual processes involved in data acquisition necessary for safe control of the vehicle.