Predicting driving impairment from visual and oculomotor impairment after alcohol intake

Abstract

Driving under the influence of alcohol has serious adverse individual and societal consequences. While the legal limit of blood alcohol content (BAC) is typically 0.08% and many people subjectively feel capable of driving unimpaired after 2–3 drinks, it is reasonable to think that any amount of alcohol in the bloodstream could impact driving ability to some degree. Here we examined the relationship between low-level neural impairment (as evidenced from visual and visuomotor testing using an ocular tracking task) and operational driving impairment (as evidenced from performance testing and modeling using a simulated lane-keeping task) after alcohol intake. For the lane-keeping task, the display (95°×63°) simulated a participant driving a virtual vehicle down a lane while facing crosswind perturbations (sum-of-sines motion: 0.1–2.19Hz). Participants (n=23, 13 females) used a steering wheel to control heading of the vehicle to keep it centered in the lane. For the ocular tracking task, the same participants tracked the step-ramp motion of a target dot (diameter: 0.64°) with its speed (16°/s–24°/s) and direction (0°–360°) randomly varied from trial to trial. We tested four BACs (0, 0.02%, 0.04%, & 0.08%). Model-independent analyses show that, while precision of lane-keeping (measured by RMS error) was affected only at the highest BAC tested (0.08%), significant effects of alcohol on visual and oculomotor function (measured by a composite ocular tracking index) are evident at the lowest BAC (0.02%). Furthermore, several ocular tracking parameters are significantly correlated with the phase lag of lane-keeping at 0.08% BAC (r2 >0.22). Model-dependent analyses however show that, while reaction time and neuromuscular stability appear degraded starting at the lowest BAC, the ability of our participants to generate compensatory lead control for lane-keeping appears to increase with BAC. This counterintuitive finding may explain the subjective, yet false, confidence in driving ability at 0.08% BAC (about 3 standard drinks).

Acknowledgement: Supported by research grants from the National Natural Science Foundation of China (31741061), Shanghai Science and Technology Committee (15DZ2270400, 17ZR1420100), and NYU-ECNU Joint Research Institute at NYU Shanghai.