Driving under low-contrast visibility conditions in Parkinson disease

Abstract

Objective: To assess driving performance in Parkinson disease (PD) under low-contrast visibility conditions.

Methods: Licensed, active drivers with mild to moderate PD (n = 67, aged 66.2 +/- 9.0 years, median Hoehn-Yahr stage = 2) and controls (n = 51, aged 64.0 +/- 7.2 years) drove in a driving simulator under high- (clear sky) and low-contrast visibility (fog) conditions, leading up to an intersection where an incurring vehicle posed a crash risk in fog.

Results: Drivers with PD had higher SD of lateral position (SDLP) and lane violation counts (LVC) than controls during fog (p < 0.001). Transition from high- to low-contrast visibility condition increased SDLP and LVC more in PD than in controls (p < 0.01). A larger proportion of drivers with PD crashed at the intersection in fog (76.1% vs 37.3%, p < 0.0001). The time to first reaction in response to incursion was longer in drivers with PD compared with controls (median 2.5 vs 2.0 seconds, p < 0.0001). Within the PD group, the strongest predictors of poor driving outcomes under low-contrast visibility conditions were worse scores on measures of visual processing speed and attention, motion perception, contrast sensitivity, visuospatial construction, motor speed, and activities of daily living score.

Conclusions: During driving simulation under low-contrast visibility conditions, drivers with Parkinson disease (PD) had poorer vehicle control and were at higher risk for crashes, which were primarily predicted by decreased visual perception and cognition; motor dysfunction also contributed. Our results suggest that drivers with PD may be at risk for unsafe driving in low-contrast visibility conditions such as during fog or twilight.

Figure 1 A scene during the drive, shown under both high-visibility (clear sky; A) and low-contrast visibility (fog; B) settings

Figure 2 Final moment in the intersection incursion scenario (A) and plots of driver reaction and vehicle kinematics after intersection incursion (B) Left column: The final moment in the intersection incursion scenario for 3 participants (A, B, and C). The multiplex view shows 4 channels of video data with superimposed digital driving data. The upper left panel shows the driver’s face and the view through the rear window (as seen in the rear view mirror). The upper right panel provides an over-the-shoulder view of the driver’s actions with superimposed steering wheel position (degrees). The lower left panel provides a record of the participant’s control of the foot pedals with superimposed percentage application of the brake and accelerator pedals. The lower right panel displays the view of the forward roadway that the driver should see, with superimposed speed (miles per hour). Right column: Diagrams of vehicle kinematics and the vehicle path after the intersection incursion was triggered (4.0 seconds before the intersection as determined by driver speed) in 3 participants. The common ordinate scale shows the driver’s vehicle speed, percentage of pedal application for accelerator and brake, and steering wheel rotations in degrees (upward deflections are counterclockwise rotations). The x-axis ends at the expected position of the incurring vehicle. The upper panel shows vehicle path inside the lane. (A) A control participant able to stop timely before colliding with the incurring vehicle. At 1.83 seconds after the trigger, he releases the accelerator and starts braking with a smooth deceleration. The brake is applied 82% at the time of stopping. (B) The driver with Parkinson disease (PD) reacts late (2.73 seconds) and collides with the incurring vehicle at 34 mph. (C) This driver with PD does not perceive the incurring vehicle, does not react at all, and crashes at 58 mph.