Highlights • The study found no statistically significant difference in driving exposure, patterns and safety critical events between drivers with and without MCI. • However, drivers with MCI drove fewer kilometres, had shorter average trip distances and trip radius, reported a lower number of trips at night and on highways/freeways compared to drivers without MCI. • Drivers (with and without MCI) who reported better contrast sensitivity experienced fewer safety critical events.

Abstract Objective The aim of the study is to compare driving exposure, patterns and factors associated with safety critical events between drivers with MCI and a comparison group without cognitive impairment. Design, setting, participants Naturalistic driving data using an in-vehicle monitoring device were collected from 36 older drivers with MCI and 35 older drivers without cognitive impairment over a two-week period in Western Australia. Measurements Naturalistic driving exposure, patterns (eg. night-time trips, peak-hour trips) and safety critical events (harsh acceleration, harsh braking and harsh cornering). Results Drivers with MCI had a lower number of safety critical events (mean = 7.20, SD = 11.44) compared to drivers without cognitive impairment (mean = 10.89, SD = 23.30) however, this was not statistically significantly. There were also no statistically significant differences between drivers with and without MCI for measures of driving exposure or any of the driving patterns including weekday trips, night-time trips and trips on highways/freeways. The results of the multivariable modelling found only binocular contrast sensitivity was associated with the rate of safety critical events. For every increase of 0.1 log units in contrast sensitivity (better contrast sensitivity) the rate of safety critical events significantly decreased by 30 % (IRR = 0.70, 95 % CI = 0.50−0.98, p = 0.04). Conclusion Drivers with MCI were found to have similar driving exposure and patterns compared to older drivers without cognitive impairment, however drivers with better contrast sensitivity experienced fewer safety critical events. Future research should consider a longitudinal study design with an extended driving monitoring period and a larger sample with a clinical diagnosis of MCI to assess changes in cognition and its impact on driving.