• Caffeine not only reduced drowsiness, but also disrupted its impact on driving errors.
• Placebo group: replicated the drowsiness-driving errors link.
• Caffeine group: driving errors significantly reduced and remained independent of drowsiness.
• Results extend the view that caffeine protects performance by reducing drowsiness, our data show a direct impact of caffeine on task ... [Show full abstract] errors.
Aim: The study examined the moderating effect of repeat-dose, chewing gum-administered caffeine on the well-established relationship between drowsiness and driving performance, under the conditions of accumulating sleep loss. Method: 50-h sleep deprivation protocol with a double-blind, placebo-controlled design. Eleven volunteers (6 male), aged 18-28 years were screened for pre-existing medical conditions (including sleep disturbances), tobacco and recreational drug use, recent time-zone travel and shift-work. They were randomly allocated to placebo or caffeine group and administered 4 oral doses of either caffeinated gum pellets (200 mg/dose) or non-caffeinated placebo gum every two hours (01:00, 03:00, 05:00, 07:00) on the first and second nights of the protocol. Participants were constantly monitored and remained awake for 50 h, while performing 15 identical, evenly-spaced 40-min monotonous driving tasks in a medium-fidelity moving-base driving simulator. Their drowsiness was monitored with a spectacle frame-mounted infra-red sensor registering ocular parameters and converting them into a Johns Drowsiness Scale (JDS) score every 60 s. Lane keeping and speed variability measures were used to assess driving performance. Results: Driving performance declined and drowsiness increased from the first simulated drive to the last. When driving performance was examined in one-minute epochs synchro-nised with JDS scores, both lateral lane positioning and speed variability were found to be associated with drowsiness. The strength of this association was significantly weaker in the caffeine group, compared to placebo. Placebo group replicated the linear relationship between drowsiness and driving errors across the full range of JDS scores. This pattern was significantly weaker under the caffeine condition, and was even reversed at the upper range of JDS, with higher JDS scores not resulting in further degradation of driving performance. This dissociation between drowsiness and driving errors persisted across the 24-h cycle under the caffeine condition, despite caffeine being administered only during early morning hours. Conclusion: Strategically timed, repeat 200 mg doses of caffeine administered via chewing gum can mitigate fatigue-induced impairments in driving performance by not only reducing drowsiness but also by significantly weakening its impact on driving errors. This dual effect of sustained drowsiness reduction and the dissociation between drowsiness levels and driving errors seems worth further investigation as it might offer an effective emergency countermeasure against driver drowsiness and its subsequent conversion into potentially fatal driving errors.