The study recruited participants who met specific inclusion criteria: they were seasonal users of one of the following second-generation antihistamines—fexofenadine, bepotastine, levocetirizine, or ketotifen—due to hay fever, without being regular users throughout the year. Additionally, they should not have had any prior experience with driving simulators. Invitation to participate was extended through application forms distributed to individuals without any vested interest in the research. Exclusion criteria for the study included people taking medicines other than those in the study and people with ophthalmological conditions. Prior to eye movement measurements, it was confirmed that no participant had ophthalmological conditions that could affect their visual field. Informed consent was obtained from all participants after they were fully briefed on the study’s objectives.

For the experiment, a Mitsubishi Precision in-vehicle driving simulator was used. Participants engaged with the simulation through a large screen, 1.9 m in height and 3.1 m in width, which displayed the driving environment. Participants operated the simulator using realistic vehicle controls, such as the steering wheel, accelerator, and brake pedals, providing an experience akin to driving a real car (Fig. 1) The participants undertook a one-back task [9] as a calling task (Fig. 2) [10]. An audio file, pre-recorded with single-digit numbers spoken randomly every two seconds, was played during the driving simulation. Participants were tasked with responding to the number previously announced immediately following the announcement of the next number. To reduce the effects of familiarity with the task, participants underwent two days of preliminary training on the driving simulator and the one-back task, identical to the setup of the main experiment (Fig. 3). They used the simulator on different days to evaluate their driving performance and eye movements under taking and not taking medications. To account for potential fatigue from operating the simulator, the calling task was conducted at the beginning of the main test day, followed by the main test without the calling task. During the medication phase of the main study, participants consistently took each medication for at least seven days beforehand to achieve a stable concentration in their bodies. Following this phase, a washout period of at least one month was implemented before proceeding with the main study without medication.

Experimental setup showing the main view with an inset detail of the eye tracker in the lower right corner

Image of one-back task as a calling task

Experimental schedule of this study

The assessment of driving performance was structured around two distinct events. For a non-emergency event, the stopped vehicle was positioned within the lane of the participant's vehicle. The measure was the smallest distance maintained between the stopped vehicle and the participant's vehicle while passing (Fig. 4A). In an emergency event, a pedestrian unexpectedly emerged from behind the stationary vehicle as the participant's vehicle approached within 30 m of it from the opposite lane. The measure was the smallest distance between the pedestrian and the participant’s vehicle (Fig. 4B). Both events demonstrate that as the distance between the object and the participant's vehicle decreases, driving performance deteriorates.

Image of a non-emergency event (A), and an emergency event (B) in the driving simulator

Eye movements including saccades were monitored using a Tobii X2 Eye Tracker, a non-intrusive device positioned on the dashboard to avoid impeding the driver’s view (Fig. 1). Horizontal eye movements were analyzed by dividing the front screen into three zones: left, center, and right. A transition from one zone to another, except for the screen’s upper area which includes the rear-view mirror, was recorded as a single movement (Fig. 4A and B). Eye movement data were collected from a distance of 100 m prior to predefined events up to the occurrence of the event itself. The ratio of eye movements was then calculated for each participant, defined as the number of eye movements by an individual relative to the maximum number of eye movements recorded across all participants in the study.

Scatter plots were generated depicting the correlation between the smallest distance in a non-emergency event or an emergency event, and the ratio of eye movements. Spearman　rank correlation coefficient was calculated to determine the strength of these relationships. Additionally, the receiver operating characteristic (ROC) curve was plotted, and the area under the curve (AUC) values were computed to evaluate the usefulness of eye movement measurements as a discriminative index for detecting potential risks of driving performance impairment [11]. All analyses were conducted using JMP17 Pro software (SAS Institute, CA) A power analysis was conducted beforehand using G-power. With an effect size of 0.6, an alpha error of 0.05, and a power of 0.8, the required total sample size was calculated to be 17 participants. Therefore, 19 participants were included in this study. Additionally, post-hoc analysis revealed that Fig. 6D had a power of 0.934.