Abstract

Advanced driving simulators are important for testing challenging driving scenarios under controllable, repeatable, and safe conditions in the context of driving rehabilitation, training, and assessment, which are particularly relevant for older drivers. Simulator sickness (e.g., nausea, fatigue, disorientation) is a common side-effect of driving simulators, not only affecting passive passengers but also the driver. The goal of the present study was to investigate how control over the vehicle may impact the severity of simulator sickness in a high-fidelity driving simulator, particularly in older adults. Thirty-four healthy participants (65+ years old; 13 women) were engaged in a driving simulator task under conditions that varied in the extent to which they had active control over the simulated vehicle (manual vs. fully automated). Various baseline measures (visual acuity, cognitive abilities, mood) were recorded, and simulator sickness was measured using the Simulator Sickness Questionnaire after each drive. No differences in simulator sickness severity were observed between the two driving control conditions (manual vs. fully automated), but women reported significantly more simulator sickness than men. In addition, a positive relationship between simulator sickness and cognitive abilities was found, indicating that better cognitive performance was associated with more simulator sickness. Additionally, in terms of sensory abilities, better visual acuity was linked to more severe simulator sickness. Our findings suggest that controllability of a vehicle may not have a large effect on the severity of simulator sickness in older adults, but that biological sex as well as cognitive and sensory abilities may be relevant factors worth considering.

Keywords: motion sickness, automation, sex, age, cognition

Acknowledgements: The authors thank Bruce Haycock and Colin Stoddart for technical support and Tim Fraser for providing photographs of the laboratories.

Funding: This work was supported by grants from the Canadian Institutes of Health Research, AGE-WELL (Alex Mihailidis), the Schwartz/Reisman Emergency Medicine Institute, Schwartz Reisman Institute in Technology and Aging, and the Vector Institute for Artificial Intelligence (Shabnam Haghzare).

Disclosures: The authors have no conflicts of interest to disclose.

Data Availability: The data presented in this article were collected as part of a larger research project but have not been published previously. The study data will be made available upon request.

Open Access License: This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License (CC BY-NC-ND). This license permits copying and redistributing the work in any medium or format for noncommercial use provided the original authors and source are credited and a link to the license is included in attribution. No derivative works are permitted under this license.

Correspondence concerning this article should be addressed to Behrang Keshavarz, KITE Research Institute, Toronto Rehabilitation Institute–University Health Network, 550 University Avenue, Toronto, ON M5G 2A2, Canada. Email: [email protected]

Driving simulators are popular tools in many domains such as training, rehabilitation, assessment, and research and development. They serve as a safe and controlled proxy for on-road driving (Bruck et al., 2021; Klüver et al., 2016; Riegler et al., 2019), as they allow for the systematic introduction of challenging driving conditions and situations that are not safe to test in the real world (e.g., impaired driving). Unfortunately, driving simulators can cause simulator sickness (Kennedy et al., 1990), a specific type of motion sickness (Cha et al., 2021; Keshavarz & Golding, 2022). Common symptoms of simulator sickness include, for instance, nausea, cold sweats, eye-strain, fatigue, dizziness, and/or headache. The exact cause of simulator sickness is not fully understood and several theories have been proposed, including the roles postural stability (Riccio & Stoffregen, 1991; Smart et al., 2002, 2014) and sensory conflict (for overviews, see Keshavarz et al., 2014; Lawson, 2014b). The sensory conflict theory is one of the most prominent theories (Oman, 1990; Reason & Brand, 1975), arguing that simulator sickness is a result of incongruent information delivered to the brain by the visual, vestibular, and somatosensory systems. For example, in a fixed-base driving simulator, the visual system indicates self-motion, whereas the vestibular and somatosensory systems indicate stasis. If this sensory conflict is novel to the user and no adequate expectation of this conflict has been previously established at a neural level, simulator sickness may occur (Reason, 1978).

The incidence rate of simulator sickness in driving simulators is difficult to estimate and can range from anywhere between 1% (Klüver et al., 2015) to 95% (Stanney & Hash, 1998). This very broad range can be attributed to the fact that various components impact simulator sickness, including visual factors (e.g., field-of-view), the driving scenario (e.g., highway vs. urban), and the physical motion capabilities of the simulator (e.g., fixed-base vs. motion base). In addition, interindividual differences such age have often been discussed as relevant contributors to simulator sickness (see Classen et al., 2011, for an overview). For traditional motion sickness, susceptibility is reported to peak around 8–10 years of age and typically decreases thereafter, with older adults being less susceptible to motion sickness compared to younger adults (Jones et al., 2019). This is often explained by habituation to sickness-inducing environments over the lifespan. In contrast, for simulator sickness, older adults (65+ years of age) have been shown to be at increased risk of experiencing more severe symptoms of simulator sickness compared to younger adults (Brooks et al., 2010; Domeyer et al., 2013; Keshavarz et al., 2018). This finding might be explained by the fact that many older drivers are likely to have accumulated a long history of real-world driving-related experiences, and, as such, have a strongly established set of expectations related to (multi)sensory motor contingencies (e.g., the associations between the motor act of braking and the expected sensory feedback associated with deceleration). Consequently, experienced drivers could be more sensitive to slight deviations of sensory feedback during simulated driving compared to real-world driving, which may result in increased simulator sickness (Johnson, 2007). Partial support for this finding is delivered by Stoffregen et al. (2017), who compared simulator sickness in highly experienced and novice drivers who were actively engaged in a virtual reality (VR)-based driving task. Results suggested a faster onset of simulator sickness symptoms in the more experienced drivers, whereas the overall severity of simulator sickness did not differ between the two groups. Interestingly, no difference between experienced and novice drivers with regards to simulator sickness was found when participants were passively exposed to a driving scene (C.-H. Chang et al., 2021). Given that older adults are one population group who might particularly benefit from modern driving simulation technologies (e.g., in the context of rehabilitation or assessments of fitness-to-drive), it is important to address the issue of simulator sickness in this age group specifically.

One aspect of simulator sickness that is not well studied is the effects of having active control over the vehicle simulator’s motion (i.e., as a driver), compared to passively experiencing these same motions (i.e., as a passenger). For traditional motion sickness, it has been demonstrated that passengers are ∼46% more prone to experiencing motion sickness compared to drivers (Schmidt et al., 2020) and that active control of a vehicle is associated with a significantly lower risk of experiencing motion sickness compared to passive transport (Dam & Jeon, 2021; Iskander et al., 2019; Turner & Griffin, 1999). This finding has been explained by the fact that the driver can correctly anticipate the vehicle’s upcoming motions, which, in turn, reduces the likelihood of a conflict between a driver’s anticipation of the vehicle motion and its actual motion (Fukuda, 1976; Rolnick & Lubow, 1991). However, the situation is more complex for simulated driving. According to Johnson (2007), simulator sickness can occur in situations that are not necessarily causing motion sickness in the real world when not properly simulated. Thus, unlike in real-world situations, drivers in simulated environments can indeed experience simulator sickness even if they are in active control of the vehicle simulator’s motion during situations when the simulation is imperfect. In the context of low-fidelity driving simulation in gaming, there is some evidence that active control in a virtual driving game (i.e., being a “driver”) results in lower levels of simulator sickness compared to passively watching the same video replayed (i.e., being a passenger; Curry et al., 2020; Dong et al., 2011; Stoffregen et al., 2014). However, these studies were conducted using driving video games presented on computer monitors or head mounted displays and did not utilize more complex, high-fidelity driving simulators with physical motion capabilities. Further, to our knowledge, no studies have compared simulator sickness during active versus passive driving in older adults, who are the primary targeted user group for many driving simulation applications. Beyond simulation design, these considerations are also broadly relevant to the rapid emergence of automated vehicle technologies, which will inevitably result in the conversion of all currently active drivers to all passive passengers. Indeed, older adults will also be one of the largest consumers of automated vehicles in the future given the overall high percentage of older drivers due to the aging population, as well as the safety benefits afforded by automated vehicle control in drivers experiencing sensory, motor, and/or cognitive decline. As such, understanding whether older adults experience greater adverse side effects during passive motion experiences (physical and/or visual) compared to actively controlled motion, could have broad implications.

The Present Study

The goal of the present study was to investigate how control over a simulated vehicle may affect the severity of simulators sickness, in particular among older adults who may be at higher risk of experiencing simulator sickness compared to younger adults. Thus, the present study utilized a high-fidelity driving simulator to compare simulator sickness in older adults during active (i.e., manual) versus passive (i.e., fully automated) driving conditions, while also considering the role of biological sex. Specifically, it has been previously suggested that female participants may report more simulator sickness than male participants (Garcia et al., 2010; Mourant & Thattacherry, 2000). Based on the sensory conflict theory, as well as previous literature, we hypothesized that (1) manual driving would cause less severe simulator sickness than automated driving in older adults and (2) that women would report more severe simulator sickness than men.

In addition, we collected a variety of baseline measures of cognition, visual acuity, mood, and driving history, as these factors often change over the course of typical aging and might also affect the level of simulator sickness. For instance, it is possible that poorer visual acuity (even within normal ranges) reduces a potential visual-vestibular conflict, resulting in less severe simulator sickness. It is also possible that participants who have a longer history of driving are more prone to simulator sickness, given that more experienced vehicle operators may be more susceptible to simulator sickness than less experienced operators (e.g., as in the case of pilots; C. M. Webb et al., 2009). Changes in mood such as fatigue, tension, or anger might also be related to simulator sickness, considering that fatigue-related symptoms (sopite syndrome, Lawson & Mead, 1998) are often prominent during simulator sickness. Although we did not have specific hypotheses regarding the relationship between these baseline measures (i.e., cognitive-, sensory-, mood-, and driving history-related factors) and simulator sickness, we investigated potential associations among these factors using exploratory analyses.

Method

Participants

Forty-one participants were recruited for this study. Three participants discontinued the study after the practice trials and three did not complete all drives due to simulator sickness (two men, one women) and were therefore excluded from the data analysis. Another participant with self-reported vestibular deficiencies was also excluded, resulting in a final sample size of 34 older adults (13 women, 21 men) with ages ranging from 65 to 90 years (M_age = 73.41, SD_age = 6.10). Participants reported to be healthy with no severe medical conditions (no recent history of stroke, psychological disorders, musculoskeletal disorders) and had normal or corrected-to-normal vision. In addition, participants were prescreened for their susceptibility to motion sickness using the Motion Sickness History Questionnaire (Kennedy et al., 1992), inquiring about the frequency of motion sickness in various situations (never, seldom, often, always). Highly susceptible participants (i.e., reporting motion sickness often or always in two or more situations) were not invited to participate in this study. All participants had a valid driver’s license and were actively driving, with an average of 52 years of driving experience (range 25–67 years). Ten participants (26%) reported having some previous experience with driving simulators. This research complied with the American Psychological Association’s Code of Ethics and was approved by the institutional review board at the University Health Network. Informed consent was obtained from each participant. Compensation of $15/hr was provided to the participants for their time.

Study Design

A one-factorial design including the within-subject factor vehicle controllability (manual, automated) was implemented. For each controllability level, participants were asked to perform three separate drives representing different driving conditions (clear daytime with low traffic, rainy daytime with low traffic, clear daytime with high traffic) in order to test across a representative range of driving environments and challenges, for a total of 6 drives total per participant. The mean simulator sickness score was calculated from these three drives for each controllability level for further statistical analyses (no significant differences were found between the three driving conditions for the manual and automated levels). The order of driving scenarios was counterbalanced across participants via a Latin square design to control for potential carryover effects.1

Apparatus

The experiment was conducted using DriverLab (Figure 1), a high-fidelity driving simulator located at the KITE Research Institute, the research arm of the Toronto Rehabilitation Institute, University Health Network. DriverLab is equipped with a full-sized passenger vehicle (Audi A3) mounted on a rotating turntable inside a high-resolution projection system dome creating a seamless 360-degrees field-of-view with surround sound. A hydraulic hexapod motion platform (Rexroth) provided physical motion cues along 6 degrees-of-freedom (Bruck et al., 2021). Together with the rotations about the yaw axis provided by the turntable, DriverLab offers physical motion along a total of 7 degrees-of-freedom.

Stimuli

The level of controllability was manipulated to include a manual (active) and an automated (passive) driving condition. In the manual condition, participants were entirely in control of the vehicle and manually operated the vehicle’s motions. Participants were asked to drive as they naturally would on-road and to obey all traffic rules (e.g., speed limit, right of way, traffic signals). To ensure identical routes across all participants, direction signs were positioned before each intersection to convey the instructed directions. In the automated condition, the vehicle was fully controlled by the simulated automation system using SCANeR Studio (Version 1.7, OKTAL), and it was programmed to follow the same route as the manual scenarios.

All experimental drives were designed and displayed using SCANeR Studio, simulating the same time of day (i.e., noon) with identical road structures including three four-way intersections. Each drive took approximately 8–10 min to complete. To minimize carryover and adaptation effects (e.g., related to practice or anticipation of the upcoming scene), each drive started and ended at a different point within the scenario. In addition, the study included two shorter (approx. 5 min) practice scenarios including one manual and one automated practice scenario. The practice scenarios were included in the design to familiarize the participants with the driving simulator and their task and to ease them into the simulated environment. As such, both practice scenarios started in a straight, low-speed (40 km/h) road with no visual distractions while gradually leading to curved, higher speed roads, and increasing visual elements (e.g., buildings).

Dependent Measures

Participants’ levels of simulator sickness were measured once at baseline, once after each of the two practice scenarios, and once after each of the six experimental driving scenarios using the Simulator Sickness Questionnaire (SSQ; Kennedy et al., 1993). The SSQ evaluates 16 symptoms of simulator sickness (e.g., general discomfort, nausea, eyestrain, vertigo, fatigue) on a 4-point scale ranging from 0 (not at all) to 3 (severe). The SSQ yields the three subscores nausea, disorientation, and oculomotor issues as well as a total score, all of which were calculated using the specific weighting procedure recommended by the authors.

Baseline Measures

A variety of baseline measures were assessed to characterize the participant sample and to account for health-related outcomes that may affect the participants’ driving performance and/or their level of simulator sickness. These measures included:

Montreal Cognitive Assessment

Cognitive abilities were screened using the Montreal Cognitive Assessment (MoCA; Nasreddine et al., 2005), which is a validated tool to screen for mild cognitive impairment, with a cutoff score of 26 or greater out of 30 representing normal cognition.

Visual Acuity

Visual acuity was measured using the Early Treatment Diabetic Retinopathy Study (ETDRS) test, an eye chart that consists of 14 lines of five letters with varying letter sizes. A logarithm of the minimal angle of resolution (logMAR) score was calculated from the ETDRS chart for statistical analyses, with lower logMAR scores indicating better visual acuity. Visual acuity was measured with visual correction for each eye independently and was then averaged.

Mood

Participants completed the Grove and Prapavessis (1992) abbreviated, validated version of the Profile of Mood States (POMS), which includes 40 questions across seven subscales of Tension, Anger, Fatigue, Depression, Esteem-related Affect, Vigor, and Confusion. Total Mood Disturbance is calculated by summing the totals for all the subscales.

Driving History, Health History, and Other Demographics

Participants were asked to indicate their years of driving experience and how often they drive on average (never, sometimes, every day). In addition, participants self-reported their health status indicating the presence of vestibular disorders and arthritis as well as vision and hearing problems. Basic demographics such as biological sex, age, handedness, and highest level of education were also included.

Procedure

The study consisted of two visits to the study site. The first visit was approximately 1 hr in duration and included the administration of the consent form, baseline tests, health history questionnaire, driving history questionnaire, and demographics questionnaire followed by the completion of two practice driving scenarios in the driving simulator. Immediately upon the completion of each practice scenario, participants completed the SSQ. The second visit was approximately 2 hr long and included the completion of the six experimental driving scenarios. Participants completed the SSQ immediately after each of the six experimental driving scenarios. Between the driving scenarios, participants were asked to step out of the driving simulator and were provided with a short rest break.

Data Analysis

For the data analysis, the SSQ scores for each of the three drives varying in daylight, traffic, and weather conditions were averaged within each controllability condition because no differences in simulator sickness showed between these three drives, resulting in a single SSQ score (per subscale and total score) for manual driving and a separate single SSQ score for automated driving for each participant. Data analysis was conducted using the statistical software R(R Core Team, 2021). A priori significance level was set to α = .05. Note that the study data will be made available to researchers upon request.

Results

Effect of Controllability and Biological Sex

A 2 x 2 mixed factorial analysis of variance including the within-subject factor vehicle controllability (manual, automated) and between-subjects factor biological sex (male, female) was conducted for each of the SSQ subscales nausea, disorientation, oculomotor, and the total score. Figure 2 shows the mean scores for each SSQ subscale separated by vehicle controllability and biological sex. No significant differences in the level of simulator sickness showed between the manual and automated driving conditions, including for the SSQ subscores nausea, F(1, 32) = 0.241, p = .627, η² = .007, disorientation, F(1, 32) = 0.469, p = .498, η² = .014, oculomotor issues, F(1, 32) = 0.153, p = .698, η² = .005, and the total score, F(1, 32) = 0.042, p = .839, η² = .001. A main effect of biological sex was found for the SSQ subscales nausea, F(1, 32) = 8.135, p = .008, η² = .203, disorientation, F(1, 32) = 5.243, p = .029, η² = .141, and total score, F(1, 32) = 5.858, p = .021, η² = .155, where women reported significantly higher scores than men. No sex-related differences were found in the SSQ oculomotor subscore, F(1, 32) = 2.076, p = .159. No interaction between biological sex and vehicle automation was found for any of the SSQ scores.

Relationship Between Cognitive Ability, Visual Acuity, Mood, Driving History, and Simulator Sickness

Table 1 summarizes the descriptive results for the MoCA, ETDRS, POMS, and driving history. For the MoCA, a total of 10 participants scored 26 or lower. To explore the relationship between simulator sickness, cognitive abilities (as measured by the MoCA), visual acuity, mood, and driving history (in years) exploratory Spearman correlations were calculated. Given that no differences in SSQ scores were found for the two vehicle controllability levels, we averaged the SSQ scores across these two factors for the correlation analysis. The results are given in Table 2 and revealed positive and moderately strong, significant correlations between all SSQ scores and the MoCA scores, suggesting that participants with better cognitive abilities tended to report more simulator sickness. Moderately strong, negative correlations were also found for the level of visual acuity and all SSQ subscales, suggesting that those with better visual acuity (i.e., lower ETDRS scores) reported more severe simulator sickness. No significant correlations were observed for mood (POMS) or for the number of years of driving experience.

Discussion

The aim of the present study was to investigate the occurrence and severity of simulator sickness in older adults during simulated manual and automated driving. In contrast to our original hypothesis, we found no significant differences between manual, active control of the vehicle and automated, passive driving with regards to simulator sickness. Further, while women experienced more severe simulator sickness than men overall, this finding was independent of vehicle controllability. Interestingly, higher degrees of simulator sickness were associated with better cognitive function and better visual acuity.

Simulator Sickness and Vehicle Controllability

Previous work using VR technologies demonstrated that active control of visual motion significantly reduced the severity of simulator sickness compared to passive exposure in younger adults (Curry et al., 2020; Dong et al., 2011; Stanney & Hash, 1998; Stoffregen et al., 2014); however, this has not been carefully evaluated for older adults and in the context of high-fidelity driving simulation in particular. In the present study, we did not find supporting evidence showing that simulator sickness is lower in manual versus automated driving modes in older adults within the conditions tested. This may point toward potential age differences in simulator sickness susceptibility when comparing automated and manual driving levels that could be tested in future studies with both a younger and older adult participant group.

Several reasons might explain the observed null finding with regards to vehicle controllability. The most obvious explanation is that there is indeed no difference in automated and manual driving with respect to simulator sickness in older adults. Further, although the raw data does not indicate any trending patterns of differences between the two controllability levels, we cannot dismiss insufficient power (1-β = .67 for large effects) as a potential reason for our findings, necessitating a careful interpretation of the null findings. Additionally, participants were prescreened for motion sickness susceptibility and only those who were less susceptible participated in this study. Susceptibility to traditional motion sickness is not necessarily a strong predictor of simulator sickness (Golding et al., 2021; Keshavarz et al., 2023), but it is possible that in a more motion sickness-susceptible population the severity of simulator sickness would have been generally higher and a difference between manual and automated driving might have emerged.

The associations between simulator sickness and active versus passive driving have also become increasingly relevant considering the looming debut of fully automated vehicles as an anticipated disruptive technology in transportation (Society of Automotive Engineers Level 5). The possibility of experiencing motion sickness during on-road automated driving is a serious concern, which could potentially hinder the successful adoption of fully automated vehicles (Diels & Bos, 2016; Diels et al., 2016). Reassuringly, the results of the present study suggest that, for older adults, passively experiencing vehicle motion during fully automated driving conditions may not significantly increase the risk of adverse side effects, such as motion sickness.

Biological Sex and Simulator Sickness

In the present study, we found sex-related differences with regards to simulator sickness severity, with women reporting more severe symptoms than men. Importantly, this sex-related difference was irrespective of vehicle controllability and was observed during both active and passive driving conditions. The role of biological sex has been discussed in the literature with controversy (see Lawson, 2014a, for an overview). For motion sickness, large-scale surveys showed higher self-reported motion sickness susceptibility in women compared to men (Dobie et al., 2001; Golding, 2006; Zhang & Sun, 2020). For instance, in a recent online survey by Schmidt et al. (2020) with more than 4,400 participants across different nations, the authors found that women reported experiencing more motion sickness as car passengers compared to men overall, and biological sex was one of the strongest predictors of self-reported motion sickness susceptibility. Experimental studies have found heterogenous results, with some reporting higher motion sickness severity in women compared to men (Flanagan et al., 2002; Klosterhalfen et al., 2005), while others reporting no sex-related differences (Cheung & Hofer, 2003; Classen et al., 2021).

For motion sickness induced by visual stimulation such as in simulators or VR applications (more generally referred to as visually induced motion sickness, VIMS; Kennedy et al., 2010; Keshavarz et al., 2014), the impact of biological sex is even more convoluted. Large-scale surveys on self-reported VIMS found mixed results (Keshavarz et al., 2023; Lukacova et al., 2023), and experimental studies are also highly inconsistent with respect to whether VIMS is higher in women compared to men (see E. Chang et al., 2020; Saredakis et al., 2020, for reviews). In a recent review article, Lawson and Bolkhovsky (2023) summarized a total of 76 studies that investigated any type of motion sickness and found that only 37 studies (48.7%) found increased motion sickness severity in women compared to men. The authors conclude that “… it is necessary to reserve [the researcher’s] judgement on this issue until conclusive evidence has been obtained” (p. 144).

A convincing explanation for potential sex-related differences does not yet exist. It has been argued that differences in the endocrine system may contribute to a potential sex-related difference. Changes in the hormonal system have been shown to affect susceptibility to motion sickness and simulator sickness (Grunfeld & Gresty, 1998; Matchock et al., 2008), with women’s susceptibility to motion sickness being potentially affected by their menstruation cycle (Clemes & Howarth, 2005; Golding et al., 2005, but see Bannigan et al., 2024; Hemmerich et al., 2019). A recent study by Stanney et al. (2020) investigated different factors that may potentially drive sex-related differences in motion sickness ratings when using VR headsets specifically and found evidence that interpupillary distance was one of the main contributing factors that led to increased VIMS ratings in women compared to men. The authors showed that when VR headsets were accurately adjusted to a user’s interpupillary distance, sex-related differences in VIMS ratings disappeared. Last, it has been speculated that women might be more honest when it comes to reporting VIMS compared to men, although empirical evidence for this assumption is weak (Cheung & Hofer, 2003; Park & Hu, 1999) and several studies have suggested that reporting bias might not be a valid explanation for sex-related differences in motion sickness research (Dobie et al., 2001; Flanagan et al., 2005). Of note, an imbalance in the ratio between male and female participants of VR studies has been pointed out by Peck et al. (2020), cautioning the interpretation of findings that investigate sex-related differences in headset-based VR-sickness research.

The Roles of Cognitive Impairment, Visual Acuity, and Mood

Through exploratory correlations, we found positive associations between simulator sickness severity and cognitive functioning, suggesting that those with better cognitive performance on the MoCA reported more simulator sickness. The reasons for this remain speculative; however, it is possible that domains of cognitive functioning such as slower processing speed and poorer executive functioning could also be associated with reduced sensitivities to sensory conflict. For example, it has been previously shown that older adults with mild cognitive impairment demonstrate a wider window of sensory integration meaning they continue to bind sensory inputs that are incongruent in space or time (Chan et al., 2015). Therefore, older adults with poorer cognition may be less sensitive to the visual-vestibular incongruencies that may be introduced by simulated driving.

We also found a significant, negative correlation between visual acuity and simulator sickness severity, indicating that better visual acuity (i.e., lower logMAR score) was associated with greater simulator sickness. The reason for this effect remains speculative, but it is possible that better visual acuity provides more reliable visual inputs, thereby amplifying the perception of a visual-vestibular conflict compared to when visual inputs are less reliable. In other words, when visual acuity is poorer, the sensation of self-motion extracted by the visual system might be diminished, resulting in a perceived sensory conflict that is less strong, resulting in less severe side effects. The relationship between visual acuity and simulator sickness has not received much attention in the past. To the best of our knowledge, the only study that measured visual acuity and motion sickness elicited through an optokinetic drum found opposite results, with lower visual acuity being linked to higher motion sickness (N. A. Webb & Griffin, 2002). However, the study by Webb and Griffin tested younger adults, which may explain the contrasting results in our study. The different findings might also be due to the different types of stimuli used in the present study and their study. That is, Webb and Griffin used an optokinetic drum with black-and-white horizontal bars, whereas our study deployed a driving simulator with more complex and richer visual scenes in which better visual acuity is more relevant for the correct perception of visual motion cues that may trigger a sensory conflict.

Finally, no significant correlations were found between simulator sickness and mood as measured by the POMS. Although anxiety (Grassini et al., 2021) and other situational factors have been previously linked to the occurrence of simulator sickness, the impact of these factors remain unclear. Here, our exploratory analysis of mood did not show any relationship with simulator sickness.

Limitations and Future Directions

As previously mentioned, the small sample size and the resulting test power require a cautious interpretation of the nonsignificant findings. Our test power was adequate to address strong effects, but we cannot rule out moderate or less strong differences in simulator sickness between manual and fully automated driving conditions with certainty. A larger sample size in future studies will help to further address this question. Also, as highlighted in the results, the overall severity of simulator sickness experiences by our participants can be considered moderate. Future studies using a more simulator sickness-inducing driving scenario (and/or including participants who are more prone to motion sickness) may help to better understand nuanced differences between active and passive driving conditions. Last, our study sample was comprised only of older adults, but did not include a younger group of drivers. Age has been frequently discussed as a potential factor for simulator sickness (Brooks et al., 2010; Domeyer et al., 2013), and it would be interesting to investigate in future studies whether a similar pattern between manual and fully automated simulated driving shows for younger healthy adults.

Conclusion

The present study investigated whether simulator sickness differs in simulated automated and manual driving in older adults. Overall, we found no significant differences in simulator sickness observed as a function of controllability, with both manual and automated driving eliciting comparable levels of simulator sickness. However, we did find a significant sex-related difference in the level of simulator sickness overall, with women reporting more severe simulator sickness than men. Interestingly, we also found that lower sickness scores were associated with poorer cognitive abilities and poorer visual acuity. In general, these findings may have implications for the use of simulated vehicle technologies in applications targeting older adults, as well as for the safe and comfortable use of automated vehicles by older drivers.

Supplementary Materials

Copyright © The Author(s) 2024

Received January 18, 2024
Revision received July 11, 2024
Accepted July 15, 2024