We conducted a pilot study with 20 active drivers aged 60 years or older recruited from the Columbus, Ohio, area. All participants held valid driver’s licenses and were active drivers, defined as those who drove at least once per week. We excluded drivers who drove less than once per week, used visual aid devices beyond habitual correction (eg, bioptic lens), or had conditions preventing them from driving, such as cognitive impairment or eye diseases. Non-English speakers and individuals with a history of severe motion sickness were also excluded. The study data were collected between November 2021 and August 2022.

We distributed information about our study through channels that had proven successful in our previous research: (1) a website listing and Facebook advertising via StudySearch, a directory managed by the Ohio State University (OSU) Center for Clinical and Translational Science [33]; (2) ResearchMatch, a registry of volunteers managed by the Vanderbilt Institutes for Clinical and Translational Research, spanning a network of 183 institutions [34]; (3) Twitter and Instagram posts facilitated by Nationwide Children’s Hospital; and (4) study flyers distributed at local community centers. Social media postings or email links directed interested individuals to complete an initial eligibility survey on our study website or to meet our study team over the phone for eligibility screening.

Following expressions of interest, our study coordinator contacted the individuals, rescreened them for eligibility, and obtained signed consent. Subsequently, the study coordinator scheduled an in-person appointment for visual function assessments at the OSU College of Optometry’s Laboratory of Vision Enhancement. Additionally, participants underwent a driving performance assessment on a high-fidelity driving simulator at the OSU Driving Simulation Laboratory. These 2 assessments took place within 1 month of each other.

Gathered demographic information included age, sex, race, marital status, employment, income level, self-assessed overall health, and eye health. Self-reported driving history included average hours driven per day and week and levels of difficulties while driving in various situations in dim light or at night.

We conducted a descriptive analysis of participants’ demographics and driving variables. We excluded 2 participants from the analyses due to unusable driving data resulting from procedure adjustment or technical difficulties. We used ANOVA, with Bonferroni correction for multiple comparisons, to compare vision scores across 3 lighting conditions. Additionally, we used ANOVA with repeated measures to assess the driving performance variable across 3 lighting conditions. We calculated Pearson correlation coefficients to evaluate the relationships between visual function and driving performance under each lighting condition, and we used the false discovery rate for multiple testing. We conducted all data analyses in SAS and completed the analyses by January 2024, with a significance level of α=.05.

The study received ethical approval from the Institutional Review Board at Nationwide Children’s Hospital (STUDY00000461). Written informed consent was obtained from all participants. The study was performed in accordance with the ethical standards as laid down in the Declaration of Helsinki and its later amendments or comparable ethical standards and the applicable local regulatory requirements and laws.

Across all 20 participants, the average age was 70.3 (ranging from 63 to 83), with 45% (n=9) female. A total of 6 participants had had cataract surgery, and 1 participant received a cataract diagnosis but had no surgical intervention. On average, participants had held their driver’s license for about 55 years, and self-reported eyesight averaged 7.6 on a scale of 1-10. In total, 50% (n=10) of the participants reported driving difficulties, ranging from a little difficulty to extreme difficulty, due to concerns about their eyesight in dim light or at night. Additionally, 45% reported driving less in dim light or at night because of their eyesight (Figure 2A). Participants indicated that they avoided driving in dim light or at night due to moderate or extreme difficulties, including driving in difficult conditions (n=8, 40%), driving in unfamiliar places (n=6, 30%), noticing objects off to the side (n=5, 25%), or reading street signs (n=4, 20%) (Figure 2B).

The average VA was –0.07 (SD 0.08) in the photopic (daytime) condition, significantly lower (indicating better function) than the average VA in the mesopic (nighttime) condition (mean 0.23, SD 0.11; P<.001) and in the mesopic with glare condition (mean 0.44, SD 0.17; P<.001) (Table 1). The average AULCSF was 1.51 (SD 0.27) in the photopic condition, significantly greater (indicating better function) than the average AULCSF in the mesopic condition (mean 0.081, SD 0.20; P<.001) and in the mesopic with glare condition (mean 0.32, SD 0.21; P<.001), which was the worst among the 3 lighting conditions. Finally, the average VUSVFM in the photopic condition was 907.08 (SD 47.05); while lower (indicating worse function) than in the mesopic condition (mean 917.89, SD 78.60; P=.60), this difference was statistically insignificant. The lighting conditions accounted for 74%, 83%, and 9% of the variation of VA, AULCSF, and VUSVFM, respectively.

Of the 18 participants included, the average driving speed was 71.55 (SD 8.89) km/h in the daytime condition, significantly slower than that in the nighttime with glare condition (74.60, SD 13.97 km/h) but not significantly different from that in the nighttime without glare condition (73.71, SD 16.33 km/h; P=.28) (Table 2). The average SDLP in the nighttime with glare condition was 0.11 (SD 0.07) meters, significantly greater than that in the daytime (mean 0.07, SD 0.04 meters; P<.001) and the nighttime without glare condition (mean 0.08, SD 0.04 meters; P<.001). Furthermore, the average reaction time was 1.23 (SD 0.29) and 1.21 (SD 0.38) seconds in the nighttime conditions with and without glare, both significantly longer than the average reaction time in the daytime condition (mean 1.07, SD 0.31 seconds; P<.001).

For functional vision in the photopic condition (Table 3), poor VA was significantly correlated with greater SDspeed (r=0.26; P<.001) and greater SDLP (r=0.31; P<.001) (Figure 3A). Poor AULCSF was associated with greater SDLP (r=–0.22; P=.007) (Figure 3B).

For functional vision in the mesopic condition (Table 3), poor VUSFVM was significantly correlated with slower average speed (r=–0.24; P=.007) and larger SDSpeed (r=–0.19; P=.04), greater SDLP (r=–0.22; P=.01) (Figure 3C), and longer reaction times (r=–0.22; P=.04) (Figure 3F) while driving at night.

For functional vision in the mesopic condition with glare (Table 3); Poor VA was significantly correlated with longer reaction times (r=0.21; P=.046) while driving at night with glare (Figure 3D). Poor AULCSF was significantly correlated with slower speed (r=–0.32; P<.001), greater SDLP (r=–0.26; P<.001) (Figure 3B), and longer reaction times (r=–0.2; P=.04) (Figure 3E) while driving at night with glare. No other significant correlations were observed between visual function and driving performance under the same light conditions.

Furthermore, poor photopic VA was significantly correlated with greater SDLP while driving at night with (r=0.25; P<.001) and without (r=0.28; P<.001) glare, respectively [see Table S1 in Multimedia Appendix 1]. Poor photopic AULCSF was significantly correlated with greater SDLP across the 3 lighting conditions, longer reaction times (r=–0.24; P<.001) while driving at night, and larger SDSpeed (r=–0.41; P<.001) while driving at night with glare. Poor photopic VUSFVM was significantly correlated with a larger SDLP (r=–0.25; P<.001) while driving at night with glare. For functional vision in the mesopic condition, poor VA, AULCSF, and VUSFVM were significantly correlated with greater SDLP and/or longer reaction times while driving at night with or without glare. For functional vision in the mesopic condition with glare, poor VA and AULCSF were correlated with longer reaction times in all 3 lighting conditions.

This pilot study examined the relationships between visual function and driving performance across 3 lighting conditions among drivers aged 60 years and older. The main findings highlight that worse visual function and driving performance were observed in the nighttime condition relative to the daytime condition, particularly with glare. Poor visual functions were largely correlated with poor driving performance. Specifically, poor photopic VA and CSFs were significantly associated with greater SDLP (poor lane keeping ability) while driving in all 3 lighting conditions. Poor mesopic VA, CSFs, and VFM were significantly correlated with greater SDLP and longer reaction times while driving at night with or without glare. Poor VA and CSFs under mesopic conditions with glare were significantly correlated with longer reaction times in all 3 lighting conditions. Our results contribute to the current literature on the relationships between visual functions and nighttime driving and have potential implications for the nighttime driving safety of over 45 million licensed drivers who are aged 65 years and older [11].

Crash statistics indicate that the nighttime driving environment is dangerous for all road users, particularly older drivers [19,54]. Our findings also reveal that participants’ SDLP was significantly greater at night with glare than in the other 2 lighting conditions. Participants also took longer to respond to stimuli (eg, reacting to the boxes appearing above the roadway) during both nighttime conditions compared to the daytime condition. The observed decline in driving performance among older drivers during the nighttime could be attributed to reduced visibility [55] and a combination of age-related vision changes, including reductions in rod sensitivity, slower dark adaptation, decreased VA, and increased sensitivity to glare [5,6,8,9]. All these factors contribute to an increased risk of nighttime crashes [5,53]. Given that access to transportation by car is crucial for the well-being and independence of many older individuals, future studies using comprehensive vision tests are imperative. Such studies can help identify robust and unambiguous indicator(s) of older individuals’ fitness to drive at night, ensuring their driving privileges and road safety [7,19,56,57].

Consistent with previous findings [5-7,28], this study reveals that poor VA and poor CSF were associated with increased SDLP across all 3 lighting conditions—an indicator of poor vehicle control. Moreover, this study found that individuals with poor VA, CSF, and VFM may face greater challenges in maintaining a stable lateral position on the road and take a relatively long time to respond to an event critical to driving safety, particularly while driving at night. These results suggest that compromised visual functions may lead to driving challenges at night, particularly with glare [6]. Therefore, assessing VA alone during daylight, as current standard vision tests do, without considering the effect of nighttime vision or glare, may inaccurately characterize functional vision [6].

Puell et al [14] observed that reduced contrast sensitivity, especially in nighttime conditions with glare, might contribute to challenges during night driving. Wood and Owens [6] found that visibility significantly diminishes during night driving, with contrast sensitivity tests better predicting real-world object recognition performance than VA alone. This study also found that visual function in the mesopic and mesopic with glare conditions was significantly associated with driving performance at night, particularly with glare conditions. Poor vision in low-light conditions is correlated with poor recognition of traffic signs, as low lighting reduces the reading distance of road signs, leaving less time to react. These effects are especially pronounced in older drivers who drive at night [7]. Our results, along with those from other studies [6,7,14], suggest that assessing VA, along with CSF, VFM, and other visual functions, is crucial for characterizing age-related vision changes when evaluating the relationships between vision and driving. These metrics could also serve as targets for interventions to mitigate driving challenges, especially at night, among older adults [27,29].

This pilot study is the first to assess the relationships between visual function and driving performance under different lighting conditions among drivers aged 60 years and older. Understanding the impacts of age-related visual deficits on nighttime driving is critically needed because age-related change in vision is the most significant risk factor for crash-related injuries and deaths at night in older individuals [4-9]. Future research and practical applications could explore developing targeted strategies for addressing specific visual impairments in older drivers in the context of distinct lighting challenges [13,20]. These strategies include considering using polarizing eyeglasses or other glare mitigation techniques [58]. Future studies should also advocate for including novel vision tests that mimic the effects of nighttime conditions and a glare source [5,27,48]. Such assessments may become imperative due to demographic changes (ie, longer life expectancy) compounded by increased dependency on private car journeys among older people [59,60].

This study has several limitations. First, the small sample size and the absence of a control group, along with the lack of data on participants’ other health-related conditions (eg, comorbidities), restricted our ability to conduct more advanced statistical analyses. Adjusting for potential known confounding factors, such as participant age, motor function, and cognitive capacity, was limited. Second, our results may be susceptible to selection bias, possibly reflecting a more active or healthier subset of older drivers. Therefore, caution is needed when generalizing our findings to all drivers aged 60 years and older. Third, the driving measured in this study relied on simulated driving performance in a high-fidelity driving simulator, which may not fully reflect participants’ real-world driving behaviors. Consequently, a cautious interpretation of our results is necessary. Fourth, the participants completed all visual function tests while sitting in a stable chair rather than in a moving vehicle. Thus, the observed relationships between visual function and driving performance in this study may be under- or overestimated. Despite these limitations, our study contributes empirical data on visual function and driving across 3 lighting conditions. We used a set of comprehensive and sensitive visual function measures coupled with simulated driving performance assessed on a high-fidelity driving simulator.

Drivers aged 60 years and older face a heightened risk of fatal nighttime crashes, primarily attributed to age-related vision changes. This pilot study presents preliminary data indicating that VA, contrast sensitivity, and VFM may distinctly impact driving performance in various lighting conditions among older drivers, particularly at night and in the presence of glare. Further research with a larger and more diverse sample is essential to validating these results. Our results underscore the importance of comprehensive visual function assessments, especially under nighttime and glare conditions, for evaluating age-related vision changes. Additionally, the study findings advocate for developing therapeutic interventions and technological aids to enhance safe nighttime driving among older drivers. As the older adult population is expected to nearly double by 2050, increased research efforts in this domain are warranted to guide clinical practices and policies, ensuring the maintenance of driving privileges and road safety for individuals aged 60 years and older.