Longitudinal Simulated Driving Performance and Rates of Progressive Visual Field Loss in Glaucoma

Abstract

Purpose: The purpose of this study was to evaluate the association between longitudinal changes in driving performance, assessed through driving simulation, and rates of progressive visual field (VF) loss in patients with glaucoma.

Methods: Ninety-five patients with glaucoma underwent Standard Automated Perimetry (SAP) and driving simulations every 6 months. Rates of VF loss were estimated by changes in mean sensitivity (MS) of the integrated binocular VF over time. Driving performance was assessed using a simulator by maintaining lane position on a winding road while responding to peripheral visual stimuli to assess divided attention. Reaction time (RT) recorded the duration between the presentation of the stimuli and the participant's response. Linear mixed models evaluated longitudinal changes in SAP MS and mean RT. Multivariable linear regression models were used to predict driving performance, adjusting for age, cognitive impairment, and driving exposure.

Results: Progressive VF loss was associated with a longitudinal increase in mean RT to the divided attention task. In the multivariable model, each 1 decibel (dB)/year faster loss of integrated binocular MS was associated with a 0.024 logarithms (ln) s/year increase in mean RT (95% confidence interval [CI] = 0.007 to 0.042, P = 0.008). Baseline MS also significantly influenced driving performance, with each 10 dB worse baseline binocular MS associated with a 0.031 ln s/year increase in mean RT (95% CI = 0.016 to 0.045, P < 0.001).

Conclusions: Faster VF progression in patients with glaucoma was associated with worsening performance on a divided attention task during driving simulation.

Translational Relevance: Patients with glaucoma who exhibit faster VF progression may be at greater risk for a decline in driving performance.

Introduction

Glaucoma is a progressive optic neuropathy characterized by degeneration of retinal ganglion cells and their axons that may lead to significant visual loss.¹ The disease affects approximately 80 million people and is one of the leading causes of visual impairment in the world.^2,3 Standard automated perimetry (SAP) has traditionally been used to evaluate functional damage in glaucoma and to measure rates of disease progression.¹ However, the true clinical value of SAP depends on how well its results can be used to predict the impact of disease on quality of life (QOL) and the ability to perform activities of daily living, such as walking, reading, and driving.^4–8 It is essential to understand how changes in SAP, including the rate of visual field deterioration, are related to or predict future disability and decline in QOL. 

Several previous studies have demonstrated that SAP measurements are associated with patient-reported QOL outcomes.^4,9,10 In a longitudinal investigation, rates of progressive visual field loss were shown to correlate significantly with rates of change in Rasch-calibrated scores obtained from the 25-item National Eye Institute Visual Function Questionnaire (NEI VFQ-25).⁹ Although important, the use of surveys to gauge the impact of disease on daily living may be limited by the subjectivity of responses, which are influenced by a host of factors, such as patient perceptions, personalities, humor, and degree of anxiety.¹¹ This may lead patients to under- or overestimate their assessment of disability. In fact, some subpopulations of adults do not validly self-evaluate their everyday abilities; community-dwelling older adults with aging-associated cognitive impairment tend to overestimate functional abilities,¹² whereas depressed persons tend to underestimate abilities.^13,14 It is currently unclear whether and to what degree patients with glaucoma tend to under or overestimate their abilities. 

An alternative to the use of questionnaires is to directly evaluate the patient’s ability to perform a relevant task using performance-based measures. To be used in a clinical setting, it is essential for these performance tests to be reproducible and standardized. In the case of driving, real-world behind-the-wheel assessment of driving performance is difficult for a variety of reasons, including patient safety, costs, confounding factors, and difficult standardization.¹⁵ This generally makes it unfeasible to propose its use to longitudinally gauge how a disease that causes significant loss of vision, such as glaucoma, is affecting the ability to drive over time. In contrast, driving simulators may offer a useful alternative.^16,17 Simulators have become widely used to assess driving safety and performance under a variety of conditions.¹⁵ They have been used to assess general driving behavior in young and old populations¹⁷ and in diseases such as stroke,¹⁸ traumatic brain or spinal injury,¹⁹ Parkinson's disease,^20–22 Alzheimer’s disease,^23–25 and attention deficit disorders.²⁶ Their use can potentially help the evaluation of driving safety and performance of diseased subjects and provide insight into the different mechanisms involved in causing driving impairment. In a previous study, we have shown that metrics derived from driving simulator assessment were significantly predictive of risk of motor vehicle collisions (MVCs) in glaucoma.²⁷ 

In the current study, we propose to investigate the longitudinal relationship between visual field assessment with SAP and driving performance as measured with a simulator. We acquired longitudinal data on a cohort of subjects followed over time to assess how rates of visual field loss were related to driving parameters that have been shown to be relevant in assessing driving performance in real life. 

Methods

This study involved analysis of de-identified data obtained from a prospective cohort study designed to evaluate structural and functional damage in glaucoma (Diagnostic Innovations in Glaucoma: Functional Impairment, NIH EY021818 PI: Felipe A. Medeiros). Collection of data was performed at the Visual Performance Laboratory, University of California San Diego. Data analysis was performed at the Bascom Palmer Eye Institute of the University of Miami. The institutional review boards of both institutions approved the methods, and written informed consent was obtained from all participants. The study adhered to the laws of the Health Insurance Portability and Accountability Act, and all study methods complied with the Declaration of Helsinki guidelines for human subject research. 

All subjects participating in the study were active drivers with a valid driver’s license. Participants underwent a comprehensive ophthalmologic examination at baseline, including review of medical history, visual acuity, slit-lamp biomicroscopy, intraocular pressure measurement using Goldmann applanation tonometry, corneal pachymetry, gonioscopy, dilated ophthalmoscopy examination, stereoscopic optic disc photography, and SAP using the 24-2 Swedish Interactive Threshold Algorithm (SITA) Standard of the Humphrey Field Analyzer II-i, model 750 (Carl Zeiss Meditec, Inc., Dublin, CA, USA). Only subjects with open angles on gonioscopy were included and patients with coexisting retinal disease, uveitis, or non-glaucomatous optic disc neuropathy were excluded. Participants also completed the Montreal Cognitive Assessment (MoCA) test,²⁸ a cognitive screening tool developed to detect mild cognitive impairment, and completed a driving history questionnaire to ascertain average mileage driven per week. 

Glaucoma was defined by the presence of 2 or more consecutive abnormal SAP tests, defined as a pattern standard deviation with P < 0.05 and/or glaucoma hemifield test results outside normal limits, and evidence of glaucomatous optic neuropathy based on masked assessment of stereophotographs at baseline. A subject was considered to have glaucoma if damage was present in at least one eye. 

Driving performance was longitudinally evaluated using a driving simulator, and for inclusion in the study, subjects were required to have completed at least 2 driving simulations over a minimum follow-up period of 1 year. Visual field tests were performed approximately every 6 months during follow-up and subjects were required to have had at least 3 reliable visual field tests in each eye for the period spanning 1 year prior to the first driving simulation until the date of the last driving simulation. Over the course of the study, each patient was treated at the discretion of the attending ophthalmologist. 

Monocular and Binocular Visual Fields

Monocular SAP was performed using the 24-2 SITA Standard test at all visits during follow-up. Only reliable tests (≤33% fixation losses and ≤15% false-positive results) were included. In addition, visual fields were reviewed and excluded if there were artifacts present, such as eyelid or rim artifacts, fatigue effects, inattention, or inappropriate fixation. Visual fields were also reviewed for the presence of abnormalities that could indicate diseases other than glaucoma, such as homonymous hemianopia. To evaluate binocular visual field (BVF) loss, sensitivities of the monocular SAP threshold sensitivities of the right and left eyes were used to calculate an integrated BVF. The sensitivity for each point of the BVF was estimated using the binocular summation model described by Nelson-Quigg et al.²⁹ Evaluation of rates of visual field change was performed using the mean sensitivity (MS) of the BVF. Binocular MS was calculated as the average of the BVF threshold sensitivities for the integrated field. Furthermore, the overall BVF was classified into peripheral, central, inferior, and superior regions, following the method described by Abe et al.³⁰ The central points were located in the region encompassing approximately the central 10 degrees of the visual field, and the inferior and superior points were located in the regions encompassing the inferior and superior hemispheres, respectively. MS was calculated for each one of these regions by averaging the antilogs of the individual sensitivity thresholds and then recalculating the logarithm. Rates of visual field change were estimated using the MS of the integrated BVF, and the rates were also calculated in the same fashion per each specific region. 

Driving Simulator Protocol

A fixed-base interactive driving simulator was used. The driving simulator consisted of a typical driving seat, a steering wheel, brake and accelerator pedals, and a 40-inch screen. The feedback of the steering wheel was provided by a passive spring system. Surround sound was used to simulate wind, tires, and engine noise. The position of the seat, wheel, and pedals could be adjusted for comfort but the distance between the subject’s head and the center of the screen was set at 33.6 inches. The screen width was 35 inches resulting in a driving scene with a 45-degree horizontal field of view. The driving simulator is shown in Figure 1. 

Figure 1.

 

Subject performing the driving simulation.

In this study, we assessed the ability to divide attention by measuring reaction times to stimuli presented during a divided attention protocol during simulated driving. Previous studies have shown longer reaction times during divided attention tasks on simulated driving to be associated with increased risk of MVCs.^27,31 Subjects were required to focus on a driving task while their ability to divide attention was assessed by presenting a visual stimulus to the peripheral field. Stimuli were presented at about 20-degrees of visual angle in the upper right and upper left of the simulator screen throughout a curve negotiation task (Fig. 2) and the subject was required to push a button on the steering wheel to register perception of the stimuli. The contrast of the stimuli was randomly altered using alpha blending techniques to achieve symbol transparencies of 0.1, 0.4, and 0.9. Therefore, in the case of a 0.1 symbol transparency, the symbol intensity and color that the driver perceived was a summation of 10% of the symbol intensity and color and 90% of the background intensity and color. The equivalent Michelson contrasts were 0.04, 0.14, and 0.27 for low, medium, and high contrast stimuli, respectively. At maximum symbol intensity, the divided attention stimulus symbols were pure white, whereas the background was a constant cloudy sky. An average of 5 stimuli were presented at each contrast for each central driving task (a total of about 15 per 3 minutes or about 1 every 12 seconds) and stimuli stayed on the screen for a maximum time between 3 and 6 seconds (uniform distribution) or until the driver responded, whichever occurred first. The next stimulus appeared between 3 and 6 seconds (again uniform distribution) after the driver responded or when the maximum display time had elapsed. The main outcome measure of reaction time was defined as the time interval between the appearance of the peripheral stimulus and the subject pressing the button on the steering wheel, with a longer reaction time indicating a worse result. In our study, we report the results of low-contrast divided attention task, as previous studies demonstrated that reaction times to low-contrast stimuli were stronger predictors of MVCs than reaction times to high-contrast stimuli, with patients with glaucoma being more significantly affected by low-contrast tasks compared with similarly aged controls.^27,31 

Figure 2.

 

Screenshot of the driving simulator during the curve negotiation task illustrating the visual layout. The peripheral visual stimulus for the divided attention task is shown in the upper left corner.

During the curve negotiation task, subjects were instructed to drive in the center lane of a winding 3-lane road (see Fig. 2). The ability to divide attention was quantified by assessing the ability to attend to the central driving task (curve negotiation) and to simultaneously detect stimuli in the peripheral vision. As a subject might achieve fast reaction times by adopting a strategy in which the driving task is neglected, it was also important to evaluate and adjust for central driving task performance. Central task was quantified by calculating curve coherence, as a measure of correlation between the road curvature and path of the vehicle.³² Curve coherence was defined as the normalized cross-correlation function between the road curvature (k_road) and the vehicle path curvature (k_own) as a function of spatial shift. It is represented by the following equation:  

\begin{eqnarray*} && CurveCoherence\, =\, _{\arg \left( {delay} \right)}^{{\rm{Max}}} \\ && \quad \times \left\{ \frac{1}{n}\sum\limits_t {\frac{{\left( {{{k}_{own}}\left( t \right) - MEA{{N}_{{{k}_{own}}}}} \right)\left( {{{k}_{road}}\left( {t,delay} \right) - MEA{{N}_{{{k}_{road}}}}} \right)}}{{S{{D}_{{{k}_{own}}}}S{{D}_{{{k}_{road}}}}}}} \right\} \end{eqnarray*}

where n is the number of samples of the two signals and SD is the standard deviation of the signals, with a coherence of 1 indicating the two signals to be an exact match. 

Reaction time was chosen as the outcome variable as difficulties with divided attention tasks seem to be related, at least in part, to a slowing of visual processing speed. The visual processing speed, which is defined as the time needed to make a correct judgment about a visual stimulus, is commonly studied in behavioral research by measuring reaction times.^33–37 

Statistical Analysis

Mean reaction times were positively skewed, whereas curve coherence values were negatively skewed. Therefore, natural logarithms (ln) were calculated for further analysis to address these skewnesses. Rates of change in mean reaction times to the peripheral stimuli were estimated by linear mixed models.^38,39 Univariable and multivariable ordinary least squares linear regression models were then used to investigate factors predictive of rates of change in mean reaction time over time, such as rates of estimated BVF loss (globally and per location). Other variables examined as potentially confounding factors included baseline visual field severity, age, gender, race, cognitive impairment (MoCA score), and driving exposure (average mileage driven per week). We also considered change in visual acuity of the better eye as a potential explanatory variable. 

All statistical analyses were performed using commercially available software Stata, version 17 (StataCorp LP, College Station, TX, USA). The alpha level (type I error) was set at 0.05. 

Results

The study involved 95 subjects with glaucomatous visual field loss. Table 1 shows clinical and demographic characteristics of the included subjects. Mean age at baseline was 65.5 ± 12.1 years. There were 59 male subjects (62.1%) and 36 female subjects (37.9%). At baseline, the average SAP mean deviation (MD) was −1.7 ± 3.2 decibel (dB) and −5.0 ± 6.4 dB in the better and worse eyes, respectively. However, there was a wide range of disease severity, with MD values ranging from −28.7 to 2.6 dB. The average binocular MS at baseline was 29.3 ± 3.1 dB. Baseline binocular MS values were highly correlated to baseline MD values of the better eye (R² = 88.4%). The average MoCA score was 27.7 ± 2.1 units and the average driving exposure was 123.2 ± 88.2 miles/week. Mean reaction time to peripheral stimuli during curve negotiation was −0.79 ± 0.90 ln s and mean curve coherence was −3.47 ± 0.91 ln units at baseline. 

Table 1.

 

Demographic and Clinical Characteristics of Subjects Included in the Study (n = 95)

 

Subjects were followed for an average of 2.4 ± 0.8 years, from the date of the first visual field to the date of the last visual field closest to the last driving simulation. The median number of available SAP visual field tests during follow-up was 7 (interquartile range = 5 to 11) and the median number of available driving simulations was 5 (interquartile range, 3 to 7). The average rate of change in binocular MS was −0.29 ± 0.27 dB/year (interquartile range = −0.45 to −0.11 dB/year), the average rate of change in mean reaction time to peripheral stimuli during the curve negotiation task was 0.015 ± 0.026 ln s/year (interquartile range = 0.002 to 0.031 ln s/year), and the average rate of change in curve coherence was −0.065 ± 0.035 ln unit/year (interquartile range = −0.086 to −0.042 ln unit/year). 

Progressive visual field loss was associated with longitudinal increase in mean reaction time on the divided attention task during driving simulation. Each 1 dB/year faster loss was associated with a 0.055 ln s/year increase in mean reaction time (95% confidence interval [CI] = 0.039 to 0.071 ln s/year, P < 0.001, R² = 32.6%; Table 2; Fig. 3). Progressive visual field loss was not associated with longitudinal decline in the central task, as measured by the curve coherence parameter (P = 0.227). 

Table 2.

 

Univariable and Multivariable Regression Model Results Predicting the Rate of Change in Mean Reaction Time During Divided Attention Tasks

 

Figure 3.

 

Scatterplot with a fitted regression line showing the relationship between the rate of change in mean reaction time during divided attention tasks over time and the rate of change in integrated binocular mean sensitivity over time. Faster rates of visual field loss were significantly associated with increasing reaction times during the follow-up period.

Worse disease severity at baseline was associated with faster rate of change in mean reaction time. Each 10 dB worse mean baseline binocular MS was associated with a 0.048 ln s/year increase in mean reaction time, indicating worse performance (95% CI = 0.034 to 0.062 ln s/year, P < 0.001, R² = 32.6%; see Table 2; Fig. 4). Older age, worse cognitive impairment (MoCA score), and less driving exposure (mileage drive per week) were significantly associated with faster increase in mean reaction time to the divided attention task in univariable analyses (see Table 2). Rate of change in curve coherence over time, gender, race, and change in visual acuity of the better eye were not significantly associated with change in mean reaction time during follow-up (see Table 2). 

Figure 4.

 

Scatterplot with a fitted regression line illustrating the relationship between the rate of change in mean reaction time during divided attention tasks over time and baseline integrated binocular mean sensitivity. The plot shows that patients with lower baseline visual field sensitivity had a greater increase in reaction time over time.

In the multivariable analysis, progressive visual field loss remained significantly associated with increase in mean reaction time to the divided attention task. Each 1 dB/year faster loss of binocular MS was associated with a 0.024 ln s/year increase in mean reaction time (95% CI = 0.007 to 0.042 ln unit/year, P = 0.008; see Table 2). Additionally, baseline MS was also a significant predictor of driving performance, with each 10 dB worse baseline binocular MS associated with a 0.031 ln s/year increase in mean reaction time (95% CI = 0.016 to 0.045, P < 0.001). Age and cognitive impairment also significantly impacted reaction times. Each decade increase in age was linked to a 0.005 ln s/year increase in reaction time (95% CI = 0.001 to 0.008, P = 0.009), whereas a decrease of 5 units in the MoCA score was associated with a 0.014 ln s/year increase in reaction time (95% CI = 0.003 to 0.024, P = 0.010). 

Supplementary Table S1 presents the multivariable analysis of rates of visual field loss across different regions (central, peripheral, inferior, and superior), adjusted for potential confounders. In the analysis of central binocular visual field loss, mean reaction time increased by 0.018 ln s/year for each 1 dB/year faster loss (P = 0.032). For peripheral binocular visual field loss, mean reaction time increased by 0.026 ln s/year per 1 dB/year faster loss (P = 0.007). When evaluating inferior binocular visual field loss, mean reaction time increased by 0.022 ln s/year per 1 dB/year faster loss (P = 0.030). Similarly, for superior binocular visual field loss, mean reaction time increased by 0.015 ln s/year per 1 dB/year faster loss (P = 0.008). 

Discussion

The findings from this study demonstrate the significant impact of progressive visual field loss on driving performance in patients with glaucoma, as assessed by a driving simulator. Patients experiencing faster rates of visual field deterioration exhibited a more rapid increase in reaction time to a divided attention task during simulated driving compared to those with slower disease progression. This relationship, which persisted even after adjusting for potential confounders in the multivariable analysis, suggests that visual field decline may have substantial real-world implications, particularly concerning driving performance and the potential onset of driving disability in patients with glaucoma. 

Faster rates of estimated BVF loss were strongly associated with worsening performance in the divided attention task, as indicated by increased reaction times to peripheral stimuli while negotiating a curve. In the univariable model, each 1 dB/year increase in the rate of BVF loss resulted in a 0.055 ln s/year increase in mean reaction time. Even after adjusting for potential confounders in the multivariable model, this relationship remained significant, with each 1 dB/year increase in BVF loss corresponding to a 0.024 ln s/year increase in mean reaction time. Although the natural ln transformation was used in the statistical analyses to manage skewed data, interpreting these values can be challenging. Therefore, it is useful to express these changes as proportional increases. A 0.055 ln s/year increase translates to approximately a 5.65% increase in reaction time per year (calculated as exp(0.055) − 1 ≈ 0.0565). Similarly, in the multivariable model, a 0.024 ln s/year increase corresponds to about a 2.43% increase per year (exp(0.024) − 1 ≈ 0.0243). These incremental changes are cumulative over time, meaning that small annual increases in reaction time can lead to significant cumulative effects over several years. However, it is important to emphasize that these proportional increases are estimates rather than definitive conclusions, particularly given the relatively short average follow-up period in our cohort. 

Baseline BVF sensitivity also significantly influenced simulated driving performance. As expected, patients with greater visual field loss at baseline experienced faster increases in reaction time. A 10 dB worse baseline BVF sensitivity had an impact comparable to a 1 dB/year increase in the rate of BVF loss on reaction time. Importantly, the effects of baseline BVF sensitivity and the rate of BVF loss would be cumulative. Over a 5-year period, a patient with a baseline mean sensitivity of 25 dB progressing at a rate of −1 dB/year would be estimated to experience a cumulative increase of approximately 25.6% in reaction time. In contrast, a patient with more severe disease, having a baseline mean sensitivity of 15 dB and the same rate of visual field loss, would exhibit a 46.4% increase in reaction time over the same period. 

In a divided attention task, participants must respond to peripheral stimuli while focusing on driving, which simulates real-world scenarios where multiple visual and cognitive demands are placed on a driver.¹⁵ Reaction times are a crucial component of safe driving. Delays in responding to unexpected stimuli, such as obstacles or changes in traffic, can significantly increase the likelihood of crashes.²⁷ For patients with glaucoma who already face visual field deficits, added cognitive demands can further impair their ability to process and respond to critical driving events. A previous study by Gracitelli et al.³¹ showed a direct relationship between reaction times in driving simulation and real-world MVCs. In their prospective study, driving simulations were conducted in patients with glaucoma at baseline, and subsequent driving records were obtained from the Department of Motor Vehicles (DMV). The study found that each 1 standard deviation (SD) increase in low-contrast reaction time was associated with a 2.19-fold increase in the risk of MVCs during follow-up (hazard ratio [HR] = 2.19 per 1 SD slower, 95% CI = 1.30 to 3.69, P = 0.003). This finding underscores the predictive value of reaction time measurements in assessing driving safety among patients with glaucoma. Another study by Ogata et al.⁴⁰ further supports these findings, showing that patients with glaucoma had significantly longer reaction times than the healthy controls, especially under increased cognitive load, such as when talking on a mobile phone while driving. This added cognitive burden worsens the already reduced capacity to detect peripheral events, emphasizing the limited ability for divided attention in some patients with glaucoma. 

Our study also revealed significant associations between reaction times and other factors, such as age and cognitive function, as measured by the MoCA. Older age was significantly associated with increased reaction times, indicating that aging may exacerbate the challenges patients with glaucoma face in processing and responding to visual stimuli while driving. Similarly, lower MoCA scores, indicative of cognitive impairment, were associated with longer reaction times. These findings underscore the complex interplay among visual, cognitive, and age-related factors in influencing driving performance in patients with glaucoma. 

Despite the emerging trends toward autonomous driving, understanding the impact of visual impairments on driving performance remains essential. Fully autonomous driving systems are not yet ubiquitous or fully reliable, and many individuals will continue to rely on manual driving for the foreseeable future. As such, the ability to quickly and accurately respond to dynamic visual and cognitive demands remains crucial. Assessing driving performance, especially in individuals with visual impairments like glaucoma, is vital for ensuring public safety and preserving individual quality of life.¹⁵ Even with advancements in technology, the transition to fully autonomous vehicles will take time, and, until then, understanding and mitigating the risks associated with manual driving in visually impaired individuals will continue to be a key public health concern. 

Moreover, the findings related to reaction times in driving simulators can have broader implications beyond driving. A driving simulator can be viewed as a type of “stress test,” highlighting the impact of visual field loss on a person's ability to perform tasks that require quick responses under pressure.¹⁹ In everyday life, similar divided attention tasks are encountered frequently, whether crossing a busy street, navigating a crowded environment, or handling multiple demands at once in the household or work settings. Longer reaction times may not only increase the risk of driving accidents but could also lead to slower responses to potential hazards in various scenarios, thereby increasing the risk of injury or other adverse outcomes. 

This study has several limitations that should be acknowledged. The controlled environment of a driving simulator, while useful for standardization and safety, may not fully capture the complexities and unpredictability of real-world driving, which could lead to differences in observed performance. Although previous studies have validated the driving simulator metrics,³¹ assessing the repeatability and reproducibility of these performance measures would add further rigor. We also recognize that our cohort had a relatively short follow-up period. However, the relatively short follow-up period would more likely attenuate rather than exaggerate the associations observed, suggesting that with longer follow-up, these associations could be even stronger. Another limitation is that we did not systematically evaluate changes in lens opacity over time, which could influence visual performance. Although no significant relationship was found between changes in visual acuity and reaction times, the potential impact of cataracts or other media opacity changes cannot be entirely ruled out and warrants further investigation. Given the sample size, we were unable to stratify patients based on treatment type or surgical intervention, highlighting the need for further research to evaluate the impact of these treatments on driving performance over time. 

Furthermore, although our study focused on reaction times, it did not distinguish between visual processing delays and motor response delays. Although previous research suggests that reaction time in glaucoma is predominantly influenced by slower visual processing rather than motor impairment,⁴¹ further distinction between these effects would enhance our understanding of the specific mechanisms that affect driving performance in patients with glaucoma. Last, whereas we accounted for age and cognitive impairment using the MoCA, other cognitive domains, such as sustained attention and executive function, which also decline with age, were not specifically evaluated.^42,43 These cognitive abilities are critical for managing complex driving tasks, making quick decisions, and adapting to changing driving conditions.^44,45 Evaluating these factors in future studies could provide a more comprehensive understanding of how cognitive decline, in conjunction with visual impairment, impacts driving safety in patients with glaucoma. 

In conclusion, our study shows that progressive visual field loss in patients with glaucoma significantly affects driving performance, as evidenced by increased reaction times during divided attention tasks in a driving simulator. By linking SAP metrics with performance-based outcomes, such as driving ability, we can better understand the functional impact of glaucoma on daily activities. Further validation in additional cohorts is essential, as this knowledge is critical for clinicians in assessing and managing the risk of driving impairment. Providing targeted guidance and support based on these insights can help patients maintain their independence and quality of life. 

Acknowledgments

Supported in part by the National Institutes of Health/National Eye Institute grants EY021818 (F.A.M). 

Disclosure: D.A. Malek, None; A. Diniz-Filho, None; E.R. Boer, None; F.A. Medeiros, AbbVie (C), Annexon (C), Astellas Pharma, Inc. (C), Carl Zeiss Meditec (C), Enavate Sciences (C), Galimedix (C), Google Inc. (F), Heidelberg Engineering (F), nGoggle Inc. (P), Ocusciences, Inc. (C), ONL Therapeutics (C), Perfuse Therapeutics (C), Stealth Biotherapeutics (C), Stuart Therapeutics (C), Thea Pharmaceuticals (C), Reichert (C, F)