Visual impairment increases significantly with age as a result of the normal aging process, as well as through the increase in prevalence of ocular disease. These visual changes have a number of implications for driving performance and safety. This article discusses the impact of these age-related changes in visual function on driving ability and describes current evidence linking performance on visual tests and driving performance and crash risk.
Driving is a complex task that involves integration of a range of visual, cognitive and psychomotor skills, many of which are impaired with increasing age. As driving is considered to be a highly complex visual task, as illustrated in Figure 1, it has been suggested that the high prevalence of visual impairment in older populations contributes to the decrement in driving ability seen in older drivers. The increased prevalence of visual impairment with age results both from normal age-related changes as well as the age-related increase in ocular diseases including cataracts, glaucoma and age-related macular degeneration. These changes result in reductions in visual acuity, contrast sensitivity, visual fields, motion sensitivity and increased glare sensitivity and speed of visual processing.
Fig. 1: Illustration of the visual complexity of the vehicle and driving environment.
Visual acuity is the most commonly used vision test for determining driving eligibility, despite the fact that the link between reduced acuity and increased crash risk is unclear. The earliest studies of visual acuity and crash risk, showed only a weak correlation between acuity and crash rates . More recent studies, using a range of sample sizes and methodologies, also reported only a weak relationship between acuity and crash risk [2, 3, 4, 5, 6], and others have failed to find any association [7, 8, 9, 10, 11, 12].
This weak association between visual acuity and crash risk is not surprising given that the ability to resolve static high contrast targets is unlikely to represent the visual demands of the normal driving environment, which includes both static and moving objects of different sizes and contrast levels both in central and peripheral vision; tests of contrast sensitivity, peripheral vision and motion sensitivity may provide better measures of visual performance for driving.
Decina and Staplin  reported that contrast sensitivity, visual acuity and visual fields were strongly related to crash rates in older drivers and impaired contrast sensitivity is associated with retrospective , but not prospective crashes [9, 11, 12]. Importantly, these increased crash rates occur despite the fact that older drivers with impaired contrast sensitivity self-regulate and reduce their driving exposure [14, 15, 16]. Interestingly, crash-involved drivers with cataracts were eight times more likely to have reduced contrast sensitivity than controls . Closed road studies have also demonstrated a significant relationship between contrast sensitivity and driving performance for drivers with simulated  and true cataracts , and contrast sensitivity also predicts drivers’ recognition performance (signs, hazards and pedestrians) under both day and night-time conditions .
Studies linking visual field loss and crash risk have reported mixed results. Johnson and Keltner  demonstrated that binocular field loss more than doubled crash rates compared to controls, while other studies have failed to find a significant relationship between field loss and crash risk [1, 7, 9, 22]. However, a more recent population-based study , and smaller scale studies of drivers with glaucomatous field defects [23, 24], provide support for Johnson and Keltner’s  initial findings, demonstrating that only those with more extensive field loss have impaired driving performance and increased crash risk. This is reflected in closed road studies which demonstrated that performance is not significantly impaired until simulated field loss is more extensive . Similarly, on-road driving studies reported that it was only the drivers with more severe binocular defects that had significantly impaired driving ability [26, 27, 28]. Indeed, it has been suggested that merging the two monocular threshold fields, known as the integrated visual field (IVF), may be useful in assessing fitness to drive in patients with a range of field losses [29, 30], particularly as monocular fields are routinely assessed in patients with ocular disease; however, the link between IVF and crash risk has yet to be determined.
The ability to see moving targets has often been considered with respect to driving, given the dynamic nature of the driving environment. Recent research has shown that motion sensitivity, using computerbased tests of motion discrimination and detection, is strongly correlated with driving ability in older adults [19, 31, 32, 33, 34]. Increased age is associated with impairments in time-to-contact judgments, as well as in the perception of speed and heading [35, 36], which are critical skills in enabling smooth and fast responses on road hazards. Even brief delays in detecting moving targets may have fatal consequences, and reduced ability to discriminate speed or time to contact could result in increased crash risk.
A large body of research has also evaluated the ability of visual processing speed and divided attention to predict driving safety, using the Useful Field of View (UFOV) [37, 38]. The UFOV is a computergenerated task, involving simultaneous identification of central and peripheral targets presented in the presence or absence of distracters and relies on selective and divided attention and rapid visual processing speed which decrease with age . Reduced UFOV performance strongly predicts both retrospective [13, 39] and prospective crashes in general populations of older adults [9, 11, 12] as well as in those with ocular disease . A 40 per cent UFOV reduction was associated with a 2.2 times higher likelihood of incurring a crash than for those with no UFOV impairment . Studies have also reported strong associations between reduced UFOV scores and unsafe on-road performance [19, 40], and driving simulator performance . The UFOV is also effective at predicting prospective crash risk when administered in a driver licensing setting , providing further support for its inclusion for screening older drivers. Figure 2 shows the importance of UFOV within the context of a real-world driving scene.
Fig. 2: Illustration of the importance of the Useful Field of View in a realworld driving situation.
Older adults also experience higher levels of disability glare than younger adults, due to an increase in intraocular light scatter which reduces visual acuity and contrast sensitivity, particularly in the presence of cataracts . Driving situations in which disability glare may cause difficulties include at night from approaching headlights, and at dawn and dusk from the low-sun. Accordingly, older drivers with higher levels of disability glare self-regulate and reduce their driving exposure, particularly at night . While previous research has failed to find significant links between disability glare and at-fault crash-rates , there is some evidence that increased glare sensitivity is linked with crash involvement in older drivers . This is consistent with findings from a simulator study of older adults which demonstrated poorer driving safety when performing turns against on-coming traffic in the presence of glare, particularly for low-contrast vehicles .
With the rise in eye diseases among older adults, and associated reductions in visual functions, it is not surprising that many eye diseases are linked with reduced driving performance and increased
crash risk. The most common eye conditions affecting older adults, and their ability to drive safety, include cataracts, glaucoma and agerelated macular degeneration .
Many people live with cataracts for extended periods before cataract removal , with around a quarter of these continuing to drive, even if their vision does not meet the legal driving standards [47, 48]. Drivers with cataracts report increased driving difficulties, drive shorter distances and avoid challenging driving situations . Despite limiting their driving exposure, drivers with cataracts have 2.5 times more crashes than controls , and as discussed above, those involved in crashes are eight times more likely to have reduced contrast sensitivity . These findings are supported by closed and open road studies, which have shown that drivers with either simulated [49, 50], or true cataracts [19, 51, 52], have impaired driving performance compared to controls. Cataract surgery has a positive impact on crash rates and driving performance. A recent meta-analysis suggested that cataract surgery is associated with a 88% reduction in the risk of driving-related difficulties . Crash rates for cataract patients who undergo surgery were half those who did not have surgery . Retrospective analysis of population-linked data (1997-2006) demonstrated that cataract surgery reduced crash risk by 13% with savings of $4.3 million . Consistent with these findings, closed road driving performance improved following bilateral cataract surgery, with the improvement being best predicted by the change in better eye contrast sensitivity . These studies provide compelling evidence that cataract surgery can provide tangible benefits to road safety , and should be targeted at an earlier rather than later stage to maintain safe driving.
Patients with glaucoma also report driving difficulties, including problems with glare and night driving . Glaucoma is cited as one of the main reasons that older drivers give up driving ; patients with bilateral glaucoma are three times more likely to cease driving that those without glaucoma . Glaucoma has been shown to be an important risk factor for both self-reported crashes  and state-recorded crashes [23, 24, 61, 62]. Patients with moderate or severe glaucomatous field loss were six times more likely to have an at-fault crash and four times more likely to cause any crash than those with no field loss , while people involved in injurious crashes were 3.6 times more likely to have glaucoma than those who were crash-free . The increased crash risk for drivers with glaucoma has also been linked to their deficits in UFOV performance . On-road assessment of glaucoma patients has highlighted problems with lane-keeping, negotiating curves and anticipatory skills , and they are more likely to have a driver instructor intervention during on-road driving assessments compared to controls . Early detection and treatment of glaucoma is clearly critical, not just to minimise irreversible visual field loss, but also to maintain safe driving ability. Few studies have addressed the impact of age-related macular degeneration (AMD) on driving, with most focussing on the selfreported aspects of driving rather than objective measures of driving performance and safety. Drivers with AMD report more difficulty driving , self regulate their driving habits and avoid challenging driving situations [16, 65], and exhibit less risk taking behaviours  than drivers without AMD. While older adults with AMD exhibited poorer driving performance on an interactive driving simulator and an in-traffic road test , these impairments were not reflected in higher crash rates . The paucity of data in this area highlights the need for further research to evaluate the extent to which AMD impacts on driving performance and crash risk, and to identify strategies to improve driver safety in this population. One of the problems for older drivers is that changes in their vision
can occur gradually; they may thus be unaware that their vision has become impaired until a driving incident occurs or they are advised of visual changes following an eye examination. It is unlikely that changes in visual ability relevant to driving will be detected if high contrast acuity continues to be used to assess visual performance for driver licensing. While recent evidence strongly suggests that alternative measures of visual function, including the UFOV, contrast sensitivity and motion sensitivity may be more relevant to driving performance and safety, evaluation and validation of these tests has yet to be completed. In addition, while there is strong evidence linking the presence of ocular disease and driving difficulties, further research is needed in order that appropriate advice can be provided to older drivers with these conditions.