| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||||
RESEARCH ARTICLE |
1 School of Public Health
2 University of California Traffic Safety Center
3 Departments of Bioengineering and School of Optometry, University of California at Berkeley.
Address correspondence to Dr. William A. Satariano, Division of Epidemiology, School of Public Health, 140 Warren Hall, University of California at Berkeley, Berkeley, CA 94720-7360. E-mail: bills{at}uclink.berkeley.edu
 |
Abstract |
|---|
 TOP Abstract MethodsResultsDiscussionReferences |
|---|
Methods. Data for this study (N = 1,840) were obtained from participants in a community-based study of aging and physical performance in people age 55 or older in the city of Sonoma, California. Each of 16 visual conditions was assessed for impact on reported driving limitation due to eyesight by calculating a "risk" ratio. Then, prevalence of the condition was combined with the ratio to generate an attributable risk for that condition for vision-related limitations in driving.
Results. Each condition was significantly associated with reported limitations in driving due to eyesight. "Avoiding physical activity due to vision" (ratio = 3.4) and "trouble seeing steps up/down stairs" (ratio = 2.9) had the strongest association. However, "glasses/contacts required for driving" and "trouble with glare from sun/lights" had the highest attributable risks (35.8 and 29.4).
Discussion. The risk ratio is relevant for evaluating individuals; the attributable risk is relevant to planning countermeasures in populations. Addressing specific problems related to vision should substantially reduce driving limitations due to eyesight.
Among older adults, increased age is associated with reduced and modified driving behavior (Ball, Owsley, Stalvey, Roenker, Sloane, & Graves, 1998
; Gilhotra, Mitchell, Ivers, & Cummings, 2001; McGwin, Chapman, & Owsley, 2000; Stewart, Moore, & Marks, 1993; West et al., 1997). There is also evidence that women are more likely to stop driving (Campbell, Bush, & Hale, 1993; Stewart et al., 1993) and that reductions in driving are associated with greater depressive symptoms (Fonda, Wallace, & Herzog, 2001; Marottoli et al., 1997).
Problems with vision represent one of the leading chronic conditions associated with reduction in or cessation of driving. Vision difficulties have been assessed in a variety of ways that include diagnosed conditions and performance on visual tests. Older drivers with cataracts and other vision difficulties are more likely to restrict their own driving and to be involved in motor vehicle crashes (Dellinger, Sehgal, Sleet, & Barrett-Connor, 2001; Lyman, McGwin, & Sims, 2001; Owsley, Stalvey, Wells, & Sloane, 1999; Owsley, Stalvey, Wells, Sloane, & McGwin, 2001). Reduced performance on standardized visual assessments is also associated with a reduction in driving in older adults (Ball et al., 1998; Gallo, Rebok, & Lesikar, 1999; Gilhotra et al., 2001; Marottoli & Richardson, 1998; McGwin, Sims, Pulley, & Roseman, 2000; Sims, Owsley, Allman, Ball, & Smoot, 1998; Stutts, 1998).
There is less research on self-reported vision problems or symptoms associated with limitations or avoidance of driving among older drivers (Dellinger et al., 2001; West, Gildengorin, Haegerstrom-Portnoy, Lott, Schneck, & Brabyn, 2003). A better understanding of the barriers to driving associated with visual deficiencies experienced by older drivers is needed. This would demonstrate the extent to which a limitation or cessation of driving is due to a loss of confidence in the ability to drive and, if so, the extent to which problems with vision contribute to a loss of confidence. Moreover, research in this area should contribute to the development of strategies to extend safe driving years for older adults.
In a previous study of older adults in Sonoma, California, current drivers were much more likely to identify problems with eyesight as a reason for limiting their driving than any other medical condition (Ragland, Satariano, & MacLeod, 2004). The current report expands on that study, examining the role of (a) disease processes affecting vision, (b) reported troubles with vision, (c) physical symptoms affecting the eyes, and (d) objective vision tests in reporting limitations to driving because of vision.
|
METHODS |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
Based on the 1990 U.S. Census of residents in Sonoma, California, ages 55 and older, the original sample of 2,085 subjects overrepresented adults ages 65–73 years (41.3% vs. 38.8% in the general population of Sonoma) and underrepresented adults over age 85 years (7.3% vs. 8.7%). Although differences existed between eligible participants and nonparticipants, the differences do not suggest a consistent pattern with respect to functional disability and chronic illness (Satariano, Smith, Swanson, & Tager, 1998).
The study sample is also somewhat more affluent and educated than the population of Sonoma residents age 55 and older. Nevertheless, a comparison of the sample with California residents age 55 and older indicates that the only residents who were underrepresented in the sample were those with household incomes less than $10,000. Similarly, the modal income category of our sample ($25,000–49,000) is the same as the modal income category for the entire state. Finally, there is very little difference among Sonomans or Californians in the percentage of households with annual incomes of $50,000 or more.
The current analysis is based on data from the baseline assessment of the SPPARCS cohort. The sample was restricted to subjects who were current drivers who reported whether or not they had personally either limited or avoided driving (N = 1,840). Driving status was established from the participants' report of their driving behavior and driver license history. Driving status was determined by the participants' answer to three separate questions: (a) Have you ever had a valid driver's license? (b) Do you currently have a driver's license? (c) Have you driven a car in the last 6 months? Participants were considered "current" drivers if they responded "yes" to all three questions. Current drivers were distributed across three age groups: 29.4% were 55–64, 45.0% were 65–74, and 25.7% were 75+.
Reasons for Limiting or Avoiding Driving
Participants also reported driving limitation for a series of 14 common medical reasons (including a physician's recommendation and a family member's recommendation) and 7 common nonmedical reasons. Driver limitation or avoidance was asked in the following way: "Do you currently limit or avoid any of your driving (either before or after dark) for any of the following reasons?" The participant could respond by indicating "yes," "no," or "don't know" to each of 21 specific reasons (14 medical and 7 nonmedical; see Appendix.)
|
; Haegerstrom-Portnoy, Schneck, Lott, & Brabyn, 2000; Schneck & Haegerstrom-Portnoy, 2003). From a list of 30 separate medical conditions, participants reported whether they had ever received a diagnosis. Vision-related diseases, cataracts, glaucoma, diabetic retinopathy, and macular degeneration were selected for this analysis. Additionally, participants indicated the presence of symptoms or problems within a 30-day period. Vision-related symptoms included watery eyes, dry eyes, and runny or itchy eyes. Vision-related problems included focusing, recognizing at distance, seeing up or down stair steps, seeing through glare from the sun or lights, reading street signs at night, experiencing constricted peripheral vision, and judging distances.
Analysis
For each of the 16 eyesight-related conditions (diseases, symptoms, problems, and objective measures), the impact on reported driving limitations due to problems with eyesight was determined by calculating a risk ratio of (a) the percent reporting a limitation when the condition was present to (b) the percent reporting a limitation when the condition was absent. The significance level of this association was calculated using a chi-square statistic.
Prevalence of the condition was then combined with the risk ratio to generate an attributable risk for that condition, that is, the impact in the population. The population attributable risk (PAR) is the proportion of the outcome (in this case, driving limitations due to vision) that can be attributable to the exposure (in this case, specific vision-related disease, condition, or symptom). This can be calculated as a percentage (PAR%) as follows: PAR% = (Ppop – Punexp)/Ppop x 100. Ppop is the percentage of those in the total population reporting limitations in driving because of vision, and Punexp is the percentage of those in the "unexposed" group (i.e., those without the specific vision-related disease, condition, or symptom) who report limitations in driving because of vision. As an example, the overall percentage of those reporting limitations in driving due to vision (Ppop) is 25.4. The percentage of those reporting limitations in driving due to vision that do not report cataracts (Punexp) is 19.6. The formula above yields a PAR% of 22.8 (Last, 2001).
|
RESULTS |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
|
|
The attributable risk for each of the vision-related conditions is a function of the prevalence of the condition as well as the ratio. Glasses/contacts required for driving has the highest attributable risk (35.8%), having a moderately high ratio of association (2.0) but a very high prevalence (57.9%).
|
DISCUSSION |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
). In this report, we have shown that specific conditions related to vision (disease processes affecting vision, objective vision tests, reported troubles with vision, and physical symptoms affecting the eye) are each associated, to varying degrees, with reported limitations in driving due to vision. We have calculated the impact of each of the factors in two different ways: (a) as an association of each of the vision-related conditions to reported limitations due to eyesight and (b) as an attributable risk. The ratio of percent reporting limitations among those with and without a specific condition related to vision is relevant to evaluating individuals with that condition; the attributable risk is relevant to planning countermeasures in populations (Rose, 1985). We will discuss the relevance of both measures for each of the types of vision-related conditions.
Specific Implications of the Findings
For the set of disease processes affecting vision, the ratios of the association with vision-related limitations in driving were fairly high (1.9 and above). For individuals with these diseases (cataracts, glaucoma, and macular degeneration), the impact on driving is an important consideration, and ways to mitigate this impact should be developed. For example, there is evidence that cataract surgery reduces crash risk associated with cataracts (Owsley et al., 1999). Although the degree of association is similar for the three diseases, cataracts had the highest impact in terms of attributable risk, as the prevalence of that condition is substantially higher than that for the other two diseases. Because so few participants reported having diabetic retinopathy, it was not included in this analysis.
Standardized measures of visual capacity recorded in this study show a different pattern. The ratio of association for the SKILL Test is very low (1.2), showing that this measure of visual acuity under light and dark conditions is not very relevant for vision-related limitations of driving. Other standardized measures of vision, such as the Useful Field of Vision, although not assessed in the current study, may be more appropriate for assessing visual capacity for driving (Owsley et al., 2001). Glasses/contacts required for driving had a relatively high ratio of association (2.0). One possibility is that needing glasses/contacts is a marker for other vision-related problems. Another possibility is that some proportion of glasses/contacts are not well prescribed or well suited for the complex and varied visual tasks associated with driving. It may be that glasses/contacts could be designed to reduce vision-related driving limitations. This might be an effective strategy from not only an individual perspective but also a population viewpoint, as the prevalence and subsequently the attributable risk are high (57.9% and 35.8%, respectively).
The eight reported troubles with vision all had a high ratio of association with vision-related limitations in driving (ranging from 2.2 to 3.4). Most of these have specific implications with respect to driving. For example, problems with focusing, depth, glare, peripheral vision, and nighttime vision might be correctable. These same problems might be addressed by highway design through modification of signs, signals, highway markings, and lighting. The Federal Highway Administration has published a Highway Design Handbook (Staplin, Lococo, Byington, & Harkey, 2001) that has recommendations addressing some of these issues. Work is needed to extend and implement these recommendations. A consideration of prevalence and subsequent attributable risk of these problems might be used to judge the relative benefits of addressing particular troubles with vision. For example, trouble with glare from sun/lights and trouble reading street signs at night have the highest attributable risk and could be addressed by changes in signs and lighting.
Physical symptoms affecting the eyes (i.e., watery eyes, dry eyes, and runny/itchy eyes) showed very low ratios of association with vision-related limitations in driving; and even though the prevalence of these symptoms was relatively high, the attributable risks were also fairly low.
General Relevance of Vision to Driving
In general, findings support previous reports that difficulties with vision are highly associated with increased age and modified or avoidant driving behavior in older adults. However, unlike other studies, this study examined specific types of symptoms and difficulties with vision in association with vision-related driving limitations.
The significance of vision to driving behavior is not surprising as it has been estimated that 90% of driver information is acquired through the visual system (Bailey & Sheedy, 1988). Early vision (here defined to include the ocular media and the retinal transduction and processing) plays a significant role in driving-related tasks as it interacts with higher cognitive functions in crucial skills such as spatial judgment, timely initiation of motor responses, and attention.
Older adults are at risk for a variety of decreased functioning within the eye. With age, the eye changes in several ways, leading to several effects. The lens loses elasticity, decreasing the ability to focus, and it yellows, affecting light sensitivity and acuity. Older adults are also more prone to difficulty with "light scatter" within the eye due to either cataracts, corneal opacity, or vitreous defects, which cause problems with glare (Babizhayev, 2003). The pupil of the eye becomes smaller with advanced age, losing ability to allow light in with a consequent reduction of both visual acuity and sensitivity. Also, many retinal disorders that affect acuity and peripheral vision are associated with increased age (Bailey & Sheedy, 1988).
Although there are a number of independent dimensions of visual capacity, problems with glare and night vision are the two symptoms most likely to be associated with an older person's decision to reduce or avoid driving (Theeuwes, Alferdinck, & Perel, 2002). It is possible to enumerate a list of age-related visual conditions, defects, and diseases associated with either glare or night vision deficiencies (Babizhayev, 2003). Some of these are relatively rare, whereas others are more frequently encountered.
To confirm these results, future studies are needed to interview older drivers and former drivers about the sequence of events that led to a decision to reduce or avoid driving and, in particular, the extent to which problems with glare and night vision contributed to that decision (Dellinger et al., 2001; Theeuwes et al., 2002; West et al., 2003). It will be important to assess the independent and joint effects of a variety of factors that may operate in conjunction with vision, especially types and levels of cognitive function, to affect driving decision making and performance. It may be that levels of self-efficacy or confidence in the ability to drive represent a possible pathway between visual capacity, health status, and driving behavior. There is a growing body of research that suggests that self-efficacy is strongly associated with daily functioning in a variety of areas, including driving (Dobbs, 2003; Seeman, Unger, McAvay, & Mendes de Leon, 1999). Although self-efficacy was not assessed in the current study, future research in this area will contribute to a better understanding of the effects of vision problems on self-efficacy associated with driving in this population.
If these results are confirmed in such studies, clinical and public health strategies to prevent or reduce the effects of these problems should be considered. There is a positive societal benefit that results from retaining the mobility of individuals of advanced age, provided that this can be done without compromising their safety or the safety of others. Strategies range from those taken at a societal level (e.g., headlight regulations that lessen glare to the oncoming driver, signage character size and illumination at night, and licensure requirements) to those undertaken by individuals (e.g., decision to undergo cataract surgery, decision to forgo refractive surgery, scrupulous maintenance of vehicle windscreen clarity). Beyond such strategies, there are varieties of promising infrastructure improvements under study that can become sensory extensions for drivers. These include in-vehicle machine vision capabilities, smart signals and signs, and infrastructure communications to vehicles and their drivers that complement the sensory capabilities of drivers. Finally, as noted previously, it will be important to determine whether changes in self-efficacy represent a possible pathway between these interventions and the timing, location, and quality of driving performance in older populations.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received for publication February 26, 2003. Accepted for publication April 28, 2004.
|
References |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
| ||||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
