This preview shows page 1. Sign up to view the full content.
Unformatted text preview: Is Vision Function Related to Physical Functional Ability in Older Adults?
Catherine G. West, MD,* Ginny Gildengorin, PhD,* Gunilla Haegerstrom-Portnoy, OD, PhD,† Marilyn E. Schneck, PhD,* Lori Lott, PhD,* and John A. Brabyn, PhD* OBJECTIVES: To assess the relationship between a broad range of vision functions and measures of physical performance in older adults. DESIGN: Cross-sectional study. SETTING: Population-based cohort of community-dwelling older adults, subset of an on-going longitudinal study. PARTICIPANTS: Seven hundred eighty-two adults aged 55 and older (65% of living eligible subjects) had subjective health measures and objective physical performance evaluated in 1989/91 and again in 1993/95 and a battery of vision functions tested in 1993/95. MEASUREMENTS: Comprehensive battery of vision tests (visual acuity, contrast sensitivity, effects of illumination level, contrast and glare on acuity, visual fields with and without attentional load, color vision, temporal sensitivity, and the impact of dimming light on walking ability) and physical function measures (self-reported mobility limitations and observed measures of walking, rising from a chair and tandem balance). RESULTS: The failure rate for all vision functions and physical performance measures increased exponentially with age. Standard high-contrast visual acuity and standard visual fields showed the lowest failure rates. Nonstandard vision tests showed much higher failure rates. Poor performance on many individual vision functions was significantly associated with particular individual measures of physical performance. Using constructed combination vision variables, significant associations were found between spatial vision, field integrity, binocularity and/or adaptation, and each of the functional outcomes. CONCLUSIONS: Vision functions other than standard visual acuity may affect day-to-day functioning of older adults. Additional studies of these other aspects of vision and how they can be treated or rehabilitated are needed to determine whether these aspects play a role in strategies for reducing disability in older adults. J Am Geriatr Soc 50:136–145, 2002. Key words: vision function; physical performance; mobility From the *Smith Kettlewell Eye Research Institute, San Francisco, California; and †School of Optometry, University of California, Berkeley, California. Supported by the Smith Kettlewell Eye Research Institute and the Buck Center for Research in Aging. John Brabyn was supported by National Eye Institute Grant EY09588. Address correspondence to Catherine G. West, MD, Smith Kettlewell Eye Research Institute, 2318 Fillmore Street, San Francisco, CA 94115. dequate vision function has long been recognized as an important factor for independence in older adults. Cross-sectional studies have demonstrated that visual impairment (defined as reduced visual acuity), either selfreported or performance-based, is a correlate of physical disability and falls.1–6 In longitudinal analyses, poor visual acuity was also associated with the later development of self-reported limitations in both mobility and activities of daily living (ADLs).2 Distance visual acuity is commonly tested in clinical practice using wall charts with black letters on a white background under high light levels. Because these types of charts are known to participants and easy to administer, they have frequently been used to define visual impairment in epidemiological studies of older adults.2 However, the types of visual impairment that may compromise daily function under the frequently encountered nonoptimal conditions of reduced lighting and contrast may be underestimated by standard acuity measures. Several recent population-based studies in older adults have begun to address this issue. 7,8 These studies are confirming that vision functions such as contrast sensitivity, depth perception, visual field integrity, and vision in glare show important declines with age. However, assessment of these nonstandard aspects of vision function in relation to self-reported disability and objective tests of physical function have not been previously reported. In this cross-sectional analysis, the relationships between a broad range of vision functions and measures of self-reported and objective physical function were explored in a subset of a population-based cohort of older community-dwelling people. We also explored the independent contribution of vision functions other than visual acuity in A JAGS 50:136–145, 2002 © 2002 by the American Geriatrics Society 0002-8614/02/$15.00 JAGS JANUARY 2002–VOL. 50, NO. 1 VISION AND FUNCTION IN OLDER ADULTS 137 relation to these measures of physical ability. We hypothesized that poor performance on individual vision function tests and low values for combined vision measures would be associated with higher levels of self-reported disability and impaired physical performance. METHODS Study Subjects Subjects for the Smith-Kettlewell Eye Research Institute Vision Study were randomly selected from participants in the Buck Center for Research in Aging’s (BCRA) populationbased longitudinal study of health and function (H&F) among older adults in Marin County, California (adjacent to San Francisco). Reed et al. have described in detail the ascertainment of subjects and data collection methods for that study.9 In brief, potential subjects for the H&F cohort (noninstitutionalized residents aged 55 and older) were identified through Health Care Financing Administration lists of Medicare-eligible residents and through random digit dialing. The final age-stratified sample of 2,025 subjects (70% of eligible subjects) completed an extensive inhome evaluation comprising sociodemographics; medical diagnoses and symptoms; self-reported sensory and physical abilities; measures of cognition and depression; and tests of actual performance of physical tasks such as walking, balance, and manual dexterity. In addition to the baseline evaluation conducted in 1989/91 a follow-up evaluation was conducted in 1993/95 that included most of the items on the baseline evaluation, additional physical measurements, and a dietary assessment. A second follow-up has recently been completed. Potential subjects for the Vision Study were identified from the baseline H&F cohort; all participants aged 75 and older and a random sample of 70% of subjects under age 75 were eligible. Because the Vision Study started at the time of the H&F follow-up and not at the baseline evaluation, a number of subjects in the original random sample were not available for recruitment because of death or loss to follow-up. Subjects who did not complete the H&F follow-up were not eligible for the Vision Study. The majority of Vision Study participants (96.3%) were evaluated at the BCRA. Participants unable or unwilling to come in for assessment were evaluated at home. This research followed the tenets of the Declaration of Helsinki. Informed consent was obtained from the subjects after explanation of the nature and possible consequences of the study. The local institutional review board approved the research. Data Collection Vision Function Subjects completed a battery of vision tests and a visionrelated questionnaire. A detailed description of the vision test methodology and results has been presented elsewhere.8 The battery was designed to assess a range of vision functions relevant to performance under commonly encountered conditions of lighting and contrast. A brief practical description of the tests and pass-fail criteria are shown in Table 1. All testing, except the Amsler grid, was binocular with habitual correction for near or distance vision. Re- placement values indicating the poorest possible performance were used for subjects with severe vision impairment when individual tests could not be completed. The criterion for failure for standard high-contrast acuity was 20/70 or worse (sometimes expressed as 20/60), which is a common definition of visual impairment.2,10 This represents detail vision that is reduced 3.5 times from the normal 20/20. The failure criteria for the other vision tests were also based on a reduction of function by a factor of 3.5 to 4.25 times from what is expected in a young healthy subject. Self-Reported Disability and Physical Performance Measures At both the baseline and follow-up H&F evaluations, information on self-reported disability was obtained, and tests of physical performance were undertaken. For this study, only the disability and physical performance measurements from the follow-up H&F evaluation were used as outcome measures, because those data were obtained concurrently with the vision testing data. For self-reported disability, subjects were asked two questions: “Are you able to walk up and down stairs without help?” and “Are you able to walk a half-mile without help?” Subjects were classified as having mobility limitations if they answered “no” to both questions. Three specific activities previously assessed by other investigators as physical performance measures in older adults2 were analyzed: walking, chair stand, and tandem stand. The Buck Center walking test consisted of walking back and forth as fast as possible along a 10-foot marker on the floor. Failure on the walking test was nine or fewer 10-foot lengths completed in one minute. Failure on the chair stand was inability to rise from a straight back chair with arms crossed over the chest at least four times in one minute. Failure on the tandem stand was inability to maintain a full tandem stand (heel of one foot to toe of the other foot) for at least 10 seconds. Self-Reported Health Measures, Cognitive Function, and Depression For diagnoses of stroke, diabetes mellitus, arthritis, myocardial infarction, or hypertension at the baseline evaluation, subjects were asked, “Have you ever been told by a doctor that you have ?” At the follow-up, the question was “Since last interviewed, have you been told by a doctor that you have ?” Subjects reporting the diagnosis at either baseline or follow-up were classified as having the diagnosis. In the present study, measures of cognitive function and depressive symptoms are presented for descriptive purposes. The Short Portable Mental Status Questionnaire, a nine-item screen for dementia, was used to indicate cognitive impairment at the follow-up evaluation. 11 Four or fewer errors (corrected for education) indicate intact cognition or a mild level of impairment; 99.6% of study subjects had scores in this range. To assess level of depressive symptoms, subjects completed the Center for Epidemiologic Studies—Depression Scale, a 20-item scale that has been used in original or modified form in previous survey studies of older adults.12–14 The scores range from 0 (no depressive symptoms) to 60 (pervasive depressive symp- 138 WEST ET AL. JANUARY 2002–VOL. 50, NO. 1 JAGS Table 1. Description of Vision Tests and Pass-Fail Criteria
Test High-contrast acuity (Bailey-Lovie chart) Low-contrast acuity (Bailey-Lovie chart) SKILL Card light-side acuity Description Ability to see fine detail at a distance with good lighting and high contrast Ability to see fine detail at a distance with good lighting and low contrast Ability to see fine detail close up with good lighting and high contrast (e.g., reading a newspaper with good lighting) Ability to see fine detail close up with poor lighting and low contrast (e.g., reading low-contrast menus in dimly lit restaurants) Ability to see fine detail with low contrast in the presence of a surrounding glare source (e.g., reading signs against setting sun) Ability to see large objects with good lighting and low contrast (e.g., detecting a curb in daylight) Measure of the type of depth perception that requires accurate coordination between the two eyes (important for near tasks) Integrity of the central visual field (important in assessment of macular retinal disease) Ability to discriminate and place colored caps in color order Recovery of acuity on the SKILL card after 1-minute exposure to a glare source (e.g., recovery from the glare of passing headlights) Measure of the size of the visual field for suprathreshold targets along 5 meridians with and without a distracting central task Ability to detect slow and fast flickering targets (thought to be related to perception of moving objects) Difference in time to walk 20 feet in lighted room and immediately after room lights are turned off (adaptation to darkness) Criterion for Failure Snellen acuity 20/70 or worse Snellen acuity 20/115 or worse Snellen acuity 20/70 or worse Snellen acuity 20/145 or worse Snellen acuity 20/145 or worse Log contrast sensitivity 1.25 or worse Inability to detect a disparity of 170” More than one anomaly in either eye (blurring, gap, ordistortion) Confusion score of 100 or worse 60 seconds or longer SKILL Card dark-side acuity Acuity in glare (Berkeley glare test) Contrast sensitivity (Pelli-Robson chart) Stereopsis (Frisby test) Amsler grid Color discrimination (Farnsworth D-15) Glare recovery Visual field (standard, attentional) Flicker sensitivity and resolution Impact of lighting on walking
SKILL Horizontal diameter less than 40 Continuous variables Continuous variable Smith-Kettlewell Institute Low Luminance. toms). A score of 16 or above is generally accepted as indicating a high level of depressive symptoms. This criterion has been used in previous analyses of the BCRA cohort15 and was used to classify participants in the present study. jects also had their vision tested using the Smith-Kettlewell Institute Low Luminance (SKILL) card.17 Data Analysis Data were analyzed using SAS 6.12. Descriptive statistics were computed for all of the demographic, health and functioning, and vision variables, including means and standard deviations for continuous data and frequency distributions for each of the categorical variables. Outcome statistics are presented as odds ratios with 95% confidence intervals. For comparisons between participants and each category of nonparticipant, means and proportions were age-adjusted according to the appropriate age distributions in the 1990 U.S. Census. Differences between the groups were compared using parametric tests (analysis of variance models) for continuous variables and nonparametric tests (chi-square) for dichotomous variables. If significant differences were found, Dunnett’s method of multiple comparisons was applied to determine which groups differed from the participants. Comparison Between Vision Study Participants and Nonparticipants For these comparisons, several additional measures were used. Mean annual household income and mean years of education were ascertained at baseline only. Cognition was assessed at baseline using the East Boston Memory Test, a short story with six key features that are scored for both immediate and delayed recall.16 The total memory score consisting of the sum of immediate and delayed recall scores ranged from 0 to 12 features recalled. The number of specific medical conditions (hypertension, myocardial infarction, stroke, diabetes mellitus, arthritis, or cataract) was summed at both baseline (any diagnoses to date) and follow-up (diagnoses since baseline only). As part of the physical assessment portion of the H&F follow-up, sub- JAGS JANUARY 2002–VOL. 50, NO. 1 VISION AND FUNCTION IN OLDER ADULTS 139 The primary analyses were to assess the associations between the physical function variables and vision function after adjusting for age, sex, and other demographic confounders. Initial analyses were based on frequency tables of the outcomes and potential covariates using chi-square, Student’s t-test, and logistic regression models to gauge the degree of association. Age-adjusted odds ratios were calculated by means of a logistic regression model. In an attempt to separate out key aspects of vision function as measured by the individual vision tests, a factor analysis was performed. Factors thus identified were linear combinations of the individual test scores. Summary vision variables were then constructed using both clinical judgment and the results from the factor analysis. These variables were then used in subsequent multivariate analyses, reducing the number of measures used in the multivariate models. The four constructed vision variables used in the multivariate analyses and their component tests scores were: (1) spatial vision (average of high contrast acuity, contrast sensitivity, SKILL dark acuity, and acuity in glare—expressed in log units, with the sign of contrast sensitivity opposite the acuities; range 0.42–1.39, mean 0.08); (2) binocularity (sum of Frisby stereo category (0–2) and Amsler grid category (0–3)—expressed as integer values between 0 and 5, median 0); (3) field integrity (average of maximum horizontal diameters for standard and attentional visual fields and errors in standard and attentional fields—expressed as a percentage of maximum horizontal diameter (% of 140 ) or of total possible errors (% of 39 locations missed) with the sign reversed and the range shifted such that 100 is perfect performance; range is 100% to 0%, mean 69.2%); and (4) adaptation (sum of glare recovery and difference between walking score in a light and dark room—expressed in log seconds; range 0.869–4.25, mean 2.41). Results of the exploratory analyses were used to make informed decisions about which variables to include in the final analyses as potential covariates. The major hypotheses were evaluated using multivariate logistic regression models with binary outcomes. Stepwise regression techniques were applied to assist in the selection of variables for the final multivariate model. Because diabetes mellitus and stroke can be associated with both poor vision function and poor physical performance, and are thus potential confounders of the vision/ physical performance association, an a priori decision was made to include these diseases in all study analyses. To explore their role as potential confounders, vision function in subjects with self-reported diabetes mellitus and with self-reported stroke was assessed as part of the present study. Other conditions that were less likely to be associated with both poor physical and vision performance (arthritis, myocardial infarction, and hypertension) were also assessed. These conditions were tested in the logistic models to determine whether their inclusion changed the estimate of the relationship between vision and physical function. Because the odds ratio estimates and level of statistical significance did not change materially when these other potential confounders were included, the final models adjusted for age, sex, stroke, and diabetes mellitus only. Main effects were included in each model, adjusting for factors mentioned above. Interaction terms were inves- tigated whenever indicated. Subjects missing any values for the H&F measures were excluded from the multivariate analyses. The .05 level of significance was used for all statistical tests. RESULTS Comparison Between Vision Study Participants and Nonparticipants One thousand seven hundred forty-seven individuals from the original 2,025 H&F subjects were randomly selected post hoc in 1993 as potential participants in the Vision Study. Of the individuals in the random sample, 1,315 subjects finished the H&F follow-up study and were therefore eligible for the Vision Study. The remaining 432 subjects in the random sample were not eligible to participate in the Vision Study because of death before the H&F followup (n 292), refusal to participate in the H&F follow-up (n 82), or being unavailable for the H&F follow-up (lost, in skilled nursing facility, moved; n 58). Compared with those who completed the Vision Study, the subjects who died before the H&F follow-up were older, had lower annual incomes and fewer years of education, had more medical conditions, and were more likely to have difficulty walking than Vision Study participants (all P .05). Those who refused or were unavailable for the H&F follow-up were significantly more likely to be female than Vision Study participants and tended to be older, have fewer years of education, have lower annual incomes, and have lower memory scores. They were similar to participants in the number of reported medical conditions and performance on the walking test. Seven hundred eighty-two (59%) of the 1,315 eligible subjects completed the Vision Study. The characteristics of those who completed the Vision Study are shown in Table 2. Of eligible subjects who did not complete the Vision Study, 8.1% died before the study, 10.3% refused to participate, and 22.1% were unavailable or lost or had moved. Thus, 65% of living eligible subjects completed the Vision Study. Subjects in each of the three nonparticipant categories (died, refused, unavailable) differed significantly from Table 2. Characteristics of Participants in Vision Study
Characteristic Age in years at vision study Annual income (H&F baseline), thousand $ Years of education Medical conditions (H&F baseline), n New medical conditions (follow-up), n Memory score (H&F baseline) SKILL light-side acuity (log MAR) Standard high-contrast acuity SKILL dark-side acuity (log MAR) Low-contrast low-luminance acuity Percentage passing Buck walk test (baseline) Percentage passing Buck walk test (follow-up) Percentage female Mean SD 76.1 9.3 50.5 32.0 14.6 3.1 1.1 1.0 0.8 0.9 9.1 2.8 0.22 0.23 Snellen 20/33 0.86 0.43 Snellen 20/146 87 89 54 SD standard deviation; H&F Health and functioning; SKILL Smith-Kettlewell Institute Low Luminance; MAR minimum angle of resolution. 140 WEST ET AL. JANUARY 2002–VOL. 50, NO. 1 JAGS those who completed the study, being older and more likely to be female and having fewer years of education, more medical conditions, lower memory scores, and worse performance on the walking test both at baseline and follow-up. All groups of nonparticipants also had worse visual acuity on both the light and the dark side of the SKILL card than did participants. Their acuity under conditions of low contrast and low illumination differed from that of the participants by factors of 1.3 to 3. The acuities of the nonparticipants under these conditions of reduced contrast and lighting were worse than the legal blindness definition of 20/200. Vision Battery Performance A small percentage of those in the youngest age group (aged 55–64) failed any of the individual vision tests, but the percentage of failure increased significantly with age group for all vision measures. (See Table 3; the failure criteria are listed in Table 1.) Distance and near visual acuity assessed under optimal lighting and contrast conditions (high-contrast distance acuity and SKILL card–light acuity) were the least impaired with advancing age compared with the other vision functions; only 20% of those aged 85 and older failed. Reducing the contrast (low-contrast visual acuity) caused 40% of the oldest group to fail. Reducing the contrast and light level (SKILL card-dark acuity) or adding surrounding glare (acuity in glare) increased the failure rates in those aged 85 and older to 72% and 87%, respectively. A high percentage in the oldest age group also failed tests of stereopsis and glare recovery (60% and 65%, respectively). Very few people in the oldest group showed failures on the standard visual field test (4%). When an attentional task was added to the field test, the failure rate increased by a factor of nine, to 36% of the oldest group. The increase in failure rate with age is exponential for all measures, but it is clear from Table 3 that the rate of change varies dramatically between measures. The commonly used clinical measures of high-contrast acuity and standard visual fields revealed the smallest changes and are thus least sensitive to aging effects. Health and Function Measures The percentage in different age groups with selected medical conditions and impaired physical performance is shown in Table 4. The presence of stroke, cognitive impairment, selfreported mobility limitation, and poor performance on the physical performance tests were all significantly associated with increasing age, but diabetes mellitus and high level of depressive symptoms were not. Of participants under age 75, the percentage who failed the walking test, the chair stand, or the tandem stand was low, but the failure rate on these tests increased exponentially for the oldest age group. Even so, two-thirds of those aged 85 and older were able to perform these tasks. Although data on most health conditions of interest were available for over 90% of participants, the percentage with available data was slightly lower for the walking (88%) and tandem stand (85%) tests. Vision Function in People with Stroke and Diabetes Mellitus Using the criteria in Table 1 and adjusting for age and sex, subjects with a self-reported diagnosis of stroke (n 65) were three times more likely to have failing scores (20/70 or worse) on tests of high-contrast near visual acuity (SKILL card—light acuity) and attentional visual fields ( 40 deg diameter) and twice as likely to fail tests of contrast sensitivity (log CS 1.25), acuity with glare (20/145 or worse), and stereoscopic vision (unable to see 170” of disparity) (data not shown). Using the same failure criteria in similarly adjusted models, those who reported diabetes mellitus (n 44) were roughly three times more likely to have failing scores on tests of SKILL dark acuity (20/145 or worse), contrast sensitivity, and stereoscopic vision. Although the confidence intervals for these odds ratios did not include one, several intervals were quite wide. Relation of Vision Function to Self-Reported Mobility Limitations and Physical Performance The odds of reporting mobility limitations were significantly increased in subjects with poorer performance on Table 3. Percentage of People in Different Age Groups Who Failed Selected Vision Tests
Number with Data for Each Vision Test Vision Test High-contrast visual acuity Low-contrast visual acuity SKILL Card light acuity SKILL Card dark acuity Acuity in glare Contrast sensitivity Stereopsis Amsler grid Color discrimination Glare recovery Standard visual field Attentional visual field
SKILL Smith-Kettlewell Institute Low Luminance. 55–64 Years n 110 65–74 Years n 247 % 75–84 Years n 287 85 Years n 138 782 781 782 782 776 781 769 779 756 759 717 715 1 1 1 4 3 1 8 5 6 10 1 5 2 3 2 8 19 3 13 9 4 17 0 8 3 10 5 32 46 15 27 17 10 34 1 22 20 40 20 72 87 51 60 39 24 65 4 36 JAGS JANUARY 2002–VOL. 50, NO. 1 VISION AND FUNCTION IN OLDER ADULTS 141 Table 4. Percentage in Different Age Groups with Selected Medical Conditions and Impaired Physical Performance
Characteristic Stroke Diabetes Cognitive impairment Depression Self-reported mobility problems Fall in past year Failed Buck walking test Failed chair stand test Failed tandem stand test Number with Data for Characteristic 750 782 775 771 763 720 689 736 661 55–64 Years n 110 3 3 4 7 9 36 1 1 2 65–74 Years n 247 8 7 5 6 9 36 3 4 8 75–84 Years n 287 10 7 11 7 18 44 14 6 19 85 Years n 138 11 3 28 11 48 68 33 24 35 tests of high- and low-contrast acuity, acuity with glare, contrast sensitivity, glare recovery, standard and attentional field diameter and errors, and impact of lighting on walking (Table 5). For every line lost on the acuity charts, the odds of mobility limitations increased by about 10%. For every change in visual field performance of 10%, the odds of mobility problems rose by about 20%. For every second of difference in the time between walking in the light versus the dark, the odds of mobility problems increased by about 20%. Significant associations were found between failure on the Buck Center walking test and poor performance on tests of high- and low-contrast visual acuity, SKILL light and dark acuity, acuity in glare, contrast sensitivity, standard and attentional visual field performance, and flicker resolution and sensitivity (Table 5). For example, for every 0.1 log unit loss of high frequency flicker resolution, there was a nearly 60% increase in the likelihood of failing the Buck Center walking test. Poorer performance on the test of impact of lighting on walking was not associated with failing the Buck Center walking test. However, this test is a difference score of the time taken to walk in the light and the dark and is a measure of dark adaptation that was intended to be independent of physical limitation. A similar pattern was seen for performance on tests of acuity, contrast sensitivity, and visual fields, but not flicker, in relation to failing the chair stand test. Instead, stereoscopic vision test performance and the impact of lighting on walking were significantly associated with failing the chair stand test. For every doubling of the stereo threshold (0.3 log unit), there was a 50% increase in the likelihood of failing the chair stand. For every second delay in dark adaptation (difference in walking in a lighted and a dark room), there was a roughly 20% increase in the likelihood of failing the chair stand. For the tandem stand test (a test of balance), a different pattern was found. Subjects with poorer performance on tests of contrast sensitivity, stereoscopic vision, and attentional field performance were more likely to fail the tandem stand. Tests of various types of acuity were not significantly associated with failure on the tandem stand. Measures of Physical Function in Relation to Constructed Functional Vision Variables The four different constructed vision variables described in the methods section—spatial vision, binocularity, field integrity, and adaptation—were examined in relation to measures of physical function while adjusting for age, sex, stroke, and diabetes mellitus. Each constructed vision variable was examined individually in a logistic model to assess the impact of the key aspect of vision function it represented. All four constructed variables were then examined jointly in an attempt to isolate which of these aspects of vision function was playing the most important role in the relationship with the physical outcome measure. The results of these analyses are shown in Table 6. Spatial vision, field integrity, and adaptation were all significantly associated with self-reported mobility limitations when each was the only vision variable included in the model. When all four constructed vision variables were included in the model, significant associations remained between mobility limitations and both field integrity and adaptation; spatial vision was no longer a significant correlate. Worse spatial vision and compromised field integrity were significantly associated with failure on the Buck Center walking test when the constructed variables were considered in individual logistic models (Table 6). When all four constructed vision variables were included in the model, only field integrity remained a significant correlate of failing the Buck Center walking test. Failing the chair stand was significantly associated with spatial vision, binocularity, and field integrity when each of these constructed variables was considered individually. Compromised binocularity remained a significant correlate of failing the chair stand test when all four variables were included in the model. For every categorical change in binocularity, the likelihood of failing the chair stand increased by about 50%. Worse binocularity, compromised field integrity, and poor adaptation were all significantly associated with failing the tandem stand when each constructed vision variable was considered separately. Only field integrity remained significant when all of the constructed vision variables were included in the model. For each loss of 10% of field integrity, the likelihood of failing the tandem test increased by about 20%. DISCUSSION This study was an exploration of the association between vision function and measures of mobility limitations and 142 WEST ET AL. Table 5. Odds of Self-Reported Mobility Problems and Odds of Failing Tests of Physical Function in Relationship to Performance on Vision Tests (Adjusted for Age, Sex, Diabetes Mellitus, and Stroke)
Self-Reported Mobility Problems n 735 734 735 735 729 734 722 733 716 684 673 673 672 672 720 720 687 1.11 1.11 1.08 1.06 1.08 1.10 1.18 1.19 1.24 1.07 1.26 1.34 1.11 1.13 1.30 1.07 1.18 1.02–1.21 1.04–1.19 1.00–1.16 0.99–1.13 1.02–1.14 1.04–1.17 0.99–1.42 0.67–2.12 0.90–1.69 1.02–1.13 1.05–1.52 1.10–1.62 1.04–1.18 1.04–1.22 0.96–1.77 0.97–1.18 1.07–1.31 665 664 665 685 662 664 653 663 650 626 624 624 623 623 650 651 636 1.19 1.19 1.18 1.11 1.11 1.18 1.24 1.44 1.28 1.02 1.41 1.47 1.22 1.26 1.56 1.16 1.08 1.07–1.33 1.09–1.30 1.07–1.30 1.02–1.20 1.04–1.20 1.09–1.28 0.99–1.56 0.69–2.99 0.85–1.92 0.96–1.09 1.11–1.78 1.14–1.89 1.12–1.33 1.14–1.40 1.02–1.20 1.01–1.33 0.97–1.21 731 708 709 709 705 708 697 708 691 666 654 654 653 653 694 694 668 1.14 1.13 1.13 1.11 1.10 1.12 1.50 1.58 1.24 1.03 1.31 1.33 1.18 1.22 1.48 1.08 1.18 Odds 95% CI n Odds 95% CI n Odds Failed Walking Test Failed Chair Stand 95% CI 1.05–1.25 1.04–1.23 1.03–1.23 1.03–1.20 1.02–1.18 1.04–1.20 1.16–1.94 0.79–3.15 0.78–1.96 0.96–1.12 1.08–1.59 1.08–1.64 1.06–1.30 1.01–1.03 0.96–1.99 0.93–1.24 1.05–1.33 n 637 636 637 637 636 636 626 636 628 606 601 601 600 600 626 627 616 Failed Tandem Stand Odds 1.08 1.06 1.03 1.07 1.03 1.08 1.33 2.00 1.11 1.05 1.15 1.20 1.10 1.14 1.09 1.11 1.12 95% CI 0.98–1.20 0.97–1.15 0.94–1.14 0.99–1.16 0.97–1.11 1.00–1.16 1.07–1.62 1.00–4.00 0.77–1.59 0.99–1.11 0.89–1.50 0.90–1.05 1.03–1.19 1.05–1.50 0.75–1.58 0.99–1.21 0.99–1.26 Vision Test Units High-contrast visual acuity Low-contrast visual acuity SKILL Card light acuity SKILL Card dark acuity Acuity in glare Contrast sensitivity Stereopsis Amsler grid Color discrimination Glare recovery Standard field diameter Standard field errors Attentional diameter Attentional field errors Log flicker resolution Log flicker sensitivity Impact of lighting on walking 0.1 log unit 0.1 log unit 0.1 log unit 0.1 log unit 0.1 log unit 0.1 log unit 0.3 log unit category category 0.1 log unit 10% 10% 10% 10% 0.1 log unit 0.1 log unit 1 second JANUARY 2002–VOL. 50, NO. 1 CI confidence interval; SKILL Smith-Kettlewell Institute Low Luminance. JAGS JAGS JANUARY 2002–VOL. 50, NO. 1 VISION AND FUNCTION IN OLDER ADULTS 143 Table 6. Multivariate Logistic Models of Self-Reported Mobility Limitations and Physical Test Performance in Relation to Constructed Vision Variables
Mobility Limitations Constructed Vision Variable Age, sex, stroke, and diabetes mellitus plus one constructed variable Spatial vision Binocularity Field integrity Adaptation Age, sex, stroke and diabetes mellitus plus all constructed variables Spatial vision Binocularity Field integrity Adaptation
CI confidence interval. Failed Walking Test n Odds 95% CI Failed Chair Stand n Odds 95% CI Failed Tandem Test n Odds 95% CI Units n Odds 95% CI 0.1 log unit categorical 10% 0.1 log unit 728 720 672 648 1.11 1.16 1.25 1.06 1.03–1.19 0.98–1.36 1.10–1.42 1.01–1.12 661 651 623 603 1.19 1.20 1.50 1.03 1.09–1.31 0.98–1.48 1.28–1.78 0.97–1.10 704 696 653 633 1.15 1.42 1.39 1.02 1.05–1.26 1.14–1.77 1.16–1.66 0.94–1.10 635 625 600 588 1.08 1.31 1.23 1.06 0.99–1.18 1.09–1.58 1.07–1.43 1.00–1.12 0.1 log unit categorical 10% 0.1 log unit 612 612 612 612 1.00 1.04 1.21 1.06 0.86–1.16 0.82–1.31 1.05–1.40 1.00–1.13 573 573 573 573 0.99 1.14 1.45 1.02 0.81–1.22 0.84–1.55 1.18–1.74 0.94–1.11 601 601 601 601 0.84 1.54 1.17 1.05 0.65–1.07 1.10–2.15 0.94–1.47 0.96–1.16 561 561 561 561 0.93 1.21 1.20 1.06 0.80–1.09 0.95–1.53 1.02–1.40 1.00–1.13 physical performance, with an emphasis on aspects of vision that are not typically measured in population studies of older adults. As a cross-sectional study, the relationships found cannot be interpreted as causal but do merit further attention in longitudinal studies. The detailed vision function performance of this study population as reported by Haegerstrom-Portnoy et al. 8 was in generally good agreement with results from other large scale studies of vision function in older groups.1,18–22 In the present study, the dichotomized results show the same cross-sectional pattern with age: relatively well-preserved high contrast acuity and standard visual field performance in the face of striking deficits in SKILL card– dark acuity, acuity in glare, stereopsis, glare recovery, and attentional visual field performance. More than 60% (60– 87%) of the oldest subjects failed these nonstandard measures of vision, whereas only 20% of them failed highcontrast visual acuity. Vision Function and Self-Reported Mobility Limitations The significant associations between self-reported mobility limitation and poor performance on tests of acuity and contrast sensitivity in the present study are in agreement with the findings in other population-based studies. Highcontrast distance visual acuity was assessed in three communities as part of the Established Populations for Epidemiologic Studies of the Elderly (EPESE) protocol.21 In that group, Salive et al. found significant associations between poor acuity and both self-reported mobility limitation (same operational definition as the present study) and self-reported ADL limitations. In the Salisbury Eye Evaluation (SEE) project, participants with impairments of binocular acuity were more likely to report difficulty with ADLs and instrumental activities of daily living and a reduced level of social activities than were those without such impairments. 4 In analyses of multiple aspects of vision function in the SEE study, distance acuity and contrast sensitivity, but not acuity in glare or stereopsis, were independently associated with self-reported disability measures.7 In the present study, in contrast, self-reported difficulty with walking or climbing stairs was also significantly associated with poor performance on two tests of glare: acuity in glare and glare recovery, and with another measure that assesses adaptation, the impact of lighting on walking. Differences in the techniques used to assess acuity in glare in the two studies may account for this discrepancy in results.8 The SEE study did not measure glare recovery. As in the SEE project, we found no association between self-reported mobility problems and stereopsis. In addition to these variables, both standard and attentional visual fields were related to mobility limitations. Spatial vision, field integrity, and adaptation were the three constructed vision variables significantly associated with self-reported mobility problems. In the final model, incorporating all of the constructed vision variables, spatial vision dropped out completely. Thus, visual acuity may be less important than other vision measures such as visual fields and adaptation when attempting to predict mobility problems. Vision Function and Objective Physical Performance Significant associations were found between physical performance and a large number of vision functions. Similar tests of physical performance were assessed in relation to visual acuity in analyses of EPESE data.2 Salive et al. found that poor acuity was significantly associated with poor performance on each of the three physical performance tests. In the present study, poorer scores on acuity measures were associated with failure on the walking test and chair stand, both of which involve movement for successful completion, but not on the tandem stand, which is a test of static balance. In the analyses using the constructed vision variables, spatial vision was not a significant correlate of physical performance when other vision functions were included in the model. 144 WEST ET AL. JANUARY 2002–VOL. 50, NO. 1 JAGS The functional importance of visual field integrity was seen for both self-reported and observed physical performance. Tests of standard and attentional visual field size and errors were significantly associated with self-reported mobility limitation, and field integrity remained significant in the model that included all the constructed variables. For the physical performance tests, the field integrity measure was the only functional variable associated with failure on each of the three tasks, and this measure remained significant in the logistic models with all constructed vision variables included for the tandem stand and the Buck Center walking test. Visual field loss has been noted to have important implications for physical functioning, and marked improvements in functional mobility may result with an appropriate rehabilitation programs and prosthetic devices.23,24 The attentional component of visual field loss may require a different approach.25 Potential Confounding Issues In assessing the systemic diseases whose presence may confound the relationship between vision and health outcomes, the patterns observed in this study population are consistent with those reported by others. The prevalence of stroke was similar to, and that of diabetes mellitus lower than, the rates reported in other studies of community-dwelling older adults.9,14 In those with stroke, visual impairment can result from visual field loss including central field loss, misinterpretation of visual information, and cognitive deficits that affect attention.26 In the present study, subjects with a history of stroke had increased odds of failing tests of near visual acuity, acuity with glare, contrast sensitivity, stereopsis, and attentional visual fields. Stroke was not associated with increased odds of failing distance acuity. The only measure available for stroke in the current study was a broadly inclusive dichotomous one that did not take vascular territory, degree of deficit, or timing of the event into account. The major causes of vision loss in diabetes mellitus are accelerated cataract and diabetic retinopathy,26 each of which leads to different patterns of vision function difficulty. Acuity in glare measure is sensitive to the presence of cataract but was not significantly associated with diabetes mellitus in this sample. The odds of failing contrast sensitivity and SKILL dark acuity were 3.3 times higher in people with diabetes mellitus, which is consistent with a retinal problem. As with stroke, the operational classification of diabetes mellitus in this study could not include factors relevant for diabetic visual complications such as level of glycemic control and disease duration. This could result in misclassification and a failure to identify associations between specific vision function impairments and diabetes mellitus. Severe cognitive impairment may also confound the relationship between visual performance and physical function measures. A high degree of impairment could result in poor performance on vision tests independent of actual vision impairment if the test itself was too complex. Many pathological processes that cause dementia may result in impaired mobility. In the SEE project, Rubin et al. chose to exclude individuals with Mini-Mental State Examination (MMSE) scores indicating severe cognitive impairment.1 However, as shown by Reischies et al., visual im- pairment itself might affect scores on tests of cognitive function such as the MMSE that rely extensively on visual items.27 In the present study, the instruments used to screen for cognitive impairment have no visual component. More importantly, the level of moderate or severe cognitive dysfunction was extremely low ( 0.4%) and thus would not explain the observed relationships between vision function and physical performance. Other Methodological Issues Although one of the study’s goals was to explore the contribution of aspects of vision function independent of visual acuity, the importance of high-contrast visual acuity in a variety of functional tasks is evident. Visual acuity is also currently the most commonly used screening measure in older adults.28 In the present analyses, the constructed spatial vision variable was an unweighted combination of high-contrast distance visual acuity, contrast sensitivity, SKILL dark near acuity, and acuity in glare. Each of these measures may be important in relation to a particular functional task, but all are assessed in similar tests based on testing central vision. Because these different measures of acuity are correlated with each other, it may not be possible to separate out the effects of other types of vision function (such as contrast, luminance, or glare effects) from those of visual acuity in relation to measures of health and physical function. The instability of the odds ratio estimates for the effects of the constructed spatial vision variable in the logistic regression models in which multiple vision variables were included suggests there was residual correlation. Thus, we cannot infer that these other vision measures are not significantly related to many of the outcomes of interest. Rather, these results serve to demonstrate that other aspects of vision function may be very relevant in the day-to-day functioning of older adults. Another methodological issue relevant to interpreting the study results is the difference between study participants and nonparticipants. Vision Study participants were generally younger and healthier than nonparticipants. This would introduce a bias in assessing the relationship between vision and the physical function outcomes only if the relationship differed according to participation status. It is unlikely that poor visual function would be associated with poor physical function in participants but with better physical function in nonparticipants. However, although bias is unlikely, the better functional status of study participants does suggest that the prevalence of poor vision and physical function in the study population is an underestimate of the level of impairment in the underlying population of older adults. This is consistent with the comments of Rubin et al. in discussing the poorer visual acuity of subjects in the Beaver Dam Study relative to the SEE findings.1,20 They noted that the higher participation rate in the Beaver Dam might indicate higher participation rates in those with poor vision who typically do not participate in vision studies. SUMMARY This cross-sectional study of older adults found significant associations between poor performance on a number of tests of vision function and both self-reported and objec- JAGS JANUARY 2002–VOL. 50, NO. 1 VISION AND FUNCTION IN OLDER ADULTS 145 tive measures of physical functioning. Prospective analyses that include vision, health, and physical performance data currently being collected in this population will help clarify the nature of the associations found in the present study. Additional studies of impairment of vision functions other than visual acuity and how they can be treated are needed, to determine whether they play a role in strategies for reducing disability in older adults. With regard to the practical applications of the study findings, several of the tests used here could be readily administered as part of a geriatric vision workup. For example, contrast sensitivity and SKILL dark acuity require only a wall chart or a hand-held card. Their results can be expressed in recognizable units and may be useful in detecting functionally relevant types of loss or identify individuals at high risk for ophthalmic disease.29 Other tests, such as the attentional visual field, acuity in glare, and glare recovery, may have particular relevance for assessing the older driver but need further work in the development of measurement units relevant for clinical practice and conversion of the tests themselves to forms that do not require specialized equipment. Nevertheless, it should be noted that all tests in the present study were administered by community volunteers and could thus be administered by trained office personnel. REFERENCES
1. Rubin GS, West SK, Munoz B et al. A comprehensive assessment of visual impairment in a population of older Americans (The SEE Study). Invest Ophthalmol Visu Sci 1997;38:557–568. 2. Salive ME, Guralnik J, Glynn RJ et al. Association of visual impairment with mobility and physical function. J Am Geriatr Soc 1994;42:287–292. 3. Kington R, Rogowski J, Lilliard L. Functional associations of “trouble seeing.” J Gen Intern Med 1997;12:125–128. 4. West SK, Munoz B, Rubin GS et al. Function and visual impairment in a population-based study of older adults. Invest Ophthalmol Vis Sci 1997;38:72–82. 5. Klein BEK, Klein R, Lee KE et al. Performance-based and self-assessed measures of visual function as related to history of falls, hip fractures and measure gait time. The Beaver Dam Eye Study. Ophthalmology 1998;105: 160–164. 6. Ivers RQ, Cummings RG, Mitchell P et al. Visual impairment and falls in older adults: The Blue Mountains Eye Study. J Am Geriatr Soc 1998;46:58–64. 7. Rubin GS, Roche KB, Prasado-Rao P et al. Visual impairment and disability in older adults. Optom Vis Sci 1994;71:750–760. 8. Haegerstrom-Portnoy G, Schneck ME, Brabyn JA. Seeing into old age: Vision function beyond acuity. Opt Vis Sci 1999;76:141–158. 9. Reed D, Satariano WA, Gildengorin G et al. Health and functioning among the elderly of Marin County, California: A glimpse of the future. J Gerontol A Biol Sci Med Sci 1995;50A:M61–M69. 10. Genensky M, Zarit SH. Low-vision care in a clinical setting. In: Rosenbloom AA, Morgan MW, eds. Vision and Aging, 2nd Ed. Boston: Butterworth-Heinemann, 1993, pp 424–444. 11. Pfeiffer E. A short portable mental status questionnaire for the assessment of organic brain deficit in elderly patients. J Am Geriatr Soc 1975;23:433–441. 12. Radloff LS. The CES-D Scale: A self-report depression scale for research in the general population. Appl Psych Meas 1977;1:385–401. 13. Radloff LS, Teri L. Use of the Center for Epidemiologic Studies Depression Scale with older adults. Clin Gerontol 1986;5:119–136. 14. Cornoni-Huntley J, Brock DB, Ostfeld A et al., eds. Established Populations for the Epidemiologic Study of the Elderly, Resource Data Book (NIH publication no. 86–2443). Bethesda, MD: National Institutes of Health, 1986. 15. West C, Reed DM, Gildengorin, GL. Can money buy happiness? Depressive symptoms in an affluent older population. J Am Geriatr Soc 1998;46:49–57. 16. Scherr PA, Albert MS, Funkenstein HH et al. Correlates of cognitive function in an elderly community population. Am J Epidemiol 1988;128:1084–1101. 17. Haegerstrom-Portnoy G, Brabyn JA, Schneck ME et al. The SKILL card. An acuity test of reduced contrast and luminance. Invest Ophthalmol Vis Sci 1997;38:207–218. 18. Kahn HA, Leibowitz HM, Ganley JP et al. The Framingham Eye Study: Outline and major prevalence findings. Am J Epidemiol 1977;106:17–32. 19. Tielsh JM, Sommer A, Witt K et al. Blindness and visual impairment in an American urban population. The Baltimore Eye Survey. Arch Ophthalmol 1990;108:286–290. 20. Klein R, Klein BEK, Linton KLP et al. The Beaver Dam Eye Study: Visual acuity. Ophthalmology 1991;98:1310–1315. 21. Salive ME, Guralnik J, Christen W et al. Functional blindness and visual impairment in older adults from three communities. Ophthalmology 1992;99: 1840–1847. 22. Taylor HR, Livingston PM, Stanislavsky YL, et al. Visual impairment in Australia: Distance visual acuity, near vision and visual field findings of the Melbourne Visual Impairment Project. Am J Ophthalmol 1997;123:328–337. 23. Marron JA, Bailey IL. Visual factors and orientation-mobility performance. Am J Optom Physiol Optics 1982;59:413–426. 24. Maino JH. Visual deficits and mobility. Evaluation and management. Clin Geriatr Med 1996;12:803–823. 25. Ball K, Owsley C. The useful field of view test: A new technique for evaluating age-related declines in visual function. J Am Optom Assoc 1993;64:71–79. 26. Cockburn DM. Ocular implications of systemic disease in the elderly. In: Rosenbloom AA, Morgan MW, eds. Vision and Aging, 2nd Ed. Boston, MA: Butterworth-Heinemann, 1993, pp 29–96. 27. Reischies FM, Geiselmann B. Age-related cognitive decline and vision impairment affecting the detection of dementia syndrome in old age. Br J Psychiatry 1997;171:449–451. 28. Besdine RW. Clinical approach: An overview. In: Cassel CK, Cohen HJ, Larson EB et al., eds. Geriatric Medicine, 3rd Ed. New York, NY: Springer, 1997, pp 155–168. 29. Woods RL, Tregear SJ, Mitchell RA. Screening for ophthalmic disease in older subjects using visual acuity and contrast sensitivity. Ophthalmology 1998;105:2318–2326. ...
View Full Document
This note was uploaded on 02/17/2011 for the course HK 490 taught by Professor Reidtky during the Fall '10 term at Purdue University-West Lafayette.
- Fall '10