- Email: [email protected]
Five-year change in visual acuity and incidence of visual impairment
Five-Year Change in Visual Acuity and Incidence of Visual Impairment The Blue Mountains Eye Study Suriya Foran, MBBS, MPH, Paul Mitchell, MD, PhD, Jie Jin Wang, MMed, PhD Purpose: To describe the 5-year change in visual acuity and the incidence of visual impairment in a population-based cohort. Design: Population-based epidemiologic study. Participants: Of the 3654 participants of the Blue Mountains Eye Study (BMES I) baseline examination (aged 49 years⫹ during 1992–1994), 2335 were reexamined during the 5-year follow-up examinations from 1997 to 1999 (BMES II), and 543 persons had died since BMES I. Methods: Visual acuity was measured using a logarithm of the minimum angle of resolution chart in both eyes separately before and after standardized refraction. Pupils were dilated and a detailed examination was performed. Main Outcome Measures: Visual impairment, after best refractive correction, was defined as any (visual acuity ⱕ20/40; ⱕ41 letters) and severe (visual acuity ⱕ20/200; 0 –5 letters) in keeping with the Beaver Dam Eye Study. Incident binocular visual impairment was defined as visual acuity ⱕ20/40 in both eyes at follow-up, where visual acuity was ⬎20/40 in both eyes at baseline. Incident binocular severe visual impairment was defined as visual acuity ⱕ20/200 in both eyes at follow-up, where visual acuity was ⬎20/200 in both eyes at baseline. The incidence for three other levels of visual impairment is also given: ⬍20/40, ⬍20/70, and ⬍20/200. Monocular visual impairment was defined as impairment in one eye only at follow-up, where both eyes were unimpaired at baseline. Incident doubling and halving of the visual angle were calculated. Results: Incidence rates for visual impairment increased significantly with age. Any incident impairment ⱕ20/40 occurred binocularly in 41 persons (1.9%) and monocularly in 150 persons (7.1%). Severe incident impairment ⱕ20/200 occurred binocularly in 3 persons (0.1%) and monocularly in 44 persons (2.1%). Incident impairment ⬍20/40 occurred binocularly in 37 persons (1.7%) and monocularly in 134 persons (6.3%). Impairment ⬍20/70 occurred binocularly in 15 persons (0.7%) and monocularly in 84 persons (3.8%). Impairment ⬍20/200 occurred binocularly in 3 persons (0.1%) and monocularly in 44 persons (1.9%). Women consistently had a higher incidence of visual impairment than men, although this was often not statistically significant after adjusting for age. Increasing age was a strong predictor of visual impairment. Conclusions: This study has documented the 5-year incidence and causes of visual impairment in an older Australian population. Ophthalmology 2003;110:41–50 © 2003 by the American Academy of Ophthalmology.
Recent population studies have provided important data on the prevalence1–3 and causes of visual impairment4 –12 in older persons, but few population reports describe the change in visual acuity or the incidence of visual impairment over time.13 Other reports show that visual impairment increases significantly with age and is associated with use of community support services,14 nursing home placement,15 falls,16,17 mortality,18,19 and factors determining self-rated Originally received: August 3, 2001. Accepted: April 17, 2002. Manuscript no. 210694. From the Department of Ophthalmology and the Save Sight and Westmead Millennium Institutes, the University of Sydney, Sydney, Australia. Supported by the Australian National Health & Medical Research Council, Canberra, Australia. Reprint requests to Paul Mitchell, MD, PhD, Department of Ophthalmology, University of Sydney, Hawkesbury Rd, Westmead, NSW, Australia, 2145; E-mail: [email protected] © 2003 by the American Academy of Ophthalmology Published by Elsevier Science Inc.
health.20 This report from the Blue Mountains Eye Study (BMES) describes the 5-year change in visual acuity and the incidence and progression of visual impairment in a representative sample of older Australians.
Materials and Methods Population The BMES is a population-based survey of vision and common eye diseases in an urban population aged 49 years or older residing in two postal codes of the Blue Mountains region, west of Sydney, Australia. A previous report from the study explained the reasons for selecting, and the methods used to identify, the target population and described the population.21 To summarize, study personnel conducted a door-to-door census from November to December 1991 for the first postcode area and from March to April 1993 for the second postcode area. All noninISSN 0161-6420/03/$–see front matter PII S0161-6420(02)01295-2
41
Ophthalmology Volume 110, Number 1, January 2003 Table 1. Baseline Characteristics of Participants and Nonparticipants in the Blue Mountains Eye Study: 5-Year Examinations Nonparticipants Participants
Alive
Characteristics
Crude %
n
Crude %
n
Total (BMES I, n ⫽ 3654) Age at baseline (yrs) ⬍55 55–64 65–74 75⫹ Gender, female Currently married Home ownership Job prestige: low Qualifications: trade certificate or higher Living alone RE visual impairment (VA 20/40 or worse) RE age-related maculopathy Early Late RE cataract Cortical Posterior subcapsular Nuclear History of Stroke Cancer Diabetes Heart disease (AMI or angina) Alcohol: any consumption Smoking status: current smoker
64.0
2335
21.2
14.3 36.3 35.0 14.3 57.5 65.7 89.6 37.6 50.1 25.8 5.6
335 848 817 335 1343 1536 2092 877 1171 599 131
3.8 0.6
Dead Age-adjusted P Value
Age-adjusted P Value
Crude %
n
776
14.8
543
17.0 33.2 29.8 20.0 60.8 62.1 81.7 41.0 44.1 25.9 10.8
132 258 231 155 472 482 634 318 342 198 84
0.1 0.1 0.0001 0.1 0.005 0.7 0.001
3.3 12.2 29.7 54.9 47.3 49.7 78.5 45.9 40.2 36.2 24.3
18 66 161 298 257 270 426 249 218 194 131
89 14
3.6 1.9
28 15
0.5 0.005
6.8 4.8
37 26
0.2 0.08
13.5 3.0 8.4
316 68 196
15.3 4.4 9.1
119 33 71
0.5 0.08 1
24.7 6.2 20.1
134 31 109
0.4 0.1 0.4
3.4 7.7 5.6 14.5 66.6 13.0
79 180 130 338 1555 303
5.0 6.2 7.7 14.1 61.3 18.2
39 48 60 109 476 141
0.08 0.1 0.03 0.5 0.1 0.0001
14.0 15.1 11.6 25.6 55.1 18.4
76 82 63 139 299 100
0.001 0.001 0.001 0.002 0.03 0.0001
0.001 0.05 0.001 0.06 0.05 0.9 0.0001
AMI ⫽ acute myocardial infarction; BMES I ⫽ Blue Mountains Eye Study I; RE ⫽ right eye; VA ⫽ visual acuity.
stitutionalized permanent residents born before January 1, 1943, at the time of the census were eligible. Persons living in the postcode area for more than 6 months of the year were defined as permanent residents. A total of 4433 noninstitutionalized and 625 institutionalized residents had birth dates before January 1, 1943. Of the 4433 eligible residents, 3654 (82.4%) participated in the examinations from 1992 to 1994. Ninety-nine percent were white. Of the 779 (17.6%) persons who did not participate, 353 (8.0%) permitted only an interview, 148 (3.3%) refused, 210 (4.7%) had moved out of the area, and 68 (1.5%) had died. A previous report compared participants and nonparticipants at baseline.21 Surviving members of the cohort were invited to return for a 5-year follow-up examination conducted from 1997 to 1999. Of the 3654 persons seen at baseline, 383 (10.5%) had moved, 394 (10.8%) refused to participate, and 543 (14.9%) had died. Thus 2335 persons (75.1% of survivors; 63.9% of original cohort) returned for a follow-up examination. Table 1 compares participants and nonparticipants. Compared with participants in BMES II, persons who were alive but did not participate in BMES II were more likely at baseline to have been in the age group 75⫹ (20.0% versus 14.3%), to have visual impairment ⬍20/40 in the right eye (10.8% versus 5.6%), or to have late age-related macular degeneration (1.9% versus 0.6%). They also had more difficulty in walking (7.9% versus 3.2%), were more likely current smokers (18.2% versus 13.0%), were less likely to have owned their home (81.7% versus 89.6%), or were less likely to have had qualifications after leaving school (44.1% versus 50.1%).
42
Procedures Previous BMES reports21–23 describe the survey methods and procedures. The Western Sydney Area Health Service Human Ethics Committee gave ethical approval for the study. All participants in the baseline and follow-up examinations gave written informed consent. Baseline and follow-up procedures were similar. Participants answered a detailed questionnaire and underwent a comprehensive eye examination after pupil dilatation. Visual acuity was measured by a logarithm of the minimum angle of resolution chart, retroilluminated with automatic calibration to 85 cd/m2 (Vectorvision CSV-100TM, Vectorvision, Inc, Dayton, OH), and read at 8 ft (244 cm).21 Distance visual acuity in each eye was initially measured with current distance glasses if worn. If participants read fewer than 54 letters correctly (20/20 Snellen equivalent), then a 1.2-mm pinhole aperture was held over the eye or glasses, and acuity was remeasured before subjective refraction. Best-corrected visual acuity was defined as the visual acuity after subjective refraction in the participant’s better eye. A Humphrey autorefractor (Model 530, Humphrey-Zeiss, Germany) provided objective refraction. Subjective refraction was performed according to the Beaver Dam Eye Study (BDES) modification of the Early Treatment Diabetic Retinopathy Study protocol.24 For each eye, visual acuity was recorded as the number of letters read correctly from 0 to 70 (20/200 –20/10). If no letters on the chart could be identified, visual acuity was assessed as count fingers at 2 ft (61 cm), hand movements, perception of light, or no perception of light. Intraocular pressure was measured using Goldmann applana-
Foran et al 䡠 Five-Year Change in Visual Acuity Table 2. Definitions of Persons at Risk for Visual Impairment and Incident Visual Impairment Used in this Report Monocular* and Binocular†
ⱕ20/40 ⬍20/40 ⬍20/70 ⱕ20/200 ⬍20/200
Level of Visual Impairment
At Risk‡
BDES any VI Driver’s license WHO ICD-10 category 1 BDES severe VI (US blindness) WHO ICD-10 category 2 (Australian blindness)
⬎20/40 ⱕ20/40 ⱖ20/40 ⬍20/40 ⱖ20/70 ⬍20/70 ⬎20/200 ⱕ20/200 ⱖ20/200 ⬍20/200
Incidence
*Monocular incidence refers to best-corrected visual acuity in one eye only at follow-up. † Binocular incidence refers to best-corrected visual acuity in both eyes at follow-up. ‡ At risk refers to best-corrected visual acuity in both eyes at baseline. BDES ⫽ Beaver Dam Eye Study, WHO ICD-10 ⫽ World Health Organization International Classification of Disease-10.
tion tonometry. Automated perimetry was performed in both eyes using the Humphrey 76-point test (Humphrey-Zeiss, Model 530, Germany). Persons with a hemifield defect of 5 or more points or other suspicious features were asked to return for the Humphrey 30-2 test. Pupils were dilated using 1.0% tropicamide and 10% phenylephrine. Detailed examination included slit-lamp (Topcon SL-7e camera, Topcon, Optical Co., Tokyo, Japan) and retroillumination (Neitz CT-R cataract camera, Neitz Instrument, Tokyo, Japan) photographs of the lens and stereoscopic photographs (Zeiss FF3, Carl Zeiss, Germany) of the optic disc and retina. For participants with incident visual impairment or doubling of the visual angle, an ophthalmologist (PM) reviewed their baseline and follow-up notes and photographs and attributed causes and their proportional contributions. The primary cause was defined as that cause estimated to be responsible for ⱖ50% of the impairment or deterioration.
Definitions For comparability this report retained the “any” and “severe” visual impairment definitions used by the Wisconsin BDES.13 We have also defined visual impairment by the visual acuity criterion
Figure 1. Mean change in number of correctly read letters in the right eye by age at baseline and gender.
used for driver’s license registration in Australia and in many parts of the United States. Finally, we have used the World Health Organization International Classification of Disease-10 WHOICD-10 criteria for visual impairment categories 1 and 2.25 Table 2 provides a summary of the definitions for persons at risk and incident cases used in this report. Unless otherwise specified, visual acuity refers to best-corrected acuity. Age was defined as age at the baseline examination. Visual impairment was defined as any, if visual acuity was 20/40 or worse (less than 41 letters read), or severe, if visual acuity was 20/200 or worse (0 –5 letters). An eye was at risk for any visual impairment if it had vision better than 20/40 and at risk for severe visual impairment if it had vision better than 20/200 at baseline. Persons developed binocular impairment if both eyes were at risk at baseline and both eyes had incident impairment at follow-up. Those who developed visual impairment in their fellow eye were thus not considered to have developed incident binocular impairment. Persons developed monocular impairment if both eyes were at risk at baseline and only one eye had incident impairment at follow-up. Incident impairment in either eye was defined if at least one eye was at risk at baseline (not necessarily both) and at least one eye at risk had impairment at the follow-up examination. The incidence of any and severe visual impairment was, respectively, defined as visual acuity 20/40 or worse and 20/200 or worse at follow-up in all eyes at risk. Thus, persons with incident monocular visual impairment had vision better than 20/40 in both eyes at baseline and had vision equal to or worse than 20/40 in only one eye at follow-up. Visual impairment also has been defined as worse than 20/40 based on the minimum legal visual acuity requirement for obtaining a driver’s license in Australia. This has been taken by this report as less than 39 letters read on the logarithm of the minimum angle of resolution chart (allowing one incorrect letter on the 20/40 line). The WHO-ICD-10 category 1 or greater defines visual impairment as worse than 20/70, taken as less than 29 letters on the logarithm of the minimum angle of resolution chart. The WHOICD-10 category 2 or greater defines visual impairment as worse than 20/200, taken as less than 4 letters on the logarithm of the minimum angle of resolution chart.25 The criterion for blind pension eligibility in Australia differs from that in the United States (20/200 or worse in both eyes) by
Figure 2. Level of visual impairment (20/40 –⬍20/70, ⬍20/70 –20/200, and ⬍20/200) among persons with incident visual impairment ⬍20/40 in either eye, by age at baseline in years, and gender.
43
Ophthalmology Volume 110, Number 1, January 2003 Table 3. Crude 5-year Incidence of Any and Severe Visual Impairment Monocularly and in the Better Eye by Age at Baseline and Gender in the Blue Mountains Eye Study Compared with Crude and Standardized Incidence in the Beaver Dam Eye Study Any Visual Impairment
Blue Mountains Eye Study
Age (Years) Monocular Women ⬍55 55–64 65–74 75⫹ Total Men ⬍55 55–64 65–74 75⫹ Total Both ⬍55 55–64 65–74 75⫹ Total Better eye Women ⬍55 55–64 65–74 75⫹ Total Men ⬍55 55–64 65–74 75⫹ Total Both ⬍55 55–64 65–74 75⫹ Total
At Risk
n
178 0 460 20 435 45 144 39 1217 104 146 354 310 94 904
Crude %
0.0 4.3 10.3 27.1
1 6 22 17 46
0.7 1.7 7.1 18.1
324 1 814 26 745 67 238 56 2121 150
0.3 3.2 9.0 23.5
181 477 467 186 1311
0 11 35 44 90
0.0 2.3 7.5 23.7
150 367 339 121 977
0 2 19 24 45
0.0 0.5 5.6 19.8
331 0 844 13 806 54 307 68 2288 135
0.0 1.5 6.7 22.1
95% Confidence Interval
Beaver Dam Eye Study At Risk
Crude %
626 507 426 126 8.5 (7.0, 10.1) 1685
1.3 3.0 11.5 27.0 6.3
5.1 (3.7, 6.5)
545 440 321 78 1384
0.7 2.3 10.0 24.4 4.7
7.1 (6.0, 8.2)
1171 947 747 204 3069
1.0 2.6 10.8 26.0 5.6
6.9 (5.5, 8.2)
656 551 521 208 1936
0.2 1.3 3.7 19.7 3.5
4.6 (3.3, 5.9)
574 471 373 110 1528
0.4 1.3 2.7 14.6 2.2
5.9 (4.9, 6.9)
1230 1022 894 318 3464
0.2 1.3 3.2 17.9 2.9
Severe Visual Impairment Standardized to Blue Mountains Eye Study 95% Confidence Interval
At Risk
n
Crude %
0.0 0.8 1.5 10.5
8.6 (7.3, 10.0)
183 0 475 4 466 7 190 20 1314 31
0.0 0.3 1.5 5.6
7.0 (5.7, 8.3)
148 0 361 1 330 5 124 7 963 13
0.0 0.6 1.5 8.6
7.9 (6.9, 8.8)
331 0 836 5 796 12 314 27 2277 44
0.0 0.0 0.8 6.5
4.6 (3.7, 5.5)
183 0 479 0 474 4 201 13 1337 17
3.3 (2.4, 4.2)
4.0 (3.4, 4.7)
requiring Snellen vision in the better eye of worse than 20/200 or a functional visual field of less than 10°. No participants in BMES I or II were eligible using the visual field criterion. In keeping with the BDES, deterioration or doubling of the visual angle was defined as a loss of 15 letters or more read correctly at follow-up compared with baseline. Similarly, improvement or halving of the visual angle was defined as an increase of 15 letters or more read correctly at follow-up compared with baseline. An eye was considered to be at risk of deterioration if at baseline it had perception of light or better, whereas an eye capable of improvement at baseline had read fewer than 56 letters. Similar to monocular and binocular visual impairment, persons were considered to have monocular or binocular deterioration or improvement. Incidence rates for visual impairment and doubling of the visual angle were determined for the better eye, whereas the
44
Blue Mountains Eye Study
150 367 342 128 987
0 0 2 5 7
0.0 0.0 0.6 3.9
333 0 846 0 816 6 329 18 2324 24
0.0 0.0 0.7 5.5
Beaver Dam Eye Study
Standardized to Blue Mountains Eye Study
95% Confidence Interval
At Risk
2.4 (1.5, 3.2)
648 533 515 224 1920
0.0 0.6 1.4 7.1 1.4
1.7 (1.2, 2.3)
1.3 (0.6, 2.1)
560 456 358 115 1489
0.0 0.0 1.1 4.4 0.6
0.9 (0.5, 1.4)
1.9 (1.4, 2.5)
1208 989 873 339 3409
0.0 0.3 1.3 6.2 1.0
1.4 (1.0, 1.8)
1.3 (0.7, 1.9)
662 553 544 249 2008
0.0 0.2 0.0 3.2 0.5
0.6 (0.2, 0.9)
0.7 (0.2, 1.2)
576 474 381 119 1550
0.0 0.0 0.0 0.8 0.1
0.1 (⫺0.1, 0.3)
1.0 (0.6, 1.4)
1238 1027 925 368 3558
0.0 0.1 0.0 2.5 0.3
0.4 (0.2, 0.6)
Crude 95% Confidence % Interval
incidence of halving of the visual angle was determined for the worse eye. In cases in which the right and left eye read the same number of letters, they were considered to be both the worse and the better eye.
Data Handling and Statistical Analysis Data were entered into dBase (Borland International Inc, Scotts Valley, CA) for BMES I and Microsoft Access (Microsoft Corporation, Redmond, WA) for BMES II. Statistical Analysis System (SAS Institute Inc, Cary, NC) was used for analyses, including t test, chi-square test, logistic regression, and generalized linear model analyses. Generalized linear model analyses adjusting for age and gender were used to compare right and left eyes, as well as men and women. BDES incidence rates were standardized to the BMES follow-up population to aid comparability. Confidence
Foran et al 䡠 Five-Year Change in Visual Acuity Table 4. Comparison of Different Levels of Binocular and Monocular Incident Visual Impairment by Age at Baseline and Gender <20/40
Binocular Women ⬍55 55–64 65–74 75⫹ Total Men ⬍55 55–64 65–74 75⫹ Total Both ⬍55 55–64 65–74 75⫹ Total Monocular Women ⬍55 55–64 65–74 75⫹ Total Men ⬍55 55–64 65–74 75⫹ Total Both ⬍55 55–64 65–74 75⫹ Total
<20/40
<20/200
<20/70
<20/200
At Risk
Incident
%
At Risk
Incident
%
At Risk
Incident
%
At Risk
Incident
%
At Risk
Incident
%
178 460 435 144 1217
0 4 8 17 29
0.0 0.9 1.8 11.8 2.4
179 462 440 147 1228
0 2 8 16 26
0.0 0.4 1.8 10.9 2.1
182 472 455 173 1282
0 0 2 8 10
0.0 0.0 0.4 4.6 0.8
183 475 466 190 1314
0 0 1 2 3
0.0 0.0 0.2 1.1 0.2
183 475 468 192 1318
0 0 1 1 2
0.0 0.0 0.2 0.5 0.2
146 354 310 94 904
0 0 3 9 12
0.0 0.0 1.0 9.6 1.3
146 356 313 99 914
0 0 4 7 11
0.0 0.0 1.3 7.1 1.2
148 360 325 116 949
0 0 2 3 5
0.0 0.0 0.6 2.6 0.5
148 361 330 124 963
0 0 0 0 0
0.0 0.0 0.0 0.0 0.0
148 362 331 125 966
0 0 0 0 0
0.0 0.0 0.0 0.0 0.0
324 814 745 238 2121
0 4 11 26 41
0.0 0.5 1.5 10.9 1.9
325 818 753 246 2142
0 2 12 23 37
0.0 0.2 1.6 9.3 1.7
330 832 780 289 2231
0 0 4 11 15
0.0 0.0 0.5 3.8 0.7
331 836 796 314 2277
0 0 1 2 3
0.0 0.0 0.1 0.6 0.1
331 837 799 317 2284
0 0 1 1 2
0.0 0.0 0.1 0.3 0.1
178 460 435 144 1217
0 20 45 39 104
0.0 4.3 10.3 27.1 8.5
179 462 440 147 1228
0 17 37 33 87
0.0 3.7 8.4 22.4 7.1
182 472 455 173 1282
0 11 16 29 56
0.0 2.3 3.5 16.8 4.4
183 475 466 190 1314
0 4 7 20 31
0.0 0.8 1.5 10.5 2.4
183 475 468 192 1318
0 3 8 19 30
0.0 0.6 1.7 9.9 2.3
146 354 310 94 904
1 6 22 17 46
0.7 1.7 7.1 18.1 5.1
146 356 313 99 914
1 7 21 18 47
0.7 2.0 6.7 18.2 5.1
148 360 325 116 949
0 5 12 11 28
0.0 1.4 3.7 9.5 3.0
148 361 330 124 963
0 1 5 7 13
0.0 0.3 1.5 5.6 1.3
148 362 331 125 966
0 1 5 8 14
0.0 0.3 1.5 6.4 1.4
324 814 745 238 2121
1 26 67 56 150
0.3 3.2 9.0 23.5 7.1
325 818 753 246 2142
1 24 58 51 134
0.3 2.9 7.7 20.7 6.3
330 832 780 289 2231
0 16 28 40 84
0.0 1.9 3.6 13.8 3.8
331 836 796 314 2277
0 5 12 27 44
0.0 0.6 1.5 8.6 1.9
331 837 799 317 2284
0 4 13 27 44
0.0 0.5 1.6 8.5 1.9
Best corrected visual acuity shown.
intervals were calculated at 95%. Means are given with their standard deviation and mean differences with their standard error.
Results The mean age of participants at baseline was 64.5 years. The mean follow-up period for the 2335 BMES II participants was 5.1 years (minimum, 3.0 years; maximum, 7.8 years). Women accounted for 57.5% of subjects. Table 1 shows the baseline characteristics of participants.
eyes P ⬍ 0.0001). There was an age-related decrease in the numbers of correctly read letters in both eyes. In the right eye, this decrease varied from ⫺0.5 ⫾ 3.4 letters in persons aged less than 55 years, ⫺1.5 ⫾ 5.9 in persons aged 55 to 64 years, ⫺3.2 ⫾ 8.3 in persons aged 65 to 74 years, to ⫺6.3 ⫾ 13.2 in persons aged 75 years or older. Figure 1 shows the mean change in number of correctly read letters in the right eye by age and gender. Age categorized into four levels (⬍55, 55– 64, 65–74 and ⬎75 years) was a significant determinant of decrease in visual acuity in both eyes (P ⬍ 0.0001). Women had a greater change in the number of letters read incorrectly than men in the worse eye (age-adjusted means ⫺3.1 ⫾ 0.2 vs. ⫺2.6 ⫾ 0.3, P ⫽ 0.2), and in the better eye (age-adjusted means ⫺2.4 ⫾ 0.2 vs. ⫺2.1 ⫾ 0.2, P ⫽ 0.1).
Change in Visual Acuity The change in numbers of letters read correctly (5-year follow-up compared with baseline examinations) in the right eye was ⫺2.6 ⫾ 8.2 (mean ⫾ standard deviation) and ⫺2.5 ⫾ 7.2 in the left eye. After age and gender adjustment, the difference between right and left eyes in mean number of letters read correctly was not statistically significant (P ⫽ 0.53). The numbers of letters correctly read by right and left eyes were not significantly different (Pearson correlation between
Presenting versus Best-Corrected Visual Acuity Compared with best-corrected visual acuity, presenting visual acuity (with current glasses if worn) was worse by approximately 1 line at both the baseline and follow-up examinations. At baseline, right eyes read 45.5 ⫾ 12.8 letters correctly at presentation compared with 51.5 ⫾ 11.1 after best subjective refraction. At followup, presenting and best-corrected vision were 44.4 ⫾ 13.1 and 50.5
45
Ophthalmology Volume 110, Number 1, January 2003 Table 5. Crude 5-year Incidence Doubling and Halving of the Visual Angle by Age at Baseline and Gender in the Blue Mountains Doubling of the Visual Angle
Beaver Dam Eye Study
Standard Blue Mountains Eye Study
95% Confidence Interval
At Risk
Crude %
95% Confidence Interval
6.7 (5.3, 8.0)
650 537 521 229 1937
0.8 2.8 6.0 14.4 4.3
5.4 (4.4, 6.4)
6.1 (4.6, 7.6)
565 461 361 113 1500
1.1 2.2 7.8 11.5 3.8
5.2 (4.1, 6.3)
6.4 (5.4, 7.4)
1215 998 882 342 3437
0.9 2.5 6.7 13.5 4.1
5.3 (4.6, 6.1)
1.7 (1.0, 2.4)
663 553 545 255 2016
0.2 1.5 1.5 8.6 1.9
2.4 (1.7, 3.1)
0.8 (0.3, 1.4)
576 474 381 120 1551
0.0 0.6 2.1 8.3 1.4
2.0 (1.3, 2.7)
1.3 (0.9, 1.8)
1239 1027 926 375 3567
0.1 1.1 1.7 8.5 1.7
2.2 (1.7, 2.7)
Blue Mountains Eye Study
Age (Years) Monocular Women ⬍55 55–64 65–74 75⫹ Total Men ⬍55 55–64 65–74 75⫹ Total Both ⬍55 55–64 65–74 75⫹ Total Better eye Women ⬍55 55–64 65–74 75⫹ Total Men ⬍55 55–64 65–74 75⫹ Total Both ⬍55 55–64 65–74 75⫹ Total
At Risk
Crude %
n
183 479 473 202 1337
1 17 37 34 89
0.5 3.5 7.8 16.8
150 365 341 127 983
1 6 32 21 60
0.7 1.6 9.4 16.5
333 844 814 329 2320
2 23 69 55 149
0.6 2.7 8.5 16.7
183 479 474 204 1340
0 1 7 15 23
0.0 0.2 1.5 7.4
150 367 343 127 987
0 1 3 4 8
0.0 0.3 0.9 3.1
333 846 817 331 2327
0 2 10 19 31
0.0 0.2 1.2 5.7
Incidences of doubling of the visual angle given monocularly and in the better eye, halving of the visual angle given monocularly and in the worse eye.
⫾ 11.7 letters, respectively. In both right and left eyes, the mean differences in best-corrected and presenting vision at baseline and at follow-up were not statistically different. For instance, in the right eye at baseline the mean improvement in best-corrected vision over presenting vision was 6.0 ⫾ 7.4 letters read correctly compared with 6.2 ⫾ 7.2 letters at follow-up.
Incidence of Visual Impairment Any Visual Impairment (<20/40). Five-year incident visual impairment occurred in 267 eyes: 137 (51.3%) right compared with 130 (48.6%) left eyes (P ⫽ 0.51). Figure 2 shows the incidence of visual impairment in either eye by level of impairment. Incident binocular impairment occurred in 41 persons (1.9%, 95% confi-
46
dence interval [CI] 1.3–2.5) and monocular impairment in 150 persons (7.1%, 95% CI 6.0 – 8.2) (Table 3). Women accounted for approximately 70% of binocular and monocular impairment compared with men. After controlling for gender, each year of increasing age was associated with roughly a 20% increased risk of incident binocular visual impairment (P ⫽ 0.0001, gender-adjusted odds ratio [OR] 1.21, 95% CI 1.15–1.26). After adjusting for age, women had a 73% increased risk of incident binocular impairment compared with men, but this did not reach statistical significance (P ⫽ 0.1). Similarly, after controlling for gender, each year of increasing age was associated with a 13% significantly increased risk for incident monocular impairment (P ⫽ 0.0001, gender-adjusted OR 1.13, 95% CI 1.10 –1.15). After adjusting for
Foran et al 䡠 Five-Year Change in Visual Acuity Eye Study Compared with the Crude and Standardized Incidence in the Beaver Dam Eye Study Halving of the Visual Angle
Blue Mountains Eye Study Crude %
At risk
n
48 151 259 181 637
2 3 4 8 17
4.2 2.0 1.5 4.4
28 74 149 105 356
0 0 2 5 7
0.0 0.0 1.3 4.8
76 225 408 286 995 Worse eye
2 3 6 13 24
2.6 1.3 1.5 4.5
98 317 385 198 998
2 3 6 7 18
2.0 0.9 1.6 3.5
61 191 247 120 619
0 0 6 6 12
0.0 0.0 2.4 5.0
159 508 632 318 1617
2 3 12 13 30
1.3 0.6 1.9 4.1
Beaver Dam Eye Study
Standard Blue Mountains Eye Study
95% Confidence Interval
At Risk
Crude %
95% Confidence Interval
2.7 (1.4, 3.9)
153 197 294 188 832
4.6 4.6 7.5 4.8 5.7
5.8 (4.2, 7.4)
2.0 (0.5, 3.9)
60 76 121 80 337
3.3 0 5 5 3.6
3.8 (1.8, 5.9)
2.4 (1.5, 3.4)
213 273 415 268 1169
4.2 3.3 6.8 4.9 5.1
5.3 (4.0, 6.5)
1.8 (⫺0.1, 0.7)
329 397 455 245 1426
1.5 1.8 4.2 4.1 2.9
3.2 (2.2, 4.1)
1.9 (0.0, 2.2)
202 217 263 108 790
1.0 0.0 1.9 4.6 1.5
1.7 (0.8, 2.7)
1.9 (0.1, 1.1)
531 614 718 353 2216
1.3 1.1 3.3 4.3 2.4
2.6 (1.9, 3.3)
age, women had a 70% significantly increased risk of incident monocular impairment compared with men (P ⫽ 0.005, ageadjusted OR 1.70, 95% CI 1.18 –2.47). Visual Impairment <20/40. Using the visual impairment definition of ⬍20/40 (legal driving cutpoint), 236 eyes (5.1%) developed incident visual impairment in 127 (53.8%) right and 109 (46.2%) left eyes (P ⫽ 0.2). Binocular impairment ⬍20/40 occurred in 37 persons (1.7%), 4 persons less than for binocular impairment ⱕ20/40. Monocular impairment ⬍20/40 occurred in 134 persons (6.3%), 16 persons less than for monocular impairment ⱕ20/40. For incident binocular impairment ⬍20/40, 70.3% of persons were women and 29.7% were men. For incident monocular impairment 64.9% of persons were women and 35.1% were men. Although women had higher incidence rates than men, these
differences were not significant after controlling for age. For each year of increasing age, the incidence of binocular impairment increased by 22% (gender-adjusted OR 1.22, 95% CI 1.16 –1.29), and the incidence of monocular impairment increased by 12% (gender-adjusted OR 1.12, 95% CI 1.09 –1.15). Table 4 shows the incidence of binocular and monocular impairment using different classifications by age and gender. Visual Impairment <20/70. Using the WHO-ICD-10 definition of ⬍20/70 (category 1 or worse) to define visual impairment, 124 eyes (2.7%) developed incident visual impairment in 72 (58.1%) right and 52 (41.9%) left eyes (P ⫽ 0.06). Binocular impairment occurred in 15 persons (0.7%), of whom 10 (66.7%) were women and 5 (33.3%) were men; this was not statistically significant after controlling for age (P ⫽ 0.6). Monocular impair-
47
Ophthalmology Volume 110, Number 1, January 2003
Figure 3. Proportionate causes of incident monocular visual impairment, severe visual impairment, and doubling of the visual angle by age at baseline in years. ARM ⫽ Age-related maculopathy.
ment developed in 84 persons (3.8%), with similar gender differences. Age (per year) was the significant determinant of both binocular (P ⫽ 0.0001, gender-adjusted OR 1.23, 95% CI 1.14 – 1.32) and monocular impairment (P ⫽ ⬍0.001, gender-adjusted OR 1.13, 95% CI 1.10 –1.7). Severe Visual Impairment <20/200. Five-year incident visual impairment ⱕ20/200 developed in 52 eyes (1.1%): 34 right (65.4%) and 18 left eyes (34.6%). Right eyes were twice as likely to develop severe impairment than left eyes after controlling for age and gender (P ⫽ 0.02, age-gender adjusted OR 1.99, CI 1.11–3.56). Three persons (0.1%, 95% CI 0.0 – 0.3) had incident severe binocular incident impairment, all of whom were women, aged 72, 78, and 80 at baseline. Forty-four persons (1.9%, 95% CI 1.4 –2.5) had incident severe monocular impairment in 29 (65.9%) right eyes and 15 (34.1%) left eyes. Of this group, 31 (70.4%) were women and 13 (29.6%) were men. Each year of increasing age was associated with a 15% increased risk of incident severe monocular visual impairment (P ⫽ ⬍0.001, gender-adjusted OR 1.14, 95% CI 1.10 –1.19) after controlling for gender. However, female gender was not associated with a significantly increased risk of incident visual impairment (P ⫽ 0.2, age-adjusted OR 1.63, 95% CI 0.0.84 –3.17). Severe Visual Impairment <20/200. Five-year incident visual impairment ⬍20/200 occurred in 50 eyes (1.1%): 33 right (66.7%) compared with 17 left eyes (33.3%). Similar to severe visual impairment ⬍20/200, right eyes were twice as likely to develop severe impairment than left eyes after controlling for age and gender (P ⫽ 0.02, age-gender adjusted OR 2.03, 95% CI 1.12– 3.69). Two persons (0.1%, 95% CI 0.0 – 0.2) had incident severe binocular incident impairment; both were women, aged 72 and 78. Similar to the incidence of monocular impairment ⱕ20/200, 44 persons (1.9%, 95% CI 1.4 –2.5) had incident severe monocular impairment in 29 (65.9%) right eyes and 15 (34.1%) left eyes. Of this group, 30 (68.2%) were women and 14 (31.8%) were men. Again, each year of increasing age was associated with a 15% increased risk of incident severe monocular visual impairment (P ⫽ 0.0001, gender-adjusted OR 1.15, 95% CI 1.11–1.20) after controlling for gender. However, female gender was not associated with a significantly increased risk of incident visual impairment (P ⫽ 0.3, age-adjusted OR 1.45, 95% CI 0.75–2.78). Doubling of the Visual Angle (Deterioration). The doubling
48
of the visual angle (a decrease in 15 or more letters read correctly) that occurred binocularly in 31 persons (1.3%, 95% CI 0.9 –1.8) was age related: none was aged ⬍55 years, 2 (0.2%) were aged 55 to 64 years, 10 (1.2%) were aged 65 to 74 years, and 19 persons (5.8%) were aged 75 years. Of the 149 persons (6.4%, 95% CI 5.4 –7.4) with monocular doubling of the visual angle, 2 persons (0.6%) were aged ⬍55 years, 23 (2.7%) 55 to 64 years, 69 (8.5%) 65 to 74 years, and 55 persons (16.7%) aged 75 years and older (Table 5). Monocular doubling of the visual angle was significantly associated with age (P ⫽ 0.0001, OR 1.10, 95% CI 1.08 –1.12) but not with gender (P ⫽ 0.9, age-adjusted OR 1.01, 95% CI 0.72–1.44). Similar associations were found with bilateral doubling of the visual angle. Halving of the Visual Angle (Improvement). Halving of the visual angle (an increase in 15 or more letters read correctly) occurred in both eyes in 5 persons (0.5%, 95% CI 0.1– 0.9) and in one eye in 24 persons (2.4%, 95% CI 1.5–3.4) (Table 5). Persons with improvement were more likely to have had incident cataract surgery, that is, cataract surgery performed after the baseline examination but before the follow-up examination. Of the 24 persons with monocular improvement, 14 (58.3%) had incident cataract surgery in that eye (P ⫽ ⬍0.001). After controlling for age and gender, monocular improvement was strongly associated with incident cataract surgery in that eye (P ⫽ ⬍0.001, ageadjusted OR 9.6, 95% CI 4.0 –22.9). For both binocular and monocular improvement, age and gender were not statistically significant (P ⬎ 0.1). Attributable Causes. The most frequent primary cause of any incident visual impairment was cataract, whereas for severe incident visual impairment the most frequent primary cause was age-related maculopathy. Figure 3 shows the proportion of attributable causes for any and severe visual impairment and for doubling of the visual angle.
Discussion Incident figures from this report can be used to project the number of incident cases of visual impairment within the Australian population. By using population estimates for the year 2001 from the Australian Bureau of Statistics26 with our incident visual impairment figures, we estimate that in 5 years approximately 94,700 persons will develop visual impairment ⬍20/40 in both eyes; these persons will no longer be able to obtain a driver’s license based on the current visual acuity criterion. This report from the BMES supplements data from the BDES on the change in visual acuity among members of an older population.13 Previous reports from these two studies13,21 have indicated similar findings for the prevalence of visual impairment in these similar, older, largely white communities in the United States and Australia. The mean number of letters read correctly after refraction was almost identical between these two studies within the same age and gender strata. The BMES follow-up examined 75.1% of survivors from the initial study. We may have underestimated the decrease in visual acuity or the incidence of visual impairment, because persons returning were younger, healthier, and better educated than those who had died since BMES I (Table 1). Persons who were alive but did not participate in the follow-up were similar in many parameters to those who did
Foran et al 䡠 Five-Year Change in Visual Acuity participate. They were significantly less likely to own their home or to have gained further qualifications after leaving school. They had a significantly higher self-reported history of diabetes and also were more likely to be current smokers. In keeping with this, they had more visual impairment and late age-related maculopathy at baseline. As a result, we may have underestimated the incidence of visual impairment, especially severe impairment. In Western populations, smoking is the strongest modifiable risk factor for the development of late-stage age-related maculopathy.27–29 Although there is no absolute visual acuity threshold, in Australia, cataract surgery is usually not performed unless visual acuity is worse than 20/40. Persons who were not visually impaired at baseline, but who had cataract surgery between the baseline and follow-up examinations, could also potentially have been classified as incident cases of visual impairment. If these persons had been included, the incidence of visual impairment would be higher. Among those at risk for monocular visual impairment ⱕ20/40 who did not develop incident impairment, 30 persons had binocular cataract surgery and 45 persons had monocular surgery. Including these persons would have increased the incidence of binocular impairment from 1.9% to 3.3% (74% higher) and monocular impairment from 7.1% to 9.2% (30% higher). Women generally had higher rates of impairment than men even after adjusting for age. This may be partially because women have a higher prevalence of cataract, especially cortical cataract, than men30 and that women are also less likely to have cataract surgery. We also found that right eyes were more likely to be impaired than left eyes and that there was a significant difference for severe impairment ⱕ20/200 and ⬍20/200. This could be due to measurement bias; to learning by the participant, because right eyes were systematically tested first; or to chance. We also found substantially higher rates of incident binocular impairment than expected if the incident monocular rates for each eye were independent. This confirms the symmetry of common eye diseases particularly cataract and age-related maculopathy. Compared with the BDES standardized rate (7.9%), the BMES found a lower rate of any monocular impairment ⱕ20/40 overall (7.1%), although confidence intervals were overlapping (Table 3). This difference was most marked for men (5.1% BMES versus 7.0% BDES). In contrast, we found higher incidences of any impairment in the better eye (nonoverlapping confidence intervals overall), monocular severe visual impairment, and better eye severe impairment (overlapping confidence intervals). Similar to any impairment, our study found a slightly higher incidence of monocular deterioration (6.4% BMES vs. 5.3% BDES) and lower incidence of deterioration in the better eye (1.3% BMES vs. 2.2% BDES). Consistent with this, improvement rates in the BMES were lower in the monocular (2.4% BMES vs. 5.3% BDES) and better eye (1.9% BMES vs. 2.6% BDES) compared with the BDES. We could not compare rates for binocular impairment, deterioration, or halving with those from the BDES, because these data were not reported.13 The BDES found a strong relationship between improvement of visual acuity and cataract surgery.
Incidence rates for cataract surgery, however, differed substantially between these two populations. For instance, the BDES age-adjusted incidence rate for cataract surgery31 in persons with any cortical cataract at baseline was 32.0% compared with the BMES age-adjusted rate of 8.9% (Invest Ophthalmol Vis Sci 2001;42[Suppl]:S508). Our finding that the mean differences in best-corrected and presenting vision at baseline and at 5-year follow-up were similar suggests that the study may not have greatly influenced the likelihood that people would seek attention for under or uncorrected refractive error. All participants whose visual acuity corrected by 2 or more lines (10⫹ letters) were advised that their vision could be improved by a change in their current glasses or by wearing distance glasses, if not previously worn. The study region has ready access to optometric and ophthalmic services. If the study had altered behavior, its effects may have been partly balanced by the age-related decline in visual acuity and the development of cataract and other conditions. In summary, our study found that increases in visual impairment were strongly age related. Women were more likely to have incident impairment than men, even after adjusting for age. Our rates were similar to those reported by the BDES. Population-based incident data are vital to health planners determining the allocation of health resources. The 10-year incidence rates of visual impairment will provide further important data. Attributable causes of incident impairment will also be necessary to determine the number of cases that could be prevented or treated.
References 1. Tielsch JM, Sommer A, Witt K, et al. Blindness and visual impairment in an American urban population. The Baltimore Eye Survey [see comments] Arch Ophthalmol 1990;108:286 –90. 2. Casson R, Giles L, Newland HS. Prevalence of blindness and visual impairment in an elderly urban population. Aust N Z J Ophthalmol 1996;24:239 – 43. 3. Taylor HR, Livingston PM, Stanislavsky YL, McCarty CA. 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 –37. 4. Wang JJ, Foran S, Mitchell P. Age-specific prevalence and causes of bilateral and unilateral visual impairment in older Australians: the Blue Mountains Eye Study. Clin Exp Ophthalmol 2000;28:268 –73. 5. Rahmani B, Tielsch JM, Katz J, et al. The cause-specific prevalence of visual impairment in an urban population. The Baltimore Eye Survey. Ophthalmology 1996;103:1721– 6. 6. Mun˜ oz B, West SK, Rubin GS, et al. Causes of blindness and visual impairment in a population of older Americans: the Salisbury Eye Evaluation Study. Arch Ophthalmol 2000;118: 819 –25. 7. Klaver CCW, Wolfs RCW, Vingerling JR, et al. Age-specific prevalence and causes of blindness and visual impairment in an older population. The Rotterdam Study. Arch Ophthalmol 1998;116:653– 8. 8. Hirvela¨ H, Laatikainen L. Visual acuity in a population aged 70 years or older; prevalence and causes of visual impairment. Acta Ophthalmol Scand 1995;73:99 –104. 9. Dana MR, Tielsch JM, Enger C, et al. Visual impairment in a
49
Ophthalmology Volume 110, Number 1, January 2003
10. 11. 12. 13. 14.
15. 16. 17. 18. 19.
rural Appalachian community: prevalence and causes. JAMA 1990;264:2400 –5. Ponte F, Giuffre` G, Giammanco R. Prevalence and causes of blindness and low vision in the Casteldaccia Eye Study. Graefes Arch Clin Exp Ophthalmol 1994;232:469 –72. Newland HS, Hiller JE, Casson RJ, Obermeder S. Prevalence and causes of blindness in the South Australian population aged 50 and over. Ophthalmic Epidemiol 1996;3:97–107. Weih LM, VanNewkirk MR, McCarty CA, Taylor HR. Agespecific causes of bilateral visual impairment. Arch Ophthalmol 2000;118:264 –9. Klein R, Klein BEK, Lee KE. Changes in visual acuity in a population. The Beaver Dam Eye Study. Ophthalmology 1996;103:1169 –78. Wang JJ, Mitchell P, Smith W, et al. Impact of visual impairment on use of community support services by elderly persons: the Blue Mountains Eye Study. Invest Ophthalmol Vis Sci 1999;40:12–9. Mitchell P, Hayes P, Wang JJ. Visual impairment in nursing home residents: the Blue Mountains Eye Study. Med J Aust 1997;166:73– 6. Ivers RQ, Cumming RG, Mitchell P, Attebo K. Visual impairment and falls in older adults: the Blue Mountains Eye Study. J Am Geriatr Soc 1998;46:58 – 64. Campbell AJ, Reinken J, Allan BC, Martinez GS. Falls in old age: a study of frequency and related clinical factors. Age Ageing 1981;10:264 –70. McCarty CA, Nanjan MB, Taylor HR. Vision impairment predicts 5 year mortality. Br J Ophthalmol 2001;85:322– 6. Wang JJ, Mitchell P, Simpson JM, et al. Visual impairment, age-related cataract and mortality. Arch Ophthalmol 2001; 119:1186 –90.
20. Wang JJ, Mitchell P, Smith W. Vision and low self-rated health: the Blue Mountains Eye Study. Invest Ophthalmol Vis Sci 2000;41:49 –54. 21. Attebo K, Mitchell P, Smith W. Visual acuity and the causes of visual loss in Australia. The Blue Mountains Eye Study. Ophthalmology 1996;103:357– 64. 22. Mitchell P, Smith W, Attebo K, Wang JJ. Prevalence of age-related maculopathy in Australia. The Blue Mountains Eye Study. Ophthalmology 1995;102:1450 – 60. 23. Mitchell P, Smith W, Attebo K, Healey PR. Prevalence of open-angle glaucoma in Australia. The Blue Mountains Eye Study. Ophthalmology 1996;103:1661–9. 24. Klein R, Klein BEK, Linton KLP, De Mets DL. The Beaver Dam Eye Study: visual acuity. Ophthalmology 1991;98:1310 –5. 25. ICD 10 Definition. Geneva: World Health Organization. Available from URL: www.who.int/pbd/pbl/img/icd10.gif. 26. Population Projections: 1997 to 2051. Canberra: Australian Bureau of Statistics, 1998;1–130. No. 3222.0. 27. Smith W, Mitchell P, Leeder SR. Smoking and age-related maculopathy: the Blue Mountains Eye Study. Arch Ophthalmol 1996;114:1518 –23. 28. Mitchell P, Chapman S, Smith W. “Smoking is a major cause of blindness:” A new cigarette pack warning. [editorial] Med J Aust 1999;171:173– 4. 29. Smith W, Assink J, Klein R, et al. Risk factors for age-related macular degeneration. Pooled findings from three continents. Ophthalmology 2001;108:697–704. 30. Mitchell P, Cumming RG, Attebo K, Panchapakesan J. Prevalence of cataract in Australia. The Blue Mountains Eye Study. Ophthalmology 1997;104:581– 8. 31. Klein BEK, Klein R, Moss SE. Incident cataract surgery. The Beaver Dam Eye Study. Ophthalmology 1997;104:573– 80.
(Questions on p. 22).
1. 2. 3. 4. 5.
B B D B C
Answers:
50
Recent population studies have provided important data on the prevalence1–3 and causes of visual impairment4 –12 in older persons, but few population reports describe the change in visual acuity or the incidence of visual impairment over time.13 Other reports show that visual impairment increases significantly with age and is associated with use of community support services,14 nursing home placement,15 falls,16,17 mortality,18,19 and factors determining self-rated Originally received: August 3, 2001. Accepted: April 17, 2002. Manuscript no. 210694. From the Department of Ophthalmology and the Save Sight and Westmead Millennium Institutes, the University of Sydney, Sydney, Australia. Supported by the Australian National Health & Medical Research Council, Canberra, Australia. Reprint requests to Paul Mitchell, MD, PhD, Department of Ophthalmology, University of Sydney, Hawkesbury Rd, Westmead, NSW, Australia, 2145; E-mail: [email protected] © 2003 by the American Academy of Ophthalmology Published by Elsevier Science Inc.
health.20 This report from the Blue Mountains Eye Study (BMES) describes the 5-year change in visual acuity and the incidence and progression of visual impairment in a representative sample of older Australians.
Materials and Methods Population The BMES is a population-based survey of vision and common eye diseases in an urban population aged 49 years or older residing in two postal codes of the Blue Mountains region, west of Sydney, Australia. A previous report from the study explained the reasons for selecting, and the methods used to identify, the target population and described the population.21 To summarize, study personnel conducted a door-to-door census from November to December 1991 for the first postcode area and from March to April 1993 for the second postcode area. All noninISSN 0161-6420/03/$–see front matter PII S0161-6420(02)01295-2
41
Ophthalmology Volume 110, Number 1, January 2003 Table 1. Baseline Characteristics of Participants and Nonparticipants in the Blue Mountains Eye Study: 5-Year Examinations Nonparticipants Participants
Alive
Characteristics
Crude %
n
Crude %
n
Total (BMES I, n ⫽ 3654) Age at baseline (yrs) ⬍55 55–64 65–74 75⫹ Gender, female Currently married Home ownership Job prestige: low Qualifications: trade certificate or higher Living alone RE visual impairment (VA 20/40 or worse) RE age-related maculopathy Early Late RE cataract Cortical Posterior subcapsular Nuclear History of Stroke Cancer Diabetes Heart disease (AMI or angina) Alcohol: any consumption Smoking status: current smoker
64.0
2335
21.2
14.3 36.3 35.0 14.3 57.5 65.7 89.6 37.6 50.1 25.8 5.6
335 848 817 335 1343 1536 2092 877 1171 599 131
3.8 0.6
Dead Age-adjusted P Value
Age-adjusted P Value
Crude %
n
776
14.8
543
17.0 33.2 29.8 20.0 60.8 62.1 81.7 41.0 44.1 25.9 10.8
132 258 231 155 472 482 634 318 342 198 84
0.1 0.1 0.0001 0.1 0.005 0.7 0.001
3.3 12.2 29.7 54.9 47.3 49.7 78.5 45.9 40.2 36.2 24.3
18 66 161 298 257 270 426 249 218 194 131
89 14
3.6 1.9
28 15
0.5 0.005
6.8 4.8
37 26
0.2 0.08
13.5 3.0 8.4
316 68 196
15.3 4.4 9.1
119 33 71
0.5 0.08 1
24.7 6.2 20.1
134 31 109
0.4 0.1 0.4
3.4 7.7 5.6 14.5 66.6 13.0
79 180 130 338 1555 303
5.0 6.2 7.7 14.1 61.3 18.2
39 48 60 109 476 141
0.08 0.1 0.03 0.5 0.1 0.0001
14.0 15.1 11.6 25.6 55.1 18.4
76 82 63 139 299 100
0.001 0.001 0.001 0.002 0.03 0.0001
0.001 0.05 0.001 0.06 0.05 0.9 0.0001
AMI ⫽ acute myocardial infarction; BMES I ⫽ Blue Mountains Eye Study I; RE ⫽ right eye; VA ⫽ visual acuity.
stitutionalized permanent residents born before January 1, 1943, at the time of the census were eligible. Persons living in the postcode area for more than 6 months of the year were defined as permanent residents. A total of 4433 noninstitutionalized and 625 institutionalized residents had birth dates before January 1, 1943. Of the 4433 eligible residents, 3654 (82.4%) participated in the examinations from 1992 to 1994. Ninety-nine percent were white. Of the 779 (17.6%) persons who did not participate, 353 (8.0%) permitted only an interview, 148 (3.3%) refused, 210 (4.7%) had moved out of the area, and 68 (1.5%) had died. A previous report compared participants and nonparticipants at baseline.21 Surviving members of the cohort were invited to return for a 5-year follow-up examination conducted from 1997 to 1999. Of the 3654 persons seen at baseline, 383 (10.5%) had moved, 394 (10.8%) refused to participate, and 543 (14.9%) had died. Thus 2335 persons (75.1% of survivors; 63.9% of original cohort) returned for a follow-up examination. Table 1 compares participants and nonparticipants. Compared with participants in BMES II, persons who were alive but did not participate in BMES II were more likely at baseline to have been in the age group 75⫹ (20.0% versus 14.3%), to have visual impairment ⬍20/40 in the right eye (10.8% versus 5.6%), or to have late age-related macular degeneration (1.9% versus 0.6%). They also had more difficulty in walking (7.9% versus 3.2%), were more likely current smokers (18.2% versus 13.0%), were less likely to have owned their home (81.7% versus 89.6%), or were less likely to have had qualifications after leaving school (44.1% versus 50.1%).
42
Procedures Previous BMES reports21–23 describe the survey methods and procedures. The Western Sydney Area Health Service Human Ethics Committee gave ethical approval for the study. All participants in the baseline and follow-up examinations gave written informed consent. Baseline and follow-up procedures were similar. Participants answered a detailed questionnaire and underwent a comprehensive eye examination after pupil dilatation. Visual acuity was measured by a logarithm of the minimum angle of resolution chart, retroilluminated with automatic calibration to 85 cd/m2 (Vectorvision CSV-100TM, Vectorvision, Inc, Dayton, OH), and read at 8 ft (244 cm).21 Distance visual acuity in each eye was initially measured with current distance glasses if worn. If participants read fewer than 54 letters correctly (20/20 Snellen equivalent), then a 1.2-mm pinhole aperture was held over the eye or glasses, and acuity was remeasured before subjective refraction. Best-corrected visual acuity was defined as the visual acuity after subjective refraction in the participant’s better eye. A Humphrey autorefractor (Model 530, Humphrey-Zeiss, Germany) provided objective refraction. Subjective refraction was performed according to the Beaver Dam Eye Study (BDES) modification of the Early Treatment Diabetic Retinopathy Study protocol.24 For each eye, visual acuity was recorded as the number of letters read correctly from 0 to 70 (20/200 –20/10). If no letters on the chart could be identified, visual acuity was assessed as count fingers at 2 ft (61 cm), hand movements, perception of light, or no perception of light. Intraocular pressure was measured using Goldmann applana-
Foran et al 䡠 Five-Year Change in Visual Acuity Table 2. Definitions of Persons at Risk for Visual Impairment and Incident Visual Impairment Used in this Report Monocular* and Binocular†
ⱕ20/40 ⬍20/40 ⬍20/70 ⱕ20/200 ⬍20/200
Level of Visual Impairment
At Risk‡
BDES any VI Driver’s license WHO ICD-10 category 1 BDES severe VI (US blindness) WHO ICD-10 category 2 (Australian blindness)
⬎20/40 ⱕ20/40 ⱖ20/40 ⬍20/40 ⱖ20/70 ⬍20/70 ⬎20/200 ⱕ20/200 ⱖ20/200 ⬍20/200
Incidence
*Monocular incidence refers to best-corrected visual acuity in one eye only at follow-up. † Binocular incidence refers to best-corrected visual acuity in both eyes at follow-up. ‡ At risk refers to best-corrected visual acuity in both eyes at baseline. BDES ⫽ Beaver Dam Eye Study, WHO ICD-10 ⫽ World Health Organization International Classification of Disease-10.
tion tonometry. Automated perimetry was performed in both eyes using the Humphrey 76-point test (Humphrey-Zeiss, Model 530, Germany). Persons with a hemifield defect of 5 or more points or other suspicious features were asked to return for the Humphrey 30-2 test. Pupils were dilated using 1.0% tropicamide and 10% phenylephrine. Detailed examination included slit-lamp (Topcon SL-7e camera, Topcon, Optical Co., Tokyo, Japan) and retroillumination (Neitz CT-R cataract camera, Neitz Instrument, Tokyo, Japan) photographs of the lens and stereoscopic photographs (Zeiss FF3, Carl Zeiss, Germany) of the optic disc and retina. For participants with incident visual impairment or doubling of the visual angle, an ophthalmologist (PM) reviewed their baseline and follow-up notes and photographs and attributed causes and their proportional contributions. The primary cause was defined as that cause estimated to be responsible for ⱖ50% of the impairment or deterioration.
Definitions For comparability this report retained the “any” and “severe” visual impairment definitions used by the Wisconsin BDES.13 We have also defined visual impairment by the visual acuity criterion
Figure 1. Mean change in number of correctly read letters in the right eye by age at baseline and gender.
used for driver’s license registration in Australia and in many parts of the United States. Finally, we have used the World Health Organization International Classification of Disease-10 WHOICD-10 criteria for visual impairment categories 1 and 2.25 Table 2 provides a summary of the definitions for persons at risk and incident cases used in this report. Unless otherwise specified, visual acuity refers to best-corrected acuity. Age was defined as age at the baseline examination. Visual impairment was defined as any, if visual acuity was 20/40 or worse (less than 41 letters read), or severe, if visual acuity was 20/200 or worse (0 –5 letters). An eye was at risk for any visual impairment if it had vision better than 20/40 and at risk for severe visual impairment if it had vision better than 20/200 at baseline. Persons developed binocular impairment if both eyes were at risk at baseline and both eyes had incident impairment at follow-up. Those who developed visual impairment in their fellow eye were thus not considered to have developed incident binocular impairment. Persons developed monocular impairment if both eyes were at risk at baseline and only one eye had incident impairment at follow-up. Incident impairment in either eye was defined if at least one eye was at risk at baseline (not necessarily both) and at least one eye at risk had impairment at the follow-up examination. The incidence of any and severe visual impairment was, respectively, defined as visual acuity 20/40 or worse and 20/200 or worse at follow-up in all eyes at risk. Thus, persons with incident monocular visual impairment had vision better than 20/40 in both eyes at baseline and had vision equal to or worse than 20/40 in only one eye at follow-up. Visual impairment also has been defined as worse than 20/40 based on the minimum legal visual acuity requirement for obtaining a driver’s license in Australia. This has been taken by this report as less than 39 letters read on the logarithm of the minimum angle of resolution chart (allowing one incorrect letter on the 20/40 line). The WHO-ICD-10 category 1 or greater defines visual impairment as worse than 20/70, taken as less than 29 letters on the logarithm of the minimum angle of resolution chart. The WHOICD-10 category 2 or greater defines visual impairment as worse than 20/200, taken as less than 4 letters on the logarithm of the minimum angle of resolution chart.25 The criterion for blind pension eligibility in Australia differs from that in the United States (20/200 or worse in both eyes) by
Figure 2. Level of visual impairment (20/40 –⬍20/70, ⬍20/70 –20/200, and ⬍20/200) among persons with incident visual impairment ⬍20/40 in either eye, by age at baseline in years, and gender.
43
Ophthalmology Volume 110, Number 1, January 2003 Table 3. Crude 5-year Incidence of Any and Severe Visual Impairment Monocularly and in the Better Eye by Age at Baseline and Gender in the Blue Mountains Eye Study Compared with Crude and Standardized Incidence in the Beaver Dam Eye Study Any Visual Impairment
Blue Mountains Eye Study
Age (Years) Monocular Women ⬍55 55–64 65–74 75⫹ Total Men ⬍55 55–64 65–74 75⫹ Total Both ⬍55 55–64 65–74 75⫹ Total Better eye Women ⬍55 55–64 65–74 75⫹ Total Men ⬍55 55–64 65–74 75⫹ Total Both ⬍55 55–64 65–74 75⫹ Total
At Risk
n
178 0 460 20 435 45 144 39 1217 104 146 354 310 94 904
Crude %
0.0 4.3 10.3 27.1
1 6 22 17 46
0.7 1.7 7.1 18.1
324 1 814 26 745 67 238 56 2121 150
0.3 3.2 9.0 23.5
181 477 467 186 1311
0 11 35 44 90
0.0 2.3 7.5 23.7
150 367 339 121 977
0 2 19 24 45
0.0 0.5 5.6 19.8
331 0 844 13 806 54 307 68 2288 135
0.0 1.5 6.7 22.1
95% Confidence Interval
Beaver Dam Eye Study At Risk
Crude %
626 507 426 126 8.5 (7.0, 10.1) 1685
1.3 3.0 11.5 27.0 6.3
5.1 (3.7, 6.5)
545 440 321 78 1384
0.7 2.3 10.0 24.4 4.7
7.1 (6.0, 8.2)
1171 947 747 204 3069
1.0 2.6 10.8 26.0 5.6
6.9 (5.5, 8.2)
656 551 521 208 1936
0.2 1.3 3.7 19.7 3.5
4.6 (3.3, 5.9)
574 471 373 110 1528
0.4 1.3 2.7 14.6 2.2
5.9 (4.9, 6.9)
1230 1022 894 318 3464
0.2 1.3 3.2 17.9 2.9
Severe Visual Impairment Standardized to Blue Mountains Eye Study 95% Confidence Interval
At Risk
n
Crude %
0.0 0.8 1.5 10.5
8.6 (7.3, 10.0)
183 0 475 4 466 7 190 20 1314 31
0.0 0.3 1.5 5.6
7.0 (5.7, 8.3)
148 0 361 1 330 5 124 7 963 13
0.0 0.6 1.5 8.6
7.9 (6.9, 8.8)
331 0 836 5 796 12 314 27 2277 44
0.0 0.0 0.8 6.5
4.6 (3.7, 5.5)
183 0 479 0 474 4 201 13 1337 17
3.3 (2.4, 4.2)
4.0 (3.4, 4.7)
requiring Snellen vision in the better eye of worse than 20/200 or a functional visual field of less than 10°. No participants in BMES I or II were eligible using the visual field criterion. In keeping with the BDES, deterioration or doubling of the visual angle was defined as a loss of 15 letters or more read correctly at follow-up compared with baseline. Similarly, improvement or halving of the visual angle was defined as an increase of 15 letters or more read correctly at follow-up compared with baseline. An eye was considered to be at risk of deterioration if at baseline it had perception of light or better, whereas an eye capable of improvement at baseline had read fewer than 56 letters. Similar to monocular and binocular visual impairment, persons were considered to have monocular or binocular deterioration or improvement. Incidence rates for visual impairment and doubling of the visual angle were determined for the better eye, whereas the
44
Blue Mountains Eye Study
150 367 342 128 987
0 0 2 5 7
0.0 0.0 0.6 3.9
333 0 846 0 816 6 329 18 2324 24
0.0 0.0 0.7 5.5
Beaver Dam Eye Study
Standardized to Blue Mountains Eye Study
95% Confidence Interval
At Risk
2.4 (1.5, 3.2)
648 533 515 224 1920
0.0 0.6 1.4 7.1 1.4
1.7 (1.2, 2.3)
1.3 (0.6, 2.1)
560 456 358 115 1489
0.0 0.0 1.1 4.4 0.6
0.9 (0.5, 1.4)
1.9 (1.4, 2.5)
1208 989 873 339 3409
0.0 0.3 1.3 6.2 1.0
1.4 (1.0, 1.8)
1.3 (0.7, 1.9)
662 553 544 249 2008
0.0 0.2 0.0 3.2 0.5
0.6 (0.2, 0.9)
0.7 (0.2, 1.2)
576 474 381 119 1550
0.0 0.0 0.0 0.8 0.1
0.1 (⫺0.1, 0.3)
1.0 (0.6, 1.4)
1238 1027 925 368 3558
0.0 0.1 0.0 2.5 0.3
0.4 (0.2, 0.6)
Crude 95% Confidence % Interval
incidence of halving of the visual angle was determined for the worse eye. In cases in which the right and left eye read the same number of letters, they were considered to be both the worse and the better eye.
Data Handling and Statistical Analysis Data were entered into dBase (Borland International Inc, Scotts Valley, CA) for BMES I and Microsoft Access (Microsoft Corporation, Redmond, WA) for BMES II. Statistical Analysis System (SAS Institute Inc, Cary, NC) was used for analyses, including t test, chi-square test, logistic regression, and generalized linear model analyses. Generalized linear model analyses adjusting for age and gender were used to compare right and left eyes, as well as men and women. BDES incidence rates were standardized to the BMES follow-up population to aid comparability. Confidence
Foran et al 䡠 Five-Year Change in Visual Acuity Table 4. Comparison of Different Levels of Binocular and Monocular Incident Visual Impairment by Age at Baseline and Gender <20/40
Binocular Women ⬍55 55–64 65–74 75⫹ Total Men ⬍55 55–64 65–74 75⫹ Total Both ⬍55 55–64 65–74 75⫹ Total Monocular Women ⬍55 55–64 65–74 75⫹ Total Men ⬍55 55–64 65–74 75⫹ Total Both ⬍55 55–64 65–74 75⫹ Total
<20/40
<20/200
<20/70
<20/200
At Risk
Incident
%
At Risk
Incident
%
At Risk
Incident
%
At Risk
Incident
%
At Risk
Incident
%
178 460 435 144 1217
0 4 8 17 29
0.0 0.9 1.8 11.8 2.4
179 462 440 147 1228
0 2 8 16 26
0.0 0.4 1.8 10.9 2.1
182 472 455 173 1282
0 0 2 8 10
0.0 0.0 0.4 4.6 0.8
183 475 466 190 1314
0 0 1 2 3
0.0 0.0 0.2 1.1 0.2
183 475 468 192 1318
0 0 1 1 2
0.0 0.0 0.2 0.5 0.2
146 354 310 94 904
0 0 3 9 12
0.0 0.0 1.0 9.6 1.3
146 356 313 99 914
0 0 4 7 11
0.0 0.0 1.3 7.1 1.2
148 360 325 116 949
0 0 2 3 5
0.0 0.0 0.6 2.6 0.5
148 361 330 124 963
0 0 0 0 0
0.0 0.0 0.0 0.0 0.0
148 362 331 125 966
0 0 0 0 0
0.0 0.0 0.0 0.0 0.0
324 814 745 238 2121
0 4 11 26 41
0.0 0.5 1.5 10.9 1.9
325 818 753 246 2142
0 2 12 23 37
0.0 0.2 1.6 9.3 1.7
330 832 780 289 2231
0 0 4 11 15
0.0 0.0 0.5 3.8 0.7
331 836 796 314 2277
0 0 1 2 3
0.0 0.0 0.1 0.6 0.1
331 837 799 317 2284
0 0 1 1 2
0.0 0.0 0.1 0.3 0.1
178 460 435 144 1217
0 20 45 39 104
0.0 4.3 10.3 27.1 8.5
179 462 440 147 1228
0 17 37 33 87
0.0 3.7 8.4 22.4 7.1
182 472 455 173 1282
0 11 16 29 56
0.0 2.3 3.5 16.8 4.4
183 475 466 190 1314
0 4 7 20 31
0.0 0.8 1.5 10.5 2.4
183 475 468 192 1318
0 3 8 19 30
0.0 0.6 1.7 9.9 2.3
146 354 310 94 904
1 6 22 17 46
0.7 1.7 7.1 18.1 5.1
146 356 313 99 914
1 7 21 18 47
0.7 2.0 6.7 18.2 5.1
148 360 325 116 949
0 5 12 11 28
0.0 1.4 3.7 9.5 3.0
148 361 330 124 963
0 1 5 7 13
0.0 0.3 1.5 5.6 1.3
148 362 331 125 966
0 1 5 8 14
0.0 0.3 1.5 6.4 1.4
324 814 745 238 2121
1 26 67 56 150
0.3 3.2 9.0 23.5 7.1
325 818 753 246 2142
1 24 58 51 134
0.3 2.9 7.7 20.7 6.3
330 832 780 289 2231
0 16 28 40 84
0.0 1.9 3.6 13.8 3.8
331 836 796 314 2277
0 5 12 27 44
0.0 0.6 1.5 8.6 1.9
331 837 799 317 2284
0 4 13 27 44
0.0 0.5 1.6 8.5 1.9
Best corrected visual acuity shown.
intervals were calculated at 95%. Means are given with their standard deviation and mean differences with their standard error.
Results The mean age of participants at baseline was 64.5 years. The mean follow-up period for the 2335 BMES II participants was 5.1 years (minimum, 3.0 years; maximum, 7.8 years). Women accounted for 57.5% of subjects. Table 1 shows the baseline characteristics of participants.
eyes P ⬍ 0.0001). There was an age-related decrease in the numbers of correctly read letters in both eyes. In the right eye, this decrease varied from ⫺0.5 ⫾ 3.4 letters in persons aged less than 55 years, ⫺1.5 ⫾ 5.9 in persons aged 55 to 64 years, ⫺3.2 ⫾ 8.3 in persons aged 65 to 74 years, to ⫺6.3 ⫾ 13.2 in persons aged 75 years or older. Figure 1 shows the mean change in number of correctly read letters in the right eye by age and gender. Age categorized into four levels (⬍55, 55– 64, 65–74 and ⬎75 years) was a significant determinant of decrease in visual acuity in both eyes (P ⬍ 0.0001). Women had a greater change in the number of letters read incorrectly than men in the worse eye (age-adjusted means ⫺3.1 ⫾ 0.2 vs. ⫺2.6 ⫾ 0.3, P ⫽ 0.2), and in the better eye (age-adjusted means ⫺2.4 ⫾ 0.2 vs. ⫺2.1 ⫾ 0.2, P ⫽ 0.1).
Change in Visual Acuity The change in numbers of letters read correctly (5-year follow-up compared with baseline examinations) in the right eye was ⫺2.6 ⫾ 8.2 (mean ⫾ standard deviation) and ⫺2.5 ⫾ 7.2 in the left eye. After age and gender adjustment, the difference between right and left eyes in mean number of letters read correctly was not statistically significant (P ⫽ 0.53). The numbers of letters correctly read by right and left eyes were not significantly different (Pearson correlation between
Presenting versus Best-Corrected Visual Acuity Compared with best-corrected visual acuity, presenting visual acuity (with current glasses if worn) was worse by approximately 1 line at both the baseline and follow-up examinations. At baseline, right eyes read 45.5 ⫾ 12.8 letters correctly at presentation compared with 51.5 ⫾ 11.1 after best subjective refraction. At followup, presenting and best-corrected vision were 44.4 ⫾ 13.1 and 50.5
45
Ophthalmology Volume 110, Number 1, January 2003 Table 5. Crude 5-year Incidence Doubling and Halving of the Visual Angle by Age at Baseline and Gender in the Blue Mountains Doubling of the Visual Angle
Beaver Dam Eye Study
Standard Blue Mountains Eye Study
95% Confidence Interval
At Risk
Crude %
95% Confidence Interval
6.7 (5.3, 8.0)
650 537 521 229 1937
0.8 2.8 6.0 14.4 4.3
5.4 (4.4, 6.4)
6.1 (4.6, 7.6)
565 461 361 113 1500
1.1 2.2 7.8 11.5 3.8
5.2 (4.1, 6.3)
6.4 (5.4, 7.4)
1215 998 882 342 3437
0.9 2.5 6.7 13.5 4.1
5.3 (4.6, 6.1)
1.7 (1.0, 2.4)
663 553 545 255 2016
0.2 1.5 1.5 8.6 1.9
2.4 (1.7, 3.1)
0.8 (0.3, 1.4)
576 474 381 120 1551
0.0 0.6 2.1 8.3 1.4
2.0 (1.3, 2.7)
1.3 (0.9, 1.8)
1239 1027 926 375 3567
0.1 1.1 1.7 8.5 1.7
2.2 (1.7, 2.7)
Blue Mountains Eye Study
Age (Years) Monocular Women ⬍55 55–64 65–74 75⫹ Total Men ⬍55 55–64 65–74 75⫹ Total Both ⬍55 55–64 65–74 75⫹ Total Better eye Women ⬍55 55–64 65–74 75⫹ Total Men ⬍55 55–64 65–74 75⫹ Total Both ⬍55 55–64 65–74 75⫹ Total
At Risk
Crude %
n
183 479 473 202 1337
1 17 37 34 89
0.5 3.5 7.8 16.8
150 365 341 127 983
1 6 32 21 60
0.7 1.6 9.4 16.5
333 844 814 329 2320
2 23 69 55 149
0.6 2.7 8.5 16.7
183 479 474 204 1340
0 1 7 15 23
0.0 0.2 1.5 7.4
150 367 343 127 987
0 1 3 4 8
0.0 0.3 0.9 3.1
333 846 817 331 2327
0 2 10 19 31
0.0 0.2 1.2 5.7
Incidences of doubling of the visual angle given monocularly and in the better eye, halving of the visual angle given monocularly and in the worse eye.
⫾ 11.7 letters, respectively. In both right and left eyes, the mean differences in best-corrected and presenting vision at baseline and at follow-up were not statistically different. For instance, in the right eye at baseline the mean improvement in best-corrected vision over presenting vision was 6.0 ⫾ 7.4 letters read correctly compared with 6.2 ⫾ 7.2 letters at follow-up.
Incidence of Visual Impairment Any Visual Impairment (<20/40). Five-year incident visual impairment occurred in 267 eyes: 137 (51.3%) right compared with 130 (48.6%) left eyes (P ⫽ 0.51). Figure 2 shows the incidence of visual impairment in either eye by level of impairment. Incident binocular impairment occurred in 41 persons (1.9%, 95% confi-
46
dence interval [CI] 1.3–2.5) and monocular impairment in 150 persons (7.1%, 95% CI 6.0 – 8.2) (Table 3). Women accounted for approximately 70% of binocular and monocular impairment compared with men. After controlling for gender, each year of increasing age was associated with roughly a 20% increased risk of incident binocular visual impairment (P ⫽ 0.0001, gender-adjusted odds ratio [OR] 1.21, 95% CI 1.15–1.26). After adjusting for age, women had a 73% increased risk of incident binocular impairment compared with men, but this did not reach statistical significance (P ⫽ 0.1). Similarly, after controlling for gender, each year of increasing age was associated with a 13% significantly increased risk for incident monocular impairment (P ⫽ 0.0001, gender-adjusted OR 1.13, 95% CI 1.10 –1.15). After adjusting for
Foran et al 䡠 Five-Year Change in Visual Acuity Eye Study Compared with the Crude and Standardized Incidence in the Beaver Dam Eye Study Halving of the Visual Angle
Blue Mountains Eye Study Crude %
At risk
n
48 151 259 181 637
2 3 4 8 17
4.2 2.0 1.5 4.4
28 74 149 105 356
0 0 2 5 7
0.0 0.0 1.3 4.8
76 225 408 286 995 Worse eye
2 3 6 13 24
2.6 1.3 1.5 4.5
98 317 385 198 998
2 3 6 7 18
2.0 0.9 1.6 3.5
61 191 247 120 619
0 0 6 6 12
0.0 0.0 2.4 5.0
159 508 632 318 1617
2 3 12 13 30
1.3 0.6 1.9 4.1
Beaver Dam Eye Study
Standard Blue Mountains Eye Study
95% Confidence Interval
At Risk
Crude %
95% Confidence Interval
2.7 (1.4, 3.9)
153 197 294 188 832
4.6 4.6 7.5 4.8 5.7
5.8 (4.2, 7.4)
2.0 (0.5, 3.9)
60 76 121 80 337
3.3 0 5 5 3.6
3.8 (1.8, 5.9)
2.4 (1.5, 3.4)
213 273 415 268 1169
4.2 3.3 6.8 4.9 5.1
5.3 (4.0, 6.5)
1.8 (⫺0.1, 0.7)
329 397 455 245 1426
1.5 1.8 4.2 4.1 2.9
3.2 (2.2, 4.1)
1.9 (0.0, 2.2)
202 217 263 108 790
1.0 0.0 1.9 4.6 1.5
1.7 (0.8, 2.7)
1.9 (0.1, 1.1)
531 614 718 353 2216
1.3 1.1 3.3 4.3 2.4
2.6 (1.9, 3.3)
age, women had a 70% significantly increased risk of incident monocular impairment compared with men (P ⫽ 0.005, ageadjusted OR 1.70, 95% CI 1.18 –2.47). Visual Impairment <20/40. Using the visual impairment definition of ⬍20/40 (legal driving cutpoint), 236 eyes (5.1%) developed incident visual impairment in 127 (53.8%) right and 109 (46.2%) left eyes (P ⫽ 0.2). Binocular impairment ⬍20/40 occurred in 37 persons (1.7%), 4 persons less than for binocular impairment ⱕ20/40. Monocular impairment ⬍20/40 occurred in 134 persons (6.3%), 16 persons less than for monocular impairment ⱕ20/40. For incident binocular impairment ⬍20/40, 70.3% of persons were women and 29.7% were men. For incident monocular impairment 64.9% of persons were women and 35.1% were men. Although women had higher incidence rates than men, these
differences were not significant after controlling for age. For each year of increasing age, the incidence of binocular impairment increased by 22% (gender-adjusted OR 1.22, 95% CI 1.16 –1.29), and the incidence of monocular impairment increased by 12% (gender-adjusted OR 1.12, 95% CI 1.09 –1.15). Table 4 shows the incidence of binocular and monocular impairment using different classifications by age and gender. Visual Impairment <20/70. Using the WHO-ICD-10 definition of ⬍20/70 (category 1 or worse) to define visual impairment, 124 eyes (2.7%) developed incident visual impairment in 72 (58.1%) right and 52 (41.9%) left eyes (P ⫽ 0.06). Binocular impairment occurred in 15 persons (0.7%), of whom 10 (66.7%) were women and 5 (33.3%) were men; this was not statistically significant after controlling for age (P ⫽ 0.6). Monocular impair-
47
Ophthalmology Volume 110, Number 1, January 2003
Figure 3. Proportionate causes of incident monocular visual impairment, severe visual impairment, and doubling of the visual angle by age at baseline in years. ARM ⫽ Age-related maculopathy.
ment developed in 84 persons (3.8%), with similar gender differences. Age (per year) was the significant determinant of both binocular (P ⫽ 0.0001, gender-adjusted OR 1.23, 95% CI 1.14 – 1.32) and monocular impairment (P ⫽ ⬍0.001, gender-adjusted OR 1.13, 95% CI 1.10 –1.7). Severe Visual Impairment <20/200. Five-year incident visual impairment ⱕ20/200 developed in 52 eyes (1.1%): 34 right (65.4%) and 18 left eyes (34.6%). Right eyes were twice as likely to develop severe impairment than left eyes after controlling for age and gender (P ⫽ 0.02, age-gender adjusted OR 1.99, CI 1.11–3.56). Three persons (0.1%, 95% CI 0.0 – 0.3) had incident severe binocular incident impairment, all of whom were women, aged 72, 78, and 80 at baseline. Forty-four persons (1.9%, 95% CI 1.4 –2.5) had incident severe monocular impairment in 29 (65.9%) right eyes and 15 (34.1%) left eyes. Of this group, 31 (70.4%) were women and 13 (29.6%) were men. Each year of increasing age was associated with a 15% increased risk of incident severe monocular visual impairment (P ⫽ ⬍0.001, gender-adjusted OR 1.14, 95% CI 1.10 –1.19) after controlling for gender. However, female gender was not associated with a significantly increased risk of incident visual impairment (P ⫽ 0.2, age-adjusted OR 1.63, 95% CI 0.0.84 –3.17). Severe Visual Impairment <20/200. Five-year incident visual impairment ⬍20/200 occurred in 50 eyes (1.1%): 33 right (66.7%) compared with 17 left eyes (33.3%). Similar to severe visual impairment ⬍20/200, right eyes were twice as likely to develop severe impairment than left eyes after controlling for age and gender (P ⫽ 0.02, age-gender adjusted OR 2.03, 95% CI 1.12– 3.69). Two persons (0.1%, 95% CI 0.0 – 0.2) had incident severe binocular incident impairment; both were women, aged 72 and 78. Similar to the incidence of monocular impairment ⱕ20/200, 44 persons (1.9%, 95% CI 1.4 –2.5) had incident severe monocular impairment in 29 (65.9%) right eyes and 15 (34.1%) left eyes. Of this group, 30 (68.2%) were women and 14 (31.8%) were men. Again, each year of increasing age was associated with a 15% increased risk of incident severe monocular visual impairment (P ⫽ 0.0001, gender-adjusted OR 1.15, 95% CI 1.11–1.20) after controlling for gender. However, female gender was not associated with a significantly increased risk of incident visual impairment (P ⫽ 0.3, age-adjusted OR 1.45, 95% CI 0.75–2.78). Doubling of the Visual Angle (Deterioration). The doubling
48
of the visual angle (a decrease in 15 or more letters read correctly) that occurred binocularly in 31 persons (1.3%, 95% CI 0.9 –1.8) was age related: none was aged ⬍55 years, 2 (0.2%) were aged 55 to 64 years, 10 (1.2%) were aged 65 to 74 years, and 19 persons (5.8%) were aged 75 years. Of the 149 persons (6.4%, 95% CI 5.4 –7.4) with monocular doubling of the visual angle, 2 persons (0.6%) were aged ⬍55 years, 23 (2.7%) 55 to 64 years, 69 (8.5%) 65 to 74 years, and 55 persons (16.7%) aged 75 years and older (Table 5). Monocular doubling of the visual angle was significantly associated with age (P ⫽ 0.0001, OR 1.10, 95% CI 1.08 –1.12) but not with gender (P ⫽ 0.9, age-adjusted OR 1.01, 95% CI 0.72–1.44). Similar associations were found with bilateral doubling of the visual angle. Halving of the Visual Angle (Improvement). Halving of the visual angle (an increase in 15 or more letters read correctly) occurred in both eyes in 5 persons (0.5%, 95% CI 0.1– 0.9) and in one eye in 24 persons (2.4%, 95% CI 1.5–3.4) (Table 5). Persons with improvement were more likely to have had incident cataract surgery, that is, cataract surgery performed after the baseline examination but before the follow-up examination. Of the 24 persons with monocular improvement, 14 (58.3%) had incident cataract surgery in that eye (P ⫽ ⬍0.001). After controlling for age and gender, monocular improvement was strongly associated with incident cataract surgery in that eye (P ⫽ ⬍0.001, ageadjusted OR 9.6, 95% CI 4.0 –22.9). For both binocular and monocular improvement, age and gender were not statistically significant (P ⬎ 0.1). Attributable Causes. The most frequent primary cause of any incident visual impairment was cataract, whereas for severe incident visual impairment the most frequent primary cause was age-related maculopathy. Figure 3 shows the proportion of attributable causes for any and severe visual impairment and for doubling of the visual angle.
Discussion Incident figures from this report can be used to project the number of incident cases of visual impairment within the Australian population. By using population estimates for the year 2001 from the Australian Bureau of Statistics26 with our incident visual impairment figures, we estimate that in 5 years approximately 94,700 persons will develop visual impairment ⬍20/40 in both eyes; these persons will no longer be able to obtain a driver’s license based on the current visual acuity criterion. This report from the BMES supplements data from the BDES on the change in visual acuity among members of an older population.13 Previous reports from these two studies13,21 have indicated similar findings for the prevalence of visual impairment in these similar, older, largely white communities in the United States and Australia. The mean number of letters read correctly after refraction was almost identical between these two studies within the same age and gender strata. The BMES follow-up examined 75.1% of survivors from the initial study. We may have underestimated the decrease in visual acuity or the incidence of visual impairment, because persons returning were younger, healthier, and better educated than those who had died since BMES I (Table 1). Persons who were alive but did not participate in the follow-up were similar in many parameters to those who did
Foran et al 䡠 Five-Year Change in Visual Acuity participate. They were significantly less likely to own their home or to have gained further qualifications after leaving school. They had a significantly higher self-reported history of diabetes and also were more likely to be current smokers. In keeping with this, they had more visual impairment and late age-related maculopathy at baseline. As a result, we may have underestimated the incidence of visual impairment, especially severe impairment. In Western populations, smoking is the strongest modifiable risk factor for the development of late-stage age-related maculopathy.27–29 Although there is no absolute visual acuity threshold, in Australia, cataract surgery is usually not performed unless visual acuity is worse than 20/40. Persons who were not visually impaired at baseline, but who had cataract surgery between the baseline and follow-up examinations, could also potentially have been classified as incident cases of visual impairment. If these persons had been included, the incidence of visual impairment would be higher. Among those at risk for monocular visual impairment ⱕ20/40 who did not develop incident impairment, 30 persons had binocular cataract surgery and 45 persons had monocular surgery. Including these persons would have increased the incidence of binocular impairment from 1.9% to 3.3% (74% higher) and monocular impairment from 7.1% to 9.2% (30% higher). Women generally had higher rates of impairment than men even after adjusting for age. This may be partially because women have a higher prevalence of cataract, especially cortical cataract, than men30 and that women are also less likely to have cataract surgery. We also found that right eyes were more likely to be impaired than left eyes and that there was a significant difference for severe impairment ⱕ20/200 and ⬍20/200. This could be due to measurement bias; to learning by the participant, because right eyes were systematically tested first; or to chance. We also found substantially higher rates of incident binocular impairment than expected if the incident monocular rates for each eye were independent. This confirms the symmetry of common eye diseases particularly cataract and age-related maculopathy. Compared with the BDES standardized rate (7.9%), the BMES found a lower rate of any monocular impairment ⱕ20/40 overall (7.1%), although confidence intervals were overlapping (Table 3). This difference was most marked for men (5.1% BMES versus 7.0% BDES). In contrast, we found higher incidences of any impairment in the better eye (nonoverlapping confidence intervals overall), monocular severe visual impairment, and better eye severe impairment (overlapping confidence intervals). Similar to any impairment, our study found a slightly higher incidence of monocular deterioration (6.4% BMES vs. 5.3% BDES) and lower incidence of deterioration in the better eye (1.3% BMES vs. 2.2% BDES). Consistent with this, improvement rates in the BMES were lower in the monocular (2.4% BMES vs. 5.3% BDES) and better eye (1.9% BMES vs. 2.6% BDES) compared with the BDES. We could not compare rates for binocular impairment, deterioration, or halving with those from the BDES, because these data were not reported.13 The BDES found a strong relationship between improvement of visual acuity and cataract surgery.
Incidence rates for cataract surgery, however, differed substantially between these two populations. For instance, the BDES age-adjusted incidence rate for cataract surgery31 in persons with any cortical cataract at baseline was 32.0% compared with the BMES age-adjusted rate of 8.9% (Invest Ophthalmol Vis Sci 2001;42[Suppl]:S508). Our finding that the mean differences in best-corrected and presenting vision at baseline and at 5-year follow-up were similar suggests that the study may not have greatly influenced the likelihood that people would seek attention for under or uncorrected refractive error. All participants whose visual acuity corrected by 2 or more lines (10⫹ letters) were advised that their vision could be improved by a change in their current glasses or by wearing distance glasses, if not previously worn. The study region has ready access to optometric and ophthalmic services. If the study had altered behavior, its effects may have been partly balanced by the age-related decline in visual acuity and the development of cataract and other conditions. In summary, our study found that increases in visual impairment were strongly age related. Women were more likely to have incident impairment than men, even after adjusting for age. Our rates were similar to those reported by the BDES. Population-based incident data are vital to health planners determining the allocation of health resources. The 10-year incidence rates of visual impairment will provide further important data. Attributable causes of incident impairment will also be necessary to determine the number of cases that could be prevented or treated.
References 1. Tielsch JM, Sommer A, Witt K, et al. Blindness and visual impairment in an American urban population. The Baltimore Eye Survey [see comments] Arch Ophthalmol 1990;108:286 –90. 2. Casson R, Giles L, Newland HS. Prevalence of blindness and visual impairment in an elderly urban population. Aust N Z J Ophthalmol 1996;24:239 – 43. 3. Taylor HR, Livingston PM, Stanislavsky YL, McCarty CA. 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 –37. 4. Wang JJ, Foran S, Mitchell P. Age-specific prevalence and causes of bilateral and unilateral visual impairment in older Australians: the Blue Mountains Eye Study. Clin Exp Ophthalmol 2000;28:268 –73. 5. Rahmani B, Tielsch JM, Katz J, et al. The cause-specific prevalence of visual impairment in an urban population. The Baltimore Eye Survey. Ophthalmology 1996;103:1721– 6. 6. Mun˜ oz B, West SK, Rubin GS, et al. Causes of blindness and visual impairment in a population of older Americans: the Salisbury Eye Evaluation Study. Arch Ophthalmol 2000;118: 819 –25. 7. Klaver CCW, Wolfs RCW, Vingerling JR, et al. Age-specific prevalence and causes of blindness and visual impairment in an older population. The Rotterdam Study. Arch Ophthalmol 1998;116:653– 8. 8. Hirvela¨ H, Laatikainen L. Visual acuity in a population aged 70 years or older; prevalence and causes of visual impairment. Acta Ophthalmol Scand 1995;73:99 –104. 9. Dana MR, Tielsch JM, Enger C, et al. Visual impairment in a
49
Ophthalmology Volume 110, Number 1, January 2003
10. 11. 12. 13. 14.
15. 16. 17. 18. 19.
rural Appalachian community: prevalence and causes. JAMA 1990;264:2400 –5. Ponte F, Giuffre` G, Giammanco R. Prevalence and causes of blindness and low vision in the Casteldaccia Eye Study. Graefes Arch Clin Exp Ophthalmol 1994;232:469 –72. Newland HS, Hiller JE, Casson RJ, Obermeder S. Prevalence and causes of blindness in the South Australian population aged 50 and over. Ophthalmic Epidemiol 1996;3:97–107. Weih LM, VanNewkirk MR, McCarty CA, Taylor HR. Agespecific causes of bilateral visual impairment. Arch Ophthalmol 2000;118:264 –9. Klein R, Klein BEK, Lee KE. Changes in visual acuity in a population. The Beaver Dam Eye Study. Ophthalmology 1996;103:1169 –78. Wang JJ, Mitchell P, Smith W, et al. Impact of visual impairment on use of community support services by elderly persons: the Blue Mountains Eye Study. Invest Ophthalmol Vis Sci 1999;40:12–9. Mitchell P, Hayes P, Wang JJ. Visual impairment in nursing home residents: the Blue Mountains Eye Study. Med J Aust 1997;166:73– 6. Ivers RQ, Cumming RG, Mitchell P, Attebo K. Visual impairment and falls in older adults: the Blue Mountains Eye Study. J Am Geriatr Soc 1998;46:58 – 64. Campbell AJ, Reinken J, Allan BC, Martinez GS. Falls in old age: a study of frequency and related clinical factors. Age Ageing 1981;10:264 –70. McCarty CA, Nanjan MB, Taylor HR. Vision impairment predicts 5 year mortality. Br J Ophthalmol 2001;85:322– 6. Wang JJ, Mitchell P, Simpson JM, et al. Visual impairment, age-related cataract and mortality. Arch Ophthalmol 2001; 119:1186 –90.
20. Wang JJ, Mitchell P, Smith W. Vision and low self-rated health: the Blue Mountains Eye Study. Invest Ophthalmol Vis Sci 2000;41:49 –54. 21. Attebo K, Mitchell P, Smith W. Visual acuity and the causes of visual loss in Australia. The Blue Mountains Eye Study. Ophthalmology 1996;103:357– 64. 22. Mitchell P, Smith W, Attebo K, Wang JJ. Prevalence of age-related maculopathy in Australia. The Blue Mountains Eye Study. Ophthalmology 1995;102:1450 – 60. 23. Mitchell P, Smith W, Attebo K, Healey PR. Prevalence of open-angle glaucoma in Australia. The Blue Mountains Eye Study. Ophthalmology 1996;103:1661–9. 24. Klein R, Klein BEK, Linton KLP, De Mets DL. The Beaver Dam Eye Study: visual acuity. Ophthalmology 1991;98:1310 –5. 25. ICD 10 Definition. Geneva: World Health Organization. Available from URL: www.who.int/pbd/pbl/img/icd10.gif. 26. Population Projections: 1997 to 2051. Canberra: Australian Bureau of Statistics, 1998;1–130. No. 3222.0. 27. Smith W, Mitchell P, Leeder SR. Smoking and age-related maculopathy: the Blue Mountains Eye Study. Arch Ophthalmol 1996;114:1518 –23. 28. Mitchell P, Chapman S, Smith W. “Smoking is a major cause of blindness:” A new cigarette pack warning. [editorial] Med J Aust 1999;171:173– 4. 29. Smith W, Assink J, Klein R, et al. Risk factors for age-related macular degeneration. Pooled findings from three continents. Ophthalmology 2001;108:697–704. 30. Mitchell P, Cumming RG, Attebo K, Panchapakesan J. Prevalence of cataract in Australia. The Blue Mountains Eye Study. Ophthalmology 1997;104:581– 8. 31. Klein BEK, Klein R, Moss SE. Incident cataract surgery. The Beaver Dam Eye Study. Ophthalmology 1997;104:573– 80.
(Questions on p. 22).
1. 2. 3. 4. 5.
B B D B C
Answers:
50