The Beaver Dam Eye Study: Visual Acuity RONALD KLEIN, MD, MPH, l BARBARA E. K. KLEIN, MD, MPH, l KATHRYN L. P. LINTON, MS,t DAVID L. De METS, PhD2
Abstract: Few current population-based data on visual impairment are available. Visual acuity and impairment were measured in 4926 people between the ages of 43 and 86 years in the defined population participating in the Beaver Dam Eye Study. Visual acuity was measured after refraction, using standardized protocols. Of a possible maximum score of 70 (20/10), the mean number of letters correctly identified (right eye) varied from 55.7 (20/20, n = 1515) in people between the ages of 43 and 54 years to 41.2 (20/40, n = 795) in people 75 years of age or older. Age-specific mean visual acuity scores were consistently and significantly lower in women, who identified three fewer letters on the average than men. Rates of any visual impairment (20/40 or worse in the better eye) or legal blindness (20/200 or worse in the better eye), increased from 0.8% and 0.1 %, respectively, in people between the ages of 43 and 54 years to 21.1 % and 2.0%, respectively, in people 75 years of age or older. Multivariate analyses showed both sex (women) and age (older) to be significant and independent predictors of poorer visual acuity. Ophthalmology 1991; 98: 131 0-1315
Visual acuity is an important measure of visual function. However there are few current population-based estimates of best-corrected visual acuity available for the United States. 1.2 Such data are useful for estimating needs for medical care, counseling and rehabilitative services, projecting costs, measuring temporal trends, developing causal insights, and measuring quality of life. Thus, we have assessed visual acuity in a large geographically defined adult population and described the findings in this report.
MATERIALS AND METHODS THE POPULATION
Beaver Dam, WI, was selected because it had an excellent response rate in a previous population-based study Originally received: December 19, 1990. Revision accepted: March 5, 1991. 1
2
Department of Ophthalmology, University of Wisconsin, Madison. Department of Statistics, University of Wisconsin, Madison.
Supported by grant U10 EY06594 (Drs. Ronald Klein and Barbara E. K. Klein) from the National Eye Institute/NIH, Bethesda, Maryland. Reprint requests to Ronald Klein, MD, MPH, Department of Ophthalmology, Clinical Science Center, 600 Highland Ave, Madison, WI 53792.
1310
of diabetic retinopathy (96.0%). 3 It also had an ideal-sized population (n = 17,179, 1980 census), which was relatively stable (outmigration in Dodge County between 1975 and 1980 was 0.32% per year).4 From September 15, 1987 to May 4, 1988, a private census of the population living in the city and township of Beaver Dam, Wisconsin, was performed by the University of Wisconsin Extension-Survey Research Laboratory. Using procedures described elsewhere, the Survey Research Laboratory provided information on 6654 Beaver Dam households. 5,6 Of these, there were 42 households from which there was duplicate information and the second set of data was therefore omitted. There was a small percentage (1.3%) of the 6612 unique households for which there were incomplete surveys, either because the residents could not be reached (0.4%) or they refused (0.9%). The census interview included questions on demographic characteristics, ocular and systemic conditions, and primary eye care providers. Data from the census have been reported elsewhere. 6 Pertinent parts of the census questionnaire included questions regarding distance vision ("Do you see well enough with glasses or contact lenses to recognize a friend across the street"), near vision ("Do you see well enough with or without glasses to recognize letters in newspaper stories"), and a self-reported medical history of cataract, age-related macular degeneration, glaucoma, diabetes mellitus, or hypertension, and
KLEIN et al
•
BEAVER DAM EYE STUDY
Table 1. Selected Characteristics of Participants and Nonparticipants in the Beaver Dam Eye Study (1988 to 1990) Men Characteristic* Number Age (yrs) Mean SO History of poor distance vision (%) History of poor near vision (%) History of presence of cataract (%) History of presence of age-related macular degeneration (%) History of glaucoma (%) History of diabetes mellitus (%) History of hypertension (%) History of being seen by an eye doctor within 2 years of census
Participant
Women
Nonparticipant
2166
497
Participant 2760
All
Nonparticipant 502
Participant 4926
Nonparticipant 999
59.5 10.9 1.7 4.0 7.8
62.4t 12.4 3.4:j: 5.2 10.9§
61.3 11.5 3.0 5.1 12.0
64.6 12.4 4.6 6.6 17.7t
60.6 11.3 2.4 4.7 10.1
63.5:j: 12.4 4.0:j: 5.9 14.3t
2.9 2.9 6.8 35.4
1.6 5.0§ 7.7 33.2
4.4 3.9 6.8 37.3
5.2 5.6 7.2 39.6
3.7 3.4 6.8 36.4
3.4 5.3§ 7.4 36.4
74.8
68.8:j:
78.2
68.1t
76.7
68.5t
SO = standard deviation. * As determined from responses to questionnaire census. t P< 0.001. :j: P< 0.01. § P< 0.05.
a history of ever having seen an eye doctor. This information was elicited about all household members. Eligibility requirements for entry into the study included living in the city or township of Beaver Dam and being 43 to 84 years of age at the time of the census. Of the 6612 households identified by the census, 3715 contained at least one person satisfying the age criteria. These households reported a total of 5833 individuals between the ages of 43 and 84 years at the time of the census. After completion of the census, 76 additional households with a total of 92 people were identified and are included in the population. The data presented here are based on a total of 5833 individuals identified by the census and 92 individuals initially missed by the census. For reasons of efficiency in recruitment and examination, those who were eligible were randomly divided into 10 groups to be examined over a 30-month period beginning March I, 1988. Those identified after this date also were randomized into one of the remaining groups.
PROCEDURES Letters from their primary care physicians or from the principal investigators if no primary care physician was identified in the census were sent to those individuals who were eligible to participate in the study. The letters described the study and invited eligible people to participate. The letter was followed by a call from the study coordinator, who provided further information about the study and made an appointment for the examination. People who were not interested in participating in the examination were asked to respond by telephone to the same
questionnaire that was administered at the time of the examination. Of the 5925 eligible individuals, 4926 (83.1 %) participated in the study. Participants were examined in a specially designed clinical site at the Beaver Dam Hospital (n = 4859,98.6%). Fifty-two of the 87 people residing in nursing homes were examined, 5 (0.1 %) of whom were examined at the clinical site and 47 (0.9%) of whom were examined at the three nursing home facilities. In addition, examinations also were conducted in the homes of 20 (0.4%) eligible persons who were confined by illness. Two hundred twenty-five people (3.8%) had died before the examination, 91 people (1.5%) moved out of the area, and 23 people (0.4%) could not be located. Two hundred sixty-nine people (4.5%) permitted an interview only, and 391 (6.6%) refused to participate. Comparisons between participants and nonparticipants are found in Table I. Ninety-nine percent of the population was white. The distribution of total household income reported was: 760 (15.4%) people-less than $10,000; 1300 (26.4%) people-$lO,OOO to $19,000; 946 (19.2%) people-$20,000 to $29,000; 956 (19.4%) people-$30,000 to $44,000; 424 (8.6%) people-$45,000 to $59,000; 285 (5.8%) people$60,000 or more; and 255 (5.2%) people-did not know or refused to answer. The parts of the examination pertinent to this report consisted of a standardized refraction and measurement of the visual acuity using the Humphrey 530 refractor. The refraction, obtained using the refractor, was placed in a trial lens frame and the best-corrected visual acuity was remeasured using the Early Treatment Diabetic Retinopathy Study (ETDRS) protocol using chart Rand 1311
OPHTHALMOLOGY
•
AUGUST 1991
•
VOLUME 98
•
NUMBER 8
Table 2. Mean Number of Letters* Read Correctly by Eye and by Sex in the Beaver Dam Eye Study (1988 to 1990) Age (yrs) Sex Men
Eye 00
as Women
00
as Men and women
00
as
N Mean SO N Mean SO N Mean SO N Mean SO N Mean SO N Mean SO
43-54
55-64
65-74
75+
Total
721 56.5 13.2 721 58.0 11.8 794 55.0 10.5 794 . 56.4 8.4 1515 55.7 11.9 1515 57.2 10.2
611 55.4 10.1 611 55.2 15.6 698 52.6 11.0 697 54.1 8.8 1309 53.9 10.7 1308 54.6 12.5
540 50.6 15.4 540 52.5 14.7 738 48.7 14.4 737 50.3 12.8 1278 49.5 14.8 1277 51.2 13.7
285 43.0 21.0 285 42.8 21.2 510 40.2 20.6 510 40.2 20.8 795 41.2 20.7 795 41.1 20.9
2157 52.9 15.0 2157 53.8 15.9 2740 49.9 15.0 2738 51.1 14.0 4897 51.2 15.1 4895 52.3 14.9
as
00 = right eye; SO = standard deviation; = left eye. * Snellen visual acuity equivalents for mean number of letters read correctly are as follows: 20/40 = 40 letters; 20/32 = 45 letters; 20/25 = 50 letters; 20/20 = 55 letters; 20/15 = 60 letters.
modified for a 2-meter distance. 7 If the best-corrected visual acuity was 20/40 or worse, an ETDRS refraction was performed and the visual acuity was measured. Visual acuity was not obtainable in both eyes for nine people, in the right eye for two 'people, and in the left eye for one person. No clinically significant interobserver variation or drift was found among the five examiners for obtaining the refractive error or the best-corrected visual acuity (data not presented).
used for analyzing the data. 9 Student's t test, both unpaired and paired, was used to compare differences in continuous response (mean number ofletters correct) between specific groups of persons and between eyes. Chi-square statistics were computed to compare frequency distributions of categorical visual acuity. Confidence intervals for correlation coefficients were computed using the method developed by Fisher. 10
DEFINITIONS
RESULTS
For each eye, the visual acuity was recorded as the number ofletters correctly identified (range, 0 [
Abstract: Few current population-based data on visual impairment are available. Visual acuity and impairment were measured in 4926 people between the ages of 43 and 86 years in the defined population participating in the Beaver Dam Eye Study. Visual acuity was measured after refraction, using standardized protocols. Of a possible maximum score of 70 (20/10), the mean number of letters correctly identified (right eye) varied from 55.7 (20/20, n = 1515) in people between the ages of 43 and 54 years to 41.2 (20/40, n = 795) in people 75 years of age or older. Age-specific mean visual acuity scores were consistently and significantly lower in women, who identified three fewer letters on the average than men. Rates of any visual impairment (20/40 or worse in the better eye) or legal blindness (20/200 or worse in the better eye), increased from 0.8% and 0.1 %, respectively, in people between the ages of 43 and 54 years to 21.1 % and 2.0%, respectively, in people 75 years of age or older. Multivariate analyses showed both sex (women) and age (older) to be significant and independent predictors of poorer visual acuity. Ophthalmology 1991; 98: 131 0-1315
Visual acuity is an important measure of visual function. However there are few current population-based estimates of best-corrected visual acuity available for the United States. 1.2 Such data are useful for estimating needs for medical care, counseling and rehabilitative services, projecting costs, measuring temporal trends, developing causal insights, and measuring quality of life. Thus, we have assessed visual acuity in a large geographically defined adult population and described the findings in this report.
MATERIALS AND METHODS THE POPULATION
Beaver Dam, WI, was selected because it had an excellent response rate in a previous population-based study Originally received: December 19, 1990. Revision accepted: March 5, 1991. 1
2
Department of Ophthalmology, University of Wisconsin, Madison. Department of Statistics, University of Wisconsin, Madison.
Supported by grant U10 EY06594 (Drs. Ronald Klein and Barbara E. K. Klein) from the National Eye Institute/NIH, Bethesda, Maryland. Reprint requests to Ronald Klein, MD, MPH, Department of Ophthalmology, Clinical Science Center, 600 Highland Ave, Madison, WI 53792.
1310
of diabetic retinopathy (96.0%). 3 It also had an ideal-sized population (n = 17,179, 1980 census), which was relatively stable (outmigration in Dodge County between 1975 and 1980 was 0.32% per year).4 From September 15, 1987 to May 4, 1988, a private census of the population living in the city and township of Beaver Dam, Wisconsin, was performed by the University of Wisconsin Extension-Survey Research Laboratory. Using procedures described elsewhere, the Survey Research Laboratory provided information on 6654 Beaver Dam households. 5,6 Of these, there were 42 households from which there was duplicate information and the second set of data was therefore omitted. There was a small percentage (1.3%) of the 6612 unique households for which there were incomplete surveys, either because the residents could not be reached (0.4%) or they refused (0.9%). The census interview included questions on demographic characteristics, ocular and systemic conditions, and primary eye care providers. Data from the census have been reported elsewhere. 6 Pertinent parts of the census questionnaire included questions regarding distance vision ("Do you see well enough with glasses or contact lenses to recognize a friend across the street"), near vision ("Do you see well enough with or without glasses to recognize letters in newspaper stories"), and a self-reported medical history of cataract, age-related macular degeneration, glaucoma, diabetes mellitus, or hypertension, and
KLEIN et al
•
BEAVER DAM EYE STUDY
Table 1. Selected Characteristics of Participants and Nonparticipants in the Beaver Dam Eye Study (1988 to 1990) Men Characteristic* Number Age (yrs) Mean SO History of poor distance vision (%) History of poor near vision (%) History of presence of cataract (%) History of presence of age-related macular degeneration (%) History of glaucoma (%) History of diabetes mellitus (%) History of hypertension (%) History of being seen by an eye doctor within 2 years of census
Participant
Women
Nonparticipant
2166
497
Participant 2760
All
Nonparticipant 502
Participant 4926
Nonparticipant 999
59.5 10.9 1.7 4.0 7.8
62.4t 12.4 3.4:j: 5.2 10.9§
61.3 11.5 3.0 5.1 12.0
64.6 12.4 4.6 6.6 17.7t
60.6 11.3 2.4 4.7 10.1
63.5:j: 12.4 4.0:j: 5.9 14.3t
2.9 2.9 6.8 35.4
1.6 5.0§ 7.7 33.2
4.4 3.9 6.8 37.3
5.2 5.6 7.2 39.6
3.7 3.4 6.8 36.4
3.4 5.3§ 7.4 36.4
74.8
68.8:j:
78.2
68.1t
76.7
68.5t
SO = standard deviation. * As determined from responses to questionnaire census. t P< 0.001. :j: P< 0.01. § P< 0.05.
a history of ever having seen an eye doctor. This information was elicited about all household members. Eligibility requirements for entry into the study included living in the city or township of Beaver Dam and being 43 to 84 years of age at the time of the census. Of the 6612 households identified by the census, 3715 contained at least one person satisfying the age criteria. These households reported a total of 5833 individuals between the ages of 43 and 84 years at the time of the census. After completion of the census, 76 additional households with a total of 92 people were identified and are included in the population. The data presented here are based on a total of 5833 individuals identified by the census and 92 individuals initially missed by the census. For reasons of efficiency in recruitment and examination, those who were eligible were randomly divided into 10 groups to be examined over a 30-month period beginning March I, 1988. Those identified after this date also were randomized into one of the remaining groups.
PROCEDURES Letters from their primary care physicians or from the principal investigators if no primary care physician was identified in the census were sent to those individuals who were eligible to participate in the study. The letters described the study and invited eligible people to participate. The letter was followed by a call from the study coordinator, who provided further information about the study and made an appointment for the examination. People who were not interested in participating in the examination were asked to respond by telephone to the same
questionnaire that was administered at the time of the examination. Of the 5925 eligible individuals, 4926 (83.1 %) participated in the study. Participants were examined in a specially designed clinical site at the Beaver Dam Hospital (n = 4859,98.6%). Fifty-two of the 87 people residing in nursing homes were examined, 5 (0.1 %) of whom were examined at the clinical site and 47 (0.9%) of whom were examined at the three nursing home facilities. In addition, examinations also were conducted in the homes of 20 (0.4%) eligible persons who were confined by illness. Two hundred twenty-five people (3.8%) had died before the examination, 91 people (1.5%) moved out of the area, and 23 people (0.4%) could not be located. Two hundred sixty-nine people (4.5%) permitted an interview only, and 391 (6.6%) refused to participate. Comparisons between participants and nonparticipants are found in Table I. Ninety-nine percent of the population was white. The distribution of total household income reported was: 760 (15.4%) people-less than $10,000; 1300 (26.4%) people-$lO,OOO to $19,000; 946 (19.2%) people-$20,000 to $29,000; 956 (19.4%) people-$30,000 to $44,000; 424 (8.6%) people-$45,000 to $59,000; 285 (5.8%) people$60,000 or more; and 255 (5.2%) people-did not know or refused to answer. The parts of the examination pertinent to this report consisted of a standardized refraction and measurement of the visual acuity using the Humphrey 530 refractor. The refraction, obtained using the refractor, was placed in a trial lens frame and the best-corrected visual acuity was remeasured using the Early Treatment Diabetic Retinopathy Study (ETDRS) protocol using chart Rand 1311
OPHTHALMOLOGY
•
AUGUST 1991
•
VOLUME 98
•
NUMBER 8
Table 2. Mean Number of Letters* Read Correctly by Eye and by Sex in the Beaver Dam Eye Study (1988 to 1990) Age (yrs) Sex Men
Eye 00
as Women
00
as Men and women
00
as
N Mean SO N Mean SO N Mean SO N Mean SO N Mean SO N Mean SO
43-54
55-64
65-74
75+
Total
721 56.5 13.2 721 58.0 11.8 794 55.0 10.5 794 . 56.4 8.4 1515 55.7 11.9 1515 57.2 10.2
611 55.4 10.1 611 55.2 15.6 698 52.6 11.0 697 54.1 8.8 1309 53.9 10.7 1308 54.6 12.5
540 50.6 15.4 540 52.5 14.7 738 48.7 14.4 737 50.3 12.8 1278 49.5 14.8 1277 51.2 13.7
285 43.0 21.0 285 42.8 21.2 510 40.2 20.6 510 40.2 20.8 795 41.2 20.7 795 41.1 20.9
2157 52.9 15.0 2157 53.8 15.9 2740 49.9 15.0 2738 51.1 14.0 4897 51.2 15.1 4895 52.3 14.9
as
00 = right eye; SO = standard deviation; = left eye. * Snellen visual acuity equivalents for mean number of letters read correctly are as follows: 20/40 = 40 letters; 20/32 = 45 letters; 20/25 = 50 letters; 20/20 = 55 letters; 20/15 = 60 letters.
modified for a 2-meter distance. 7 If the best-corrected visual acuity was 20/40 or worse, an ETDRS refraction was performed and the visual acuity was measured. Visual acuity was not obtainable in both eyes for nine people, in the right eye for two 'people, and in the left eye for one person. No clinically significant interobserver variation or drift was found among the five examiners for obtaining the refractive error or the best-corrected visual acuity (data not presented).
used for analyzing the data. 9 Student's t test, both unpaired and paired, was used to compare differences in continuous response (mean number ofletters correct) between specific groups of persons and between eyes. Chi-square statistics were computed to compare frequency distributions of categorical visual acuity. Confidence intervals for correlation coefficients were computed using the method developed by Fisher. 10
DEFINITIONS
RESULTS
For each eye, the visual acuity was recorded as the number ofletters correctly identified (range, 0 [
