Documente Academic
Documente Profesional
Documente Cultură
METHODS
Population
Proyecto VER (Vision and Eye Research Project) is a population-based
survey of visual impairment and blindness among noninstitutionalized
Hispanics, aged 40 years or more living in Nogales or Tucson, Arizona.
Sampling methods have been described elsewhere.3 In summary, a
stratified random sample of block groups were selected based on
census data of the Hispanic population in 1990. Every other household
in Nogales block groups and two of every three households in Tucson
block groups were chosen for determination of eligibility. All members
of selected households who were self-described Hispanics aged 40 or
more years were eligible to participate.
3393
3394
Broman et al.
Subscale
General Health
General Vision
Near Vision
Distance Vision
Driving
Ocular Pain
Role Difficulties
Dependency
Social Function
Mental Health
Color Vision
Peripheral Vision
Item
Five-level health rating
Six-level vision rating
Reading ordinary print in newspapers
Work or hobbies that require you to see well
up close
Finding something on a crowded shelf
Reading street or shop signs
Going down stairs in dim light
Going out to see movies, plays, sports events
Driving during the day
Driving during the night
Eye pain keeps you from activities
Amplitude of eye pain
Accomplish less because of vision
Limited in length of activities because of
vision
Stay home because of eyesight
Rely on others because of eyesight
Need help from others because of eyesight
Seeing how people react to things you say
Visiting with people
Worry about eyesight
Worry about embarrassing myself because of
eyesight
Feel frustrated because of eyesight
Have less control of what I do because of
eyesight
Picking out and matching your clothes
Noticing objects off to the side
Home Interview
The questionnaire was administered by trained personnel and offered
in English and Spanish. The majority of home interviews (80%) were
conducted in Spanish. The questionnaire contained specific questions
on education, socioeconomic and health status, history of vision problems, visit to an eye care professional, and Native-American ancestry.
Also included was the NEI-VFQ-25, and a series of questions on health
care utilization, access, and insurance from the Hispanic Health and
Nutrition Examination Survey (HHANES).14 An acculturation index was
created by summarizing specific questions from HHANES regarding
preferred language, country of birth, and other questions relating to
adapting to U.S. culture.15 The acculturation score ranges from 1 to 5,
where 1 is low acculturation, and 5 is high acculturation.
The NEI-VFQ-25 is a subset of the original 51-item NEI-VFQ.16
Subscales and items for the 1996 version used in this study are shown
in Table 1. Items were answered in the following manner: amount of
difficulty, amount of time, amount true, and amplitude. Item
responses were adjusted for directionality (high scores reflect participants with good vision/health), and were rated on a scale of 0 to 100.
Subscale scores were created by averaging their adjusted item responses. Participants responses were excluded from specific items if
they had stopped the activity for reasons other than poor eyesight or
if someone other than the participant had answered the questionnaire.
Clinical Measures
Presenting and best-corrected distance acuity were measured at the
clinic. Acuity was measured in each eye with the participant using
Statistical Analysis
In previous work13 we showed that the NEI-VFQ-25 was not only
influenced by visual function, but by other demographic factors as
well: age, gender, level of acculturation, income level, education level,
location (Nogales or Tucson), and self-report of having medical insurance. These confounders were chosen because of their demonstrated
effects on self-reported quality of life in this population and their
association with eye disease. We adjusted for all these variables in the
ensuing multivariate regression models to obtain more precise estimates of the relationship of vision with quality-of-life domains. Although diabetes was an important comorbid condition in this population, the models do not include diabetes as a covariate, because of its
high association with DR; however, we adjusted for the presence of
hypertension as a representative of comorbid conditions. All covariates, except age, were modeled as categorical variables.
We used presenting acuity in the better-seeing eye to characterize
vision used on a daily basis. This approach was reasonable, because
acuity in the better eye has been shown to be as good a predictor of
self-reported functional status as other approaches of summarizing
better- and worse-eye acuity.20 We modeled presenting acuity as a
continuous variable (logMAR units) and analyzed the strength of the
association between lines lost and subscale score.
RESULTS
Of the 4774 participants in the study, 13 (0.3%) had questionnaires completed by proxy and 7 participants refused to undergo measurement of refraction. Of the remainder, 175 participants did not receive a complete eye disease assessment,
and 29 were diagnosed with eye disease other than glaucoma,
DR, or cataract. These latter two groups totaling 204 were
excluded from the analyses of eye disease and quality of life.
The average age for our sample was 56.9 years (range,
40 96; Table 2). Approximately two thirds of this sample had
a yearly income below $20,000 and had less than a high school
3395
Gender
Yearly income
Education
Acculturation Score
Location
Current Drivers
Medical Insurance
Hypertension
Glaucoma
Diabetic retinopathy
Cataract
Other eye disease
Uncorrected
refractive error
Category
n (%)
20/40 or better
Between 20/41 and 20/199
20/200 or worse
4368 (91.9)
363 (7.6)
23 (0.5)
4049
5059
6069
7079
80
Male
Female
$20,000
$20,000
12 years
12 years
12
23
34
44.5
Nogales
Tucson
1591 (33.4)
1361 (28.6)
982 (20.6)
631 (13.3)
195 (4.1)
1841 (38.7)
2920 (61.3)
3151 (66.2)
1492 (31.3)
3101 (65.1)
1660 (34.9)
2763 (58.0)
972 (20.4)
942 (19.8)
84 (1.8)
1620 (34.0)
3141 (66.0)
3522 (74.0)
3192 (67.1)
2085 (56.2)
100 (2.1)
290 (6.3)
374 (8.0)
29 (0.6)
405 (8.5)
3396
Broman et al.
TABLE 3. Relationship of Presenting Acuity in the Better Eye to Scale Scores of The NEI-VFQ 25, Adjusting for Demographic Factors
Change in Scale Score with Addition
of 2 Lines of Acuity Loss for Acuity
Worse Than 20/40
NEI-VFQ Subscale
Est.
General health
General vision
Near vision
Distance vision
Driving
Peripheral vision
Color vision
Ocular pain
Role difficulties
Dependency
Social functioning
Mental health
1.84
3.60
5.26
3.78
4.49
3.77
2.67
2.67
6.23
5.27
3.39
3.66
Bootstrap
95% CI
(2.65
(4.28
(6.05
(4.61
(5.72
(4.63
(1.76
(3.52
(7.14
(6.17
(4.20
(4.73
to
to
to
to
to
to
to
to
to
to
to
to
1.10)
2.99)
4.50)
2.95)
3.33)
2.99)
3.55)
1.94)
5.31)
4.28)
2.63)
2.59)
Est.
Bootstrap
95% CI
8.60
11.65
(10.2 to 6.93)
(15.0 to 8.68)
9.00
(10.7 to 7.13)
All models are adjusted for age, gender, acculturation, income, education level, location, and medical insurance. Est., estimate.
* There is a 3-line difference between 20/20 acuity and 20/40 acuity.
For these domains, the slope changed significantly after 20/40, and we show the decline after that point (spline regression).
mixed cataract (86/374; 23.0%), with the greatest acuity impairment in those with nuclear cataract (average, 0.29 logMAR)
and mixed cataract (average, 0.36 logMAR).
After adjustment for demographic variables, those with
glaucoma, DR, cataract, and uncorrected refractive error were
found to have lower scale scores than those without disease in
general (Table 6). Those with glaucoma had significantly lower
scores in all domains except General Health and Ocular Pain
than did those with no eye disease. Those with DR had significantly lower scores in all domains. The largest decrease occurred in General Health, which was expected, because these
participants had diabetes. Those with glaucoma or DR reported larger decreases in the areas of Near and Distance
Vision, Driving, and Peripheral Vision and in the psychosocial
domains of Role Difficulties, Dependency, Social Function, and
Mental Health than did those with cataract or uncorrected
refractive error.
The association of eye disease with quality-of-life-domains
was not as strong after adjustment for presenting acuity; howTABLE 4. Degree of Decrement in Subscale Scores of Those with
Monocular Presenting Acuity Impairment, Compared with Those
with 20/40 or Better Acuity in Both Eyes
Monocular Impairment
NEI-VFQ Subscale
Est.
General health
General vision
Near vision
Distance vision
Driving
Peripheral vision
Color vision
Ocular pain
Role difficulties
Dependency
Social functioning
Mental health
0.42
3.44*
1.39
2.45*
5.68*
3.02*
0.55
0.76
4.43*
2.34*
2.08*
2.69*
Bootstrap
95% CI
(2.97
(5.03
(3.18
(4.07
(8.24
(4.92
(1.67
(2.62
(6.64
(3.99
(3.21
(4.48
to
to
to
to
to
to
to
to
to
to
to
to
1.74)
1.94)
0.54)
0.85)
3.19)
1.25)
0.45)
1.22)
2.28)
0.76)
1.03)
0.83)
All models are adjusted for presenting acuity in the better eye, age,
gender, acculturation, income, education level, location, and medical
insurance. Est., estimate.
* Indicates a significant estimate.
ever, there still were significant effects of disease (Table 7). For
glaucoma, the difference in scores after adjustment for acuity
was modest, in most cases only one point or less different from
scores that were not adjusted for acuity. For DR, the change in
scores was slightly greater (except for General Health), but still
the overall decrement in scale scores with DR was large, after
adjustment for acuity. None of the scale scores was significantly different between those with cataract and those with no
eye disease, except in Driving, after adjustment for acuity. This
suggests that the effect of cataract on quality of life in this
population was due to the effect on decreasing acuity.
The models that included interactions between glaucoma,
cataract, and DR, suggested that multiple eye disease was
associated with additional decrements in quality of life, beyond
the additive decrements associated with each disease separately. Table 8 shows the proportion of our participants with
different combinations of eye disease. After adjustment for
demographic variables, there was a significant additional decrement in score (range, 20.2 to 8.5, P 0.05) in the scales
of Near and Distant Vision, Driving, Ocular Pain, Role Function, Dependency, Mental Health, and Color Vision for those
with both glaucoma and cataract. There was a significant decrease in Driving, Role Function, and Dependency (range,
17.8 to 14.5, P 0.05) for those with glaucoma and DR,
and there was also a significant decrease in the Driving score
(8.7, P 0.04) for those with DR and cataract. However, the
number of persons with combinations of eye disease was small
and did not permit more detailed analyses.
DISCUSSION
Visual acuity impairment was significantly associated with a
decrement in all measured domains of self-reported, visionrelated quality of life in this sample of Mexican-Americans. For
TABLE 5. Comparison of Mean LogMAR Acuity in the Better Eye of
Participants with Cataract, Glaucoma, or DR
Disease
Mean Acuity
Range
Cataract
Glaucoma
DR
374
100
290
0.21
0.22
0.12
0.261.7
0.181.7
0.31.5
3397
TABLE 6. Effect of Eye Disase on NEI-VFQ-25 Scale Scores, Compared with Those with No Eye Disease, and Adjusting
for Demographic Variables
Glaucoma
NEI-VFQ Scale
Est.
General health
General vision
Near vision
Distance vision
Driving
Peripheral vision
Color vision
Ocular pain
Role difficulties
Dependency
Social functioning
Mental health
1.81
6.30*
9.01*
9.60*
17.0*
9.46*
5.78*
4.05
10.2*
14.6*
8.04*
12.7*
Diabetic Retinopathy
Bootstrap
95% CI
(7.0
(10.6
(14.5
(15.3
(28.2
(15.2
(11.1
(9.48
(13.8
(22.1
(13.1
(19.3
to
to
to
to
to
to
to
to
to
to
to
to
Est.
3.4)
1.89)
2.87)
3.62)
6.04)
3.99)
1.14)
0.49)
6.87)
7.97)
3.39)
6.95)
16.0*
4.74*
8.44*
8.64*
11.7*
6.22*
2.54*
2.92*
10.2*
8.70*
4.92*
9.73*
Bootstrap
95% CI
(18.8
(6.76
(11.4
(11.4
(15.9
(9.02
(4.77
(5.53
(13.8
(12.1
(7.14
(13.0
to
to
to
to
to
to
to
to
to
to
to
to
13.5)
2.74)
5.58)
5.90)
7.60)
3.55)
0.93)
0.23)
6.87)
5.87)
2.64)
6.46)
Uncorrected
Refractive Error
Cataract
Bootstrap
95% CI
Est.
0.09
2.74*
2.87*
3.75*
4.60*
1.62
0.83
1.22
4.40*
2.87*
1.62
3.19*
(2.65
(4.78
(5.74
(6.51
(8.54
(4.16
(2.89
(1.29
(7.53
(5.88
(3.57
(6.00
to
to
to
to
to
to
to
to
to
to
to
to
Bootstrap
95% CI
Est.
2.59)
0.88)
0.38)
1.45)
1.37)
0.67)
0.81)
3.75)
1.47)
0.51)
0.05)
0.55)
0.67
2.58*
5.07*
4.67*
3.09*
1.98
1.12
3.07*
5.51*
3.79*
1.99*
5.24*
(1.79
(4.50
(7.58
(6.80
(6.23
(3.97
(2.82
(5.60
(8.64
(6.32
(3.52
(7.86
to
to
to
to
to
to
to
to
to
to
to
to
3.73)
0.78)
2.88)
2.69)
0.07)
0.005)
0.26)
0.63)
2.88)
1.66)
0.51)
2.82)
TABLE 7. Effect of Eye Disease on NEI-VFQ-25 Scale Scores, Adjusted for Presenting Acuity and Other Demographic Variables
Glaucoma
NEI-VFQ Subscale
Est.
General health
General vision
Near vision
Distance vision
Driving
Peripheral vision
Color vision
Ocular pain
Role difficulties
Dependency
Social functioning
Mental health
1.39
5.14*
7.69*
7.88*
14.8*
8.47*
5.19*
3.18
9.58*
13.0*
7.28*
10.5*
Diabetic Retinopathy
Bootstrap
95% CI
(6.49
(9.31
(12.6
(13.4
(25.4
(14.0
(10.4
(8.57
(15.4
(20.3
(12.0
(16.4
to
to
to
to
to
to
to
to
to
to
to
to
3.87)
1.21)
2.10)
2.68)
4.32)
3.23)
0.94)
1.19)
3.63)
6.94)
2.97)
4.97)
Est.
15.6*
3.75*
6.96*
6.99*
10.1*
5.14*
1.87*
2.16
8.55*
7.28*
4.07*
8.15*
Bootstrap
95% CI
(18.5
(5.68
(9.88
(9.62
(14.4
(7.92
(4.04
(4.70
(11.7
(10.4
(6.19
(11.3
to
to
to
to
to
to
to
to
to
to
to
to
13.2)
1.68)
4.24)
4.38)
6.32)
2.65)
0.26)
0.49)
5.38)
4.58)
1.99)
5.28)
Cataract
Est.
0.53
1.23
0.89
1.41
3.65*
0.17
0.09
2.35
2.03
0.84
0.45
0.62
Bootstrap
95% CI
(2.09
(3.13
(3.75
(3.93
(7.45
(2.60
(1.84
(0.12
(4.88
(3.57
(2.35
(3.24
to
to
to
to
to
to
to
to
to
to
to
to
3.20)
0.56)
1.50)
0.81)
0.38)
2.06)
1.68)
4.87)
0.97)
1.42)
1.26)
1.93)
3398
Broman et al.
85.6
6.3
5.1
1.4
0.9
0.3
0.3
0.04
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
Acknowledgments
The authors thank Gary Rubin, Karen Bandeen-Roche, and the anonymous referees for their advice and input; and team of Proyecto VER for
their skill and support.
21.
References
22.
1. Novello AC, Wise PH, Kleinman DV. Hispanic health: time for data,
time for action. JAMA. 1991;265:253255.
2. Munoz B, West SK, Rodriguez J, et al. Blindness, visual impairment
and the problem of uncorrected refractive error in a Mexican-
23.