Myopia and Incident Cataract and Cataract Surgery:
The Blue Mountains Eye Study
Christine Younan,1 Paul Mitchell,1 Robert G. Cumming,2 Elena Rochtchina,1
and Jie Jin Wang1
PURPOSE. To assess whether an association exists between
myopia and incident cataract and cataract surgery in an older
population-based cohort study.
METHODS. The Blue Mountains Eye Study examined 3654 par-
ticipants aged 49 years or more during 1992 to 1994 and then
2334 (75.1%) of the survivors after 5 years. A history of using
eyeglasses for clear distance vision was obtained. Objective
refraction was performed with an autorefractor, followed by
subjective refraction with a logarithm of minimum angle of
resolution (logMAR) chart. Emmetropia was defined as a spher-
ical equivalent refraction between ⫹1 D and ⫺1 D, hyperopia
as more than ⫹1 D, and myopia as less than ⫺1 D. Slit lamp and
retroillumination lens photographs were graded for presence
of cortical, nuclear, or posterior subcapsular cataract, accord-
ing to the Wisconsin Cataract Grading System. Generalized
estimating equation models analyzed data by eye.
RESULTS. There was a statistically significant association be-
tween high myopia (⫺6 D or less) and incident nuclear cata-
ract (odds ratio [OR] 3.3, 95% confidence interval [CI] 1.5–
7.4). Incident posterior subcapsular cataract was associated
with any myopia (OR 2.1, 95% CI 1.0 – 4.8), moderate to high
myopia (⫺3.5 D or less, OR 4.4, 95% CI 1.7–11.5), and use of
distance glasses before age 20 (OR 3.0, 95% CI 1.0 –9.3), after
adjustment for multiple potential confounders, including sever-
ity of nuclear opacity. Incident cataract surgery was signifi-
cantly associated with any myopia (OR 2.1, 95% CI 1.1– 4.2) as
well as moderate (⫺3.5 to more than ⫺6D; OR 2.9, 1.2–7.3)
and high myopia (OR 3.4, 95% CI 1.0 –11.3).
CONCLUSIONS. These epidemiologic data provide some evidence
of an association between myopia and incident cataract and
cataract surgery, after adjustment for multiple confounders and
severity of nuclear opacity. These data support other cross-
sectional and longitudinal population-based findings. (Invest
Ophthalmol Vis Sci. 2002;43:3625–3632)
M yopia is relatively common in older populations, with
reports indicating prevalence rates ranging from 15% (in
those 49 years and older)1 to 38.7% (in those aged 40 –79
years).2 Cataract is also common in older age groups, with
From the 1Department of Ophthalmology, and the 2Department of
Public Health and Community Medicine, University of Sydney, Sydney,
Australia.
Supported by Australian National Health and Medical Research
Council Grant 974159, and the Westmead Millennium and Save Sight
Institutes, University of Sydney.
Submitted for publication November 16, 2001; revised April 18,
2002; accepted April 26, 2002.
Commercial relationships policy: N.
The publication costs of this article were defrayed in part by page
charge payment. This article must therefore be marked “advertise-
ment” in accordance with 18 U.S.C. §1734 solely to indicate this fact.
Corresponding author: Paul Mitchell, Department of Ophthalmol-
ogy (Centre for Vision Research), University of Sydney, Westmead
Hospital, Hawkesbury Road, Westmead, NSW 2145, Australia;
paulmi@westgate.wh.usyd.edu.au.
Investigative Ophthalmology & Visual Science, December 2002, Vol. 43, No. 12
Copyright © Association for Research in Vision and Ophthalmology
increasing age associated with increased prevalence3,4 and
incidence.5 High myopia is known to be associated with cata-
ract,6 and a relationship between myopia and cataract has been
suggested.7,8 Other reports reject the association with myopia,
instead explaining that a trend to myopia is the direct conse-
quence of the presence of nuclear cataract.9
To date, few population-based studies have attempted to
assess the association between myopia and cataract. Cross-
sectional data from the Blue Mountains Eye Study have pro-
vided such evidence, showing an association between myopia
and both nuclear and posterior subcapsular cataract.10 The
cross-sectional association between myopia and nuclear cata-
ract was supported by data from the Beaver Dam Eye Study.11
The longitudinal data from Beaver Dam, however, noted in-
creased incident nuclear cataract, and possibly incident corti-
cal cataract, in hyperopic eyes. In that study no relationship
was found between myopia and 5-year incident cataract, but a
higher incidence of cataract surgery was reported in myopes.12
A laboratory-based study identified reduced antioxidant
properties in myopic eyes compared with those with typical
age-related cataract.13 Increased levels of lipid peroxidation
by-products have been found in cataractous lenses and in the
vitreous of myopes compared with control subjects and non-
myopic cataractous lenses.13,14 An association was also shown
between the degree of retinal lipid peroxidation and lens
opacity in rodents.15 These studies provide a plausible expla-
nation for the association between myopia and cataract and
suggest that increasing myopia may be related to increasing
damage to rod outer segments, which could lead to potentially
cataractous by-products.
Our purpose in the present report is to assess the associa-
tion between myopia and incident cataract and cataract sur-
gery in a group of older Australians for whom baseline and
follow-up information had been collected over a 5-year inter-
val. These data could add further epidemiologic evidence to
the debate about whether an association exists between myo-
pia and cataract and may serve to guide laboratory-based stud-
ies in the search for biological explanations of cataract risk
factors.
METHODS
The Blue Mountains Eye Study is a population-based study of vision and
common eye diseases in an urban population aged 49 years or older,
resident in two postal codes of the Blue Mountains region, west of
Sydney, Australia. The baseline survey methods and procedures have
been described.16,17 The study was approved by the Western Sydney
Area Health Service Human Ethics Committee and signed, informed
consent was obtained from all participants. The research was con-
ducted according to the recommendations of the Declaration of Hel-
sinki. During 1992 to 1994 at the baseline examinations, 3654 (82.4%)
of the 4433 eligible residents aged 49 to 97 years were assessed.
Five-year follow-up examinations were conducted during 1997 to
1999, when 2334 (75.1%) of the survivors were reexamined. Of those
not seen, 383 (12.3%) had moved from the area, and 394 (12.7%)
refused the examination.
3625
3626 Younan et al.
Questionnaire and Definitions
An interviewer-administered questionnaire enabled documentation of
a detailed general medical and ocular history, as well as general demo-
graphic information. A history of myopia was sought by asking each
participant whether he or she currently wore glasses to see clearly in
the distance (including bifocals or multifocals), or had previously done
so. If worn, the participant was asked at what age glasses were first
used for clear distance vision. Data from this question were used only
for participants with eyes having a measured myopic refractive error at
baseline. Objective refraction was performed with an autorefractor
(model 530; Humphrey, San Leandro, CA). This was followed by
subjective refraction according to the Beaver Dam Eye Study modifi-
cation of the Early Treatment Diabetic Retinopathy Study (ETDRS)
protocol using a logarithm of minimum angle of resolution (logMAR)
chart.17,18
Baseline refraction data were used for analyses. The baseline refrac-
tive state was defined as the spherical equivalent refraction (SER),
calculated by the algebraic addition of the best corrected spherical
refraction and half the cylindrical refraction. Emmetropia was defined
as a SER between ⫹1 D and ⫺1 D, hyperopia as more than ⫹1 D, and
myopia less than ⫺1 D. Myopia was further classified as low (less than
⫺1 D to more than ⫺3.5 D), moderate (⫺3.5 D or less to more than ⫺6
D) and high (⫺6 D or less). For those known to be myopic at baseline,
the age at which distance glasses were first worn was used as a proxy
for the onset and therefore duration of myopia. Hyperopia was also
divided into low (greater then ⫹1 D to less than ⫹2 D), moderate (⫹2
D to less than ⫹4 D), and high categories (⫹4 D or greater).
Participants were asked whether they smoked or consumed alco-
hol and whether oral or inhaled steroids had been prescribed in the
past. They were asked whether they had angina (also described as
“chest pain from your heart”), stroke, diabetes, or hypertension diag-
nosed by a doctor. Systolic and diastolic blood pressure was measured
with the subject seated, before the use of any eye drops. Hypertension
was defined either by history and/or a systolic measurement above 160
mm Hg and/or a diastolic measurement above 95 mm Hg. Diabetes was
defined either by history or a fasting blood glucose level of 7.0 mmol/L
or more. All blood samples were collected at a subsequent visit and
later analyzed at Westmead Hospital. Sun-related skin damage was
estimated by a clinical examiner on a four-point scale (none, mild,
moderate, and severe) by assessing the arms, hands, and face.19 Par-
ticipants had their weight (after removal of shoes and heavy clothing)
and height measured. Body mass index was calculated as weight/
height squared in kilograms per square meter, with obesity defined as
a body mass index of 30 or greater. Higher educational achievement
was defined as attainment of a qualification (certificate, diploma, or
degree) after leaving school.
Cataract Grading
Cataract was documented by both slit lamp (Topcon SL-7e camera;
Topcon Optical Co., Tokyo, Japan) and retroillumination (CT-R cata-
ract camera; Neitz Instrument Co., Tokyo, Japan) lens photographs.
Details of the photographic technique and grading3,16 used in the Blue
Mountains Eye Study have been reported. The grading closely followed
the Wisconsin Cataract Grading System,20 with good agreement found
for assessments of both inter- and intragrader reliability.3 History of
past cataract surgery was confirmed at both the examination and
photographic grading. At the follow-up study, graders were masked to
baseline cataract status.
Presence of nuclear, cortical, and posterior subcapsular cataract
was assessed in each eye.20 Presence and severity of nuclear cataract
was defined on a five-level scale by comparison with a set of four
standard slit lamp photographs. Participants with nuclear grades 1 to 3
at baseline who developed nuclear grade level 4 or 5 at 5-year fol-
low-up were defined as having incident nuclear cataract. The percent-
age area involved by cortical or posterior subcapsular cataract in each
eye was calculated from the estimated percentage area involved in
each of nine segments of the lens divided by a grid.20 Participants with
IOVS, December 2002, Vol. 43, No. 12
less than 5% cortical opacity at baseline who then developed 5% or
more of the total lens area involved at follow-up were defined as having
incident cortical cataract. Participants with no posterior subcapsular
cataract at baseline with the presence of any posterior subcapsular
cataract at follow-up were defined as having incident posterior sub-
capsular cataract. The definitions of each of these incident cataract
types were not mutually exclusive.
Statistical Analysis
Because refraction is eye specific, analyses were run according to eye
rather than subject. These were performed with all eyes combined
using a generalized estimating equation method described by Zeger et
al.21 and Liang and Zeger.22 This method allows data from both eyes to
be used while accounting for the correlation between the two eyes of
a single subject. Cataract was analyzed as a dichotomous variable.
Because age is strongly associated with cataract incidence5 and both
increased cataract prevalence3,4,23–25 and incidence5,26 has been noted
in women, all odds ratios are age- and sex-adjusted, unless otherwise
stated. For comparison, logistic regression analyses were also per-
formed in right and then left eyes for both hyperopia and myopia. All
presented results are from multivariate models that adjust for the same
potential confounders as found in the tables, including level of nuclear
opacity.
The variables included in multivariate generalized estimating equa-
tion models varied by cataract type. Variables considered for inclusion
were: age (categorically), sex, smoking (ever versus never), current
alcohol consumption (drinks per week), ever used inhaled steroids,
dark brown iris color, educational achievement, sun exposure
(none versus any sun-related skin damage), obesity, severity of
nuclear opacity (levels 1–5), and history of diabetes, hypertension,
stroke, or angina. Each model included only variables associated
with that cataract type, either from our own age- and sex-adjusted
incident analyses or from reports for prevalent or incident corti-
cal,25,27–30 nuclear,25,27,29,31,32 or posterior subcapsular cataract,25
as well as for cataract surgery.25
Statistical analysis was performed on computer (Statistical Analysis
System, ver. 6.12; SAS Institute Inc, Cary, NC). P ⬍ 0.05 was used to
indicate statistical significance. Odds ratio (OR) and 95% confidence
interval (CI) are presented.
RESULTS
Over the 5-year period, nuclear cataracts developed in 593
(23.4%) eyes, cortical in 350 (9.8%) eyes, and posterior sub-
capsular in 100 (2.5%) eyes, according to data from both eyes.
During this same period, 211 (4.7%) eyes underwent cataract
surgery. Baseline aphakia, pseudophakia, or enucleation was
present in 139 (3.0%) eyes, which were excluded from these
analyses. The reported data refer to participants who attended
both examinations and had gradable photographs from both
visits. Approximately 8.7% of eyes had missing baseline and/or
follow-up data for cortical or posterior subcapsular cataract,
because photographs were ungradable or were not taken. A
higher proportion (35.7%) had missing data for grading of
nuclear cataract, predominantly because of an intermittent
camera malfunction, as described in our previous report.3
There were no significant differences, however, between par-
ticipants with and without gradable photographs.3
Of participants who returned for the 5-year examination,
baseline refraction data were available on 4663 (99.9%) eyes,
including 2218 (47.6%) emmetropic, 1925 (41.3%) hyperopic,
and 520 (11.2%) myopic eyes. The myopic eyes included 330
(63.5%) with low myopia, 115 (22.1%) with moderate myopia,
and 75 (14.4%) eyes with high myopia. Information about the
age at which distance glasses were first worn was available for
464 eyes of participants with myopia (89.2%). In 52.6%, the age
was 40 years more, in 25.6% between ages 20 and 39 years and
IOVS, December 2002, Vol. 43, No. 12 Myopia and Incident Cataract and Cataract Surgery 3627

TABLE 1. Baseline Characteristics of Participants Who Did and significant age- and sex-adjusted association found was be-
Participants Who Did Not Attend the 5-Year Follow-up Examinations tween incident cortical cataract and moderate myopia (OR 1.8,
in the Blue Mountains Eye Study 95% CI 1.0 –3.4). This finding, however, was not statistically
significant after adjustment for multiple potential confounders.
Participants in Both
Examinations Survivors Who It is well known that nuclear cataract is associated with a
Included in Did Not Attend myopic shift.9 We therefore repeated our multivariate analyses
Baseline Characteristic Analyses (%) Follow-up (%) to adjust further for the severity of nuclear opacity. This indi-
cated no statistically significant association between moderate
Total 2278 777 myopia and incident cortical cataract. In logistic regression
Age analyses for right and left eyes, hyperopia was associated with
⬍60 712 (31.3) 260 (33.5) incidence of cortical cataract in an age- and sex-adjusted model
60–69 928 (40.7) 255 (32.8) (OR 1.4, 95% CI 1.0 –2.0), but not in a multivariate model that
70–79 529 (23.2) 191 (24.6)
80⫹ 109 (4.8) 71 (9.1)
also accounted for level of nuclear opacity (OR 1.3, 95% CI
Sex 0.9 –2.0). No statistically significant association was found,
Female 1309 (57.5) 472 (60.8) however, between hyperopia and incident cortical cataract in
Male 969 (42.5) 305 (39.3) left eyes (data not shown).
Right eye refractive error A statistically significant increased risk of incident nuclear
Emmetropia 1074 (47.3) 362 (47.3) cataract (OR 1.2, 95% CI 1.0 –1.5) was found in hyperopic
Hyperopia 944 (41.6) 304 (39.7) compared with emmetropic eyes, after adjustment for age and
Myopia 253 (11.1) 100 (13.1) sex (Table 3). However, this was not statistically significant in
Baseline prevalent cataract
Cortical 473 (20.8) 154 (19.8)
the multivariate model. An association between moderate hy-
Nuclear 239 (10.5) 88 (11.3) peropia and incident nuclear cataract was statistically signifi-
Posterior subcapsular 110 (4.8) 49 (6.3) cant in the multivariate model (OR 1.4, 95% CI 1.1–1.9). Anal-
Education 854 (39.3) 321 (45.1) yses for right eyes also showed a statistically significant
Inhaled steroids 241 (11.2) 61 (8.7) association between hyperopia and incident nuclear cataract in
Dark brown iris color 229 (10.3) 80 (10.5) an age- and sex-adjusted model (OR 1.4, 95% CI 1.0 –1.9),
though not in the model adjusting for multiple potential con-
founders (OR 1.3, 95% CI 0.9 –1.8). No statistically significant
in 21.8%, before 20 years. The hyperopic eyes included 961 association was found for left eyes (data not shown). Although
(49.9%) with low hyperopia, 836 (43.4%) with moderate hy- no association was found with low or moderate myopia, eyes
peropia, and 128 (6.6%) with high hyperopia. with high myopia had a statistically significant higher risk of
Baseline characteristics of participants in the 5-year fol- incident nuclear cataract, adjusting both for age and sex (OR
low-up examination who were included in any analyses and 3.1, 95% CI 1.5– 6.5) and in the multivariate model (OR 3.3,
survivors of the baseline examination who did not attend 95% CI 1.5–7.4).
follow-up are shown in Table 1. There were no statistically We found stronger associations between baseline refractive
significant differences (P ⬍ 0.05) between these two groups. status and incident posterior subcapsular cataract. The pres-
Of the 2334 participants in both examinations, 56 were not ence of any myopia was associated with more than a doubling
included in any analyses (44 because of nongradable photo- of the risk (5.4% vs. 2.1%) of incident posterior subcapsular
graphs at baseline and 12 because of nongradable photographs cataract (OR 2.9, 95% CI 1.7–5.2), after adjustment for age and
at follow-up). sex (Table 4). This relationship was similar after adjustment for
No statistically significant associations were found between multiple potential confounders (OR 2.6, 95% CI 1.4 –5.0). Anal-
hyperopia or any myopia (compared with emmetropia) and yses for left eyes also showed a statistically significant associa-
incident cortical cataract in either the age- and sex-adjusted or tion between incident posterior subcapsular cataract and my-
multivariate-adjusted models (Table 2). The only statistically opia in multivariate models (OR 3.5, 95% CI 1.3–10.1),

TABLE 2. Odds Ratios and 95% Confidence Intervals (CI) for the Association between Baseline Refraction and Incident Cortical Cataract in All Eyes

Multivariate Multivariate
Age and Sex Adjusted Adjusted* Adjusted†

Eyes at Incident Odds Odds Odds

Risk (n) Eyes (n) % Ratio 95% CI Ratio 95% CI Ratio 95% CI

Emmetropia 1776 151 8.5 1.0 Referent 1.0 Referent 1.0 Referent
Any hyperopia 1418 163 11.5 1.1 0.9–1.5 1.2 0.9–1.5 1.1 0.8–1.4
Hyperopia‡
Low 737 80 10.9 1.1 0.9–1.5 1.2 0.9–1.6 1.1 0.8–1.6
Moderate 594 72 12.1 1.2 0.9–1.7 1.2 0.9–1.7 1.1 0.7–1.6
High 87 11 12.6 1.3 0.6–2.6 1.4 0.7–2.8 1.1 0.5–2.7
Any myopia 381 35 9.2 1.1 0.7–1.7 1.0 0.7–1.6 0.7 0.4–1.3
Myopia§
Low 231 15 6.5 0.8 0.5–1.5 0.8 0.4–1.5 0.5 0.3–1.2
Moderate 96 14 14.6 1.8 1.0–3.4 1.7 0.8–3.3 1.3 0.5–3.2
High 54 6 11.1 1.3 0.6–3.1 1.0 0.4–2.5 0.5 0.2–2.0

* Adjusted for age, sex, education, alcohol, sun exposure, diabetes, obesity, and stroke.
† Also adjusted for severity of nuclear opacity.
‡ Low hyperopia, greater than ⫹1 D to less than ⫹2 D; moderate hyperopia, ⫹2 D to less than ⫹4 D; high hyperopia, ⫹4 D or greater.
§ Low myopia, less than ⫺1 D to more than ⫺3.5 D; moderate myopia, ⫺3.5 D to more than ⫺6 D; high myopia, ⫺6 D or less.
3628 Younan et al. IOVS, December 2002, Vol. 43, No. 12

TABLE 3. Odds Ratios and 95% Confidence Intervals (CI) for the Association between Baseline
Refraction and Incident Nuclear Cataract in All Eyes

Multivariate
Age and Sex Adjusted Adjusted*

Eyes at Incident Odds Odds

Risk (n) eyes (n) % Ratio 95% CI Ratio 95% CI

Emmetropia 1243 235 18.9 1.0 Referent 1.0 Referent

Any hyperopia 1036 312 30.1 1.2 1.0–1.5 1.1 0.9–1.5
Hyperopia†
Low 536 134 25.0 1.1 0.8–1.4 1.0 0.8–1.3
Moderate 438 163 37.2 1.6 1.2–2.1 1.4 1.1–1.9
High 62 15 24.2 0.9 0.5–1.8 0.9 0.5–1.8
Any myopia 256 46 18.0 1.1 0.7–1.6 1.0 0.7–1.5
Myopia‡
Low 151 21 13.9 0.8 0.5–1.3 0.7 0.4–1.2
Moderate 71 14 19.7 1.3 0.6–2.6 1.2 0.5–2.7
High 34 11 32.4 3.1 1.5–6.5 3.3 1.5–7.4

* Adjusted for age, sex, smoking, education, dark brown iris color, and inhaled steroids.
† Low hyperopia, greater than ⫹1 D to less than ⫹2 D; moderate hyperopia, ⫹2 D to less than ⫹4
D; high hyperopia, ⫹4 D or greater.
‡ Low myopia, ⫺1 D to more than ⫺3.5 D; moderate myopia, ⫺3.5 D to more than ⫺6 D; high
myopia, ⫺6 D or less.

although this was not found for right eyes (OR 1.4, 95% CI
0.4 –5.4). Although the odds for posterior subcapsular cataract
were increased in low myopia, only the association with mod-
erate or high myopia was statistically significant, after adjust-
ment for age and sex (OR 5.7, 95% CI 2.8 –11.6), and in the
multivariate model (OR 5.4, 95% CI 2.5–11.9). After adjustment
was made for severity of nuclear opacity, the associations
between any myopia (OR 2.1, 95% CI 1.0 – 4.8) and moderate
or high myopia (OR 4.4, 95% CI 1.7–11.5) remained strong and
statistically significant (Table 4). There was no statistically
significant association between hyperopia and incident poste-
rior subcapsular cataract.
The presence of any myopia was strongly associated with
incident cataract surgery in age- and sex-adjusted analyses (OR
2.8, 95% CI 1.8 – 4.5) and there were statistically significant
associations of low, moderate, and high myopia with incident
cataract surgery (Table 5). All findings remained statistically
significant, after adjustment for multiple potential confound-
ers. Multivariate models also showed statistically significant
associations between myopia and incident cataract surgery in
logistic regression analyses for right (OR 2.3, 95% CI 1.0 –5.6)
and left eyes separately (OR 4.1, 95% CI 1.8 –9.2). No signifi-
cant associations were found between hyperopia and incident
cataract surgery.
In attempting to quantify the effect of duration of myopia,
we used the age that participants stated as that at which they
had first worn glasses for distance vision. Duration of myopia
was not associated with incident cortical or nuclear cataract
(data not shown). Myopic subjects who began wearing dis-
tance glasses before age 40 had a significantly higher incidence
of posterior subcapsular cataract after adjustment for age and
sex. This was found both for those first wearing glasses before
age 20 (OR 3.5, 95% CI 1.7–7.3) and for those first wearing
glasses between ages 20 and 39 years (OR 4.5, 95% CI 1.8 –
11.5). The relationship, however, was only statistically signifi-
cant for the subgroup wearing distance glasses for the longest
period (Table 6), when multiple confounders were controlled
for, including the level of nuclear opacity. Both the longest-

TABLE 4. Odds Ratios and 95% Confidence Intervals (CI) for the Association between Baseline Refraction and Incident Posterior Subcapsular
Cataract in All Eyes

Multivariate Multivariate
Age and Sex Adjusted Adjusted* Adjusted†

Eyes at Incident Odds Odds Odds

Risk (n) Eyes (n) % Ratio 95% CI Ratio 95% CI Ratio 95% CI

Emmetropia 1925 41 2.1 1.0 Referent 1.0 Referent 1.0 Referent

Any hyperopia 1647 38 2.3 0.8 0.5–1.4 0.8 0.5–1.4 1.0 0.5–1.8
Hyperopia‡
Low 825 21 2.6 1.0 0.6–1.7 1.0 0.6–1.8 1.3 0.7–2.5
Moderate 721 13 1.8 0.6 0.3–1.2 0.6 0.3–1.2 0.6 0.2–1.4
High 101 4 4.0 1.6 0.6–4.5 1.7 0.6–5.1 2.3 0.7–8.1
Any myopia 388 21 5.4 2.9 1.7–5.2 2.6 1.4–5.0 2.1 1.0–4.8
Myopia§
Low 240 8 3.3 1.9 0.9–4.2 1.9 0.8–4.7 1.3 0.4–4.3
Moderate/high 148 13 8.8 5.7 2.8–11.6 5.4 2.5–11.9 4.4 1.7–11.5

* Adjusted for age, sex, education, obesity, and hypertension.

† Also adjusted for severity of nuclear opacity.
‡ Low hyperopia, greater than ⫹1 D to less than ⫹2 D; moderate hyperopia, ⫹2 D to less than ⫹4 D; high hyperopia, ⫹4 D or greater.
§ Low myopia, less than ⫺1 D to more than ⫺3.5 D; moderate myopia, ⫺3.5 D to more than ⫺6 D; high myopia, ⫺6 D or less.
IOVS, December 2002, Vol. 43, No. 12 Myopia and Incident Cataract and Cataract Surgery 3629

TABLE 5. Odds Ratios and 95% Confidence Intervals (CI) for the Association between Baseline Refraction and Incident Cataract Surgery in All Eyes

Multivariate Multivariate
Age and Sex Adjusted Adjusted* Adjusted†

Eyes at Incident Odds Odds Odds

Risk (n) Eyes (n) % Ratio 95% CI Ratio 95% CI Ratio 95% CI

Emmetropia 2137 76 3.6 1.0 Referent 1.0 Referent 1.0 Referent

Any hyperopia 1875 84 4.5 0.8 0.5–1.1 0.8 0.6–1.2 0.8 0.5–1.3
Hyperopia‡
Low 938 35 3.7 0.8 0.5–1.2 0.9 0.6–1.3 0.8 0.5–1.5
Moderate 827 44 5.3 0.8 0.5–1.2 0.8 0.5–1.4 0.9 0.5–1.6
High 110 5 4.6 0.9 0.3–2.4 1.3 0.5–3.1 0.9 0.2–4.1
Any myopia 479 50 10.4 2.8 1.8–4.5 3.2 1.9–5.3 2.1 1.1–4.2
Myopia§
Low 294 30 10.2 2.5 1.4–4.5 2.7 1.3–5.3 1.5 0.5–4.3
Moderate 113 8 7.1 2.7 1.2–5.8 3.6 1.5–8.6 2.9 1.2–7.3
High 72 12 16.7 5.0 2.4–10.3 5.5 2.3–13.2 3.4 1.0–11.3

* Adjusted for age, sex, education, inhaled steroids, diabetes, dark brown iris color, and angina.
† Also adjusted for severity of nuclear opacity.
‡ Low hyperopia, greater than ⫹1 D to less than ⫹2 D; moderate hyperopia, ⫹2 D to less than ⫹4 D; high hyperopia, ⫹4 D or greater.
§ Low myopia, less than ⫺1 D to more than ⫺3.5 D; moderate myopia, ⫺3.5 D to more than ⫺6 D; high myopia, ⫺6 D or less.

and shortest-duration subgroups had statistically significant as-
sociations with incident cataract surgery after adjustment for
age and sex and for multiple confounders, but not when the
level of nuclear opacity was included in the model (Table 7).
DISCUSSION
Our previous report of cross-sectional associations between
refractive error and cataract suggested a strong relationship
between myopia and posterior subcapsular cataract that re-
flected the apparent duration of myopia.10 The present report
of incident posterior subcapsular cataract provides support for
this hypothesis. Because posterior subcapsular cataract rapidly
progresses to cataract surgery, the significant relationship
found between myopia and incident cataract surgery, particu-
larly for long-duration myopes, adds further support.
An important limitation of this report is that using the age at
which distance glasses were first worn as a proxy for onset of
myopia does not take into account the many factors that
impact on the decision to start wearing glasses. As pointed out
by Wong et al.11 in the Beaver Dam Eye Study report on the
relationship between refractive errors and incident cataract,
poor memory and interpretation of this question may be rele-
vant factors and may have played a role in the reported incon-
sistencies between a history of wearing distance glasses, the
age at first wearing, and prevalent age-related cataract.11,33
Similarly, the decision to undergo cataract surgery, particu-
larly on a second eye, takes into account many factors other
than severity of cataract. In our effort to counter the possibility
that statistically significant associations with myopia may have
been due to the presence of baseline nuclear cataract, inter-
pretation of those adjusted results was limited by the fact that
a significant proportion of participants had missing baseline
data for nuclear cataract because of random camera malfunc-
tion (so reducing the power of our analyses). Our definitions of
incident cortical and nuclear cataract may also pose a limita-
tion. Presence of less than 5% baseline cortical opacity and
baseline nuclear grades 1 to 3 were not considered to be
cataract. Although we sought to encompass a level of clinically
significant cataract to define cataract incidence, inclusion of
any baseline cataract may have had a minor influence on our
findings.
Strengths of our study include its high participation rate
from a well-defined urban residential population. Information
was collected about a large number of potential confounders as
well as detailed information on refractive status.1 Documenta-
tion of cataract status, based on reproducible grading of lens
photographs according to the Wisconsin System, is a further
strength. Graders of the prospective lens photographs were
masked to refractive state and so could not be influenced by
selection bias.
The development of age-related cataract is widely known to
be associated with a myopic shift in refraction, principally by
progression of the level of opacity within the lens nucleus.9
The Visual Impairment Project demonstrated a strong cross-
sectional association between myopia and nuclear opacity,24,34

TABLE 6. Odds Ratios and 95% Confidence Intervals (CI) for the Association between Baseline Age at Which Distance Glasses Were First Worn
in Myopes and Incident Posterior Subcapsular Cataract in All Eyes

Multivariate Multivariate
Age and Sex Adjusted Adjusted* Adjusted†

Age Distance Glasses Eyes at Incident Odds Odds Odds

First Worn (y) Risk (n) Eyes (n) % Ratio 95% CI Ratio 95% CI Ratio 95% CI

Never 1925 41 2.1 1.0 Referent 1.0 Referent 1.0 Referent

40⫹ 58 3 5.2 2.3 0.7–7.5 2.0 0.5–7.8 2.8 0.6–13.1
20–39 96 7 7.3 4.5 1.8–11.5 3.8 1.3–10.6 2.4 0.5–11.8
0–19 202 11 5.5 3.5 1.7–7.3 3.6 1.5–8.8 3.0 1.0–9.3

* Adjusted for age, sex, education, inhaled steroids, diabetes, dark brown iris color, obesity, and hypertension.
† Also adjusted for severity of nuclear opacity.
3630 Younan et al.
TABLE 7. Odds Ratios and 95% Confidence Intervals (CI) for the Association between Baseline Age at Which Distance Glasses Were First Worn
in Myopes and Incident Cataract Surgery in All Eyes
Age and Sex Adjusted
Age Distance Glasses Eyes at Incident Odds
First Worn (y) Risk (n) Eyes (n) % Ratio
Never 2137 76 3.6 1.0
40⫹ 86 23 26.7 4.7
20–39 112 5 4.5 1.3
0–19 238 15 6.3 2.5
* Adjusted for age, sex, education, inhaled steroids, diabetes, dark brown iris color, and angina.
† Also adjusted for severity of nuclear opacity.
as did the Barbados Eye Survey.35 In the 5-year examinations of
the Beaver Dam Eye Study, participants with the highest levels
of nuclear opacity at baseline were more likely to have a
myopic shift in refraction.36
As cataract is frequently mixed, it is reasonable to adjust for
the level of nuclear opacity when examining the impact of
myopia on development of other cataract types. Our study
indicates that myopia was related to incident posterior subcap-
sular cataract (as well as incident cataract surgery), after taking
into account the effects both from multiple confounders plus
the level of nuclear opacity. Adjustment for level of nuclear
opacity, however, may be less important in assessing longitu-
dinal associations, as we used the baseline refractive state to
evaluate the impact of refractive error on development of
cataract over time. The moderately strong relationship found
between incident posterior subcapsular cataract and either
moderate or high myopia at baseline argues that myopic shift
may not be a relevant influence. It could be expected that
further cataract-associated myopic shifts would be of a low
magnitude.
The apparent relationship found in our study between du-
ration of myopia and both incident posterior subcapsular cat-
aract and incident surgery provides further support for a true
association between myopia and this cataract type. This was
seen for both longer-duration subgroups, but was strongest and
statistically significant only in the fully adjusted model for those
myopic subjects who reported wearing glasses for the longest
period.
Apart from the association found with posterior subcapsular
cataract, our data provide no evidence for a relationship be-
tween myopia, or other refractive error, and cortical cataract.
Incident nuclear cataract was unrelated to baseline low or
moderate magnitude myopia, but a relationship was found
with high myopia, after adjustment for confounders. A weak
relationship was observed between incident nuclear cataract
and moderate hyperopia. There was, however, no statistically
significant relationship with any hyperopia, and the odds ratios
in the hyperopia categories did not show a dose–response
relationship, suggesting that this finding could be spurious.
Although our prevalence findings suggested a protective
influence of hyperopia for posterior subcapsular cataract,10
this was not supported longitudinally, either in our current
data or from the Beaver Dam Eye Study data.11 the Beaver Dam
findings reported a significant trend (P ⫽ 0.02) for increasing
myopia and higher prevalence of posterior subcapsular cata-
ract. This was not, however, supported by their incidence data.
In contrast, our longitudinal data supported the cross-sectional
associations found with any myopia or high myopia and in
persons with onset of myopia in their youth.
Small posterior subcapsular opacities may cause significant
visual disturbance because of their central location in the visual
axis. Because these opacities thus have a propensity to
IOVS, December 2002, Vol. 43, No. 12
Multivariate Multivariate
Adjusted* Adjusted†
Odds Odds
95% CI Ratio 95% CI Ratio 95% CI
Referent 1.0 Referent 1.0 Referent
2.3–9.4 4.9 2.2–11.1 2.2 0.8–6.3
0.3–5.6 1.4 0.2–8.8 1.1 0.1–13.0
1.3–4.8 3.0 1.5–6.3 2.0 0.8–5.3
progress to cataract surgery relatively quickly compared with
nuclear or cortical opacities, it seems reasonable to consider
incident cataract surgery as a surrogate for the development of
posterior subcapsular cataract. The number of incident poste-
rior subcapsular cataract cases may also be limited because of
varying thresholds for cataract surgery.
The Beaver Dam Eye Study found a statistically significant
association between baseline myopia and 5-year incident cata-
ract surgery,11,12 as well as a significant trend (P ⫽ 0.003) for
the relationship between increasing levels of myopia and in-
creased incidence of cataract surgery.11 Our longitudinal re-
sults provide strong support for this finding, with significant
associations present between myopia at all levels and in all
models examined, apart from low myopia in the multivariate
model that adjusted for severity of nuclear cataract.
Our finding of a link between myopia and cataract is sup-
ported by data from several other studies.2,24,37– 41 The Visual
Impairment Project reported a statistically significant associa-
tion between myopia (defined as ⱖ1 D) and cortical, nuclear,
and posterior subcapsular cataract.24,37 The Melton Mowbray
Eye Study reported a higher prevalence of cataract in
myopes.38 Participants who had worn eye glasses before age
20 (used as an indicator for myopia) had a higher relative risk
(RR) for nuclear cataract in the Longitudinal Study of Cataract
(RR 1.37, 95% CI 0.97–1.95),40 and a higher risk of develop-
ment of mixed cataract in the Lens Opacities Case–Control
Study (OR 1.44, 95% CI 1.06 –1.94).39 In an Oxfordshire case–
control study of patients who had had cataract surgery and
control subjects aged between 50 and 79 years,42 a statistically
significant increased risk of cataract was found in participants
with a history of childhood myopia (RR 1.68, 95% CI 1.2–2.4).
In several reports in which the records of patients who had
undergone cataract surgery were reviewed,7,8 myopes were
found to be more likely to undergo surgery than nonmyopes.
Biological plausibility of this association has been provided
by a number of laboratory-based studies. Development of cat-
aract in myopic lenses has been shown to be related to oxida-
tive changes in lens proteins,43 with glutathione a potential
inhibitor of this oxidation.44 Compared with healthy control
subjects, cataractous lenses had lower levels of glutathione,
with the lowest levels found in myopic lenses.13 Increased
levels of malondialdehyde (MDA) have been found in catarac-
tous lenses and in the vitreous of myopes compared with
control subjects and nonmyopic cataract lenses.13,14 Because
MDA is a breakdown product of lipid peroxidation, the vitre-
ous finding suggests a retinal source. Rod outer segments are
particularly susceptible to lipid peroxidation because of the
high concentration of polyunsaturated lipid in their mem-
branes.
One study has shown a correlation between degree of
retinal lipid peroxidation and extent of lens damage.15 A sub-
sequent study has shown a correlation between the level of
IOVS, December 2002, Vol. 43, No. 12
thiobarbituric acid reactive substances (also indicating lipid
peroxidation) in the subretinal fluid of patients who undergo
retinal detachment surgery and the degree of myopia.45 These
results suggest that increasing myopia may be related to in-
creasing damage of rod outer segments and that by-products of
this process may affect various ocular structures, including the
lens. This could explain the higher prevalence and incidence,
particularly of posterior subcapsular cataract and nuclear cat-
aract, in myopic eyes.
Myopia increases the risk of glaucoma35,46 as well as retinal
detachment, myopic retinal degeneration, visual impairment,
and blindness.47 The present study builds on our earlier find-
ing10 in suggesting that myopia may increase the risk of pos-
terior subcapsular cataract, one of the important predictors of
cataract surgery. This finding is potentially important, given
the increasing prevalence of myopia in East Asia,48,49 the
United States,50 and elsewhere.51
In this study, we assessed the association between baseline
refraction and the 5-year incidence of cataract and cataract
surgery. We have been able to provide epidemiologic evidence
to confirm the previously noted increased risk of posterior
subcapsular cataract in eyes with high myopia. We have also
provided evidence to support the hypothesis that this type of
cataract may develop more frequently in eyes with lower levels
of myopia. These longitudinal data support our cross-sectional
findings as well as the findings from some, but not all,11
population-based reports. Previously, this association was con-
sidered because of the myopic shift from increasing nuclear
opacity and was thought unlikely to be causally related.37 Our
data suggest the possibility of a causal relationship. Although
some laboratory-based studies support a plausible biological
basis, further studies are needed.
References
1. Attebo K, Ivers RQ, Mitchell P. Refractive errors in an older
population: the Blue Mountains Eye Study. Ophthalmology. 1999;
106:1066 –1072.
2. Wong TY, Foster PJ, Hee J, et al. Prevalence and risk factors for
refractive errors in adult Chinese in Singapore. Invest Ophthalmol
Vis Sci. 2000;41:2486 –2494.
3. Mitchell P, Cumming RG, Attebo K, Panchapakesan J. Prevalence
of cataract in Australia: the Blue Mountains eye study. Ophthal-
mology. 1997;104:581–588.
4. Klein BE, Klein R, Linton KL. Prevalence of age-related lens opac-
ities in a population: the Beaver Dam Eye Study. Ophthalmology.
1992;99:546 –552.
5. Klein BE, Klein R, Lee KE. Incidence of age-related cataract: the
Beaver Dam Eye Study. Arch Ophthalmol. 1998;116:219 –225.
6. Duke-Elder S: Diseases of the lens and vitreous; glaucoma and
hypotony. System of Ophthalmology. London: Kimpton; 1970:
225.
7. Perkins ES. Cataract: refractive error, diabetes, and morphology.
Br J Ophthalmol. 1984;68:293–297.
8. Weale R. A note on a possible relation between refraction and a
disposition for senile nuclear cataract. Br J Ophthalmol. 1980;64:
311–314.
9. Brown NA, Hill AR. Cataract: the relation between myopia and
cataract morphology. Br J Ophthalmol. 1987;71:405– 414.
10. Lim R, Mitchell P, Cumming RG. Refractive associations with
cataract: the Blue Mountains Eye Study. Invest Ophthalmol Vis Sci.
1999;40:3021–3026.
11. Wong TY, Klein BE, Klein R, Tomany SC, Lee KE. Refractive errors
and incident cataracts: the Beaver Dam Eye Study. Invest Ophthal-
mol Vis Sci. 2001;42:1449 –1454.
12. Klein BE, Klein R, Moss SE. Incident cataract surgery: the Beaver
Dam eye study. Ophthalmology. 1997;104:573–580.
13. Micelli Ferrari T, Vendemiale G, Grattagliano I, et al. Role of lipid
peroxidation in the pathogenesis of myopic and senile cataract.
Br J Ophthalmol. 1996;80:840 – 843.
Myopia and Incident Cataract and Cataract Surgery 3631
14. Simonelli F, Nesti A, Pensa M, et al. Lipid peroxidation and human
cataractogenesis in diabetes and severe myopia. Exp Eye Res.
1989;49:181–187.
15. Zigler JS Jr, Bodaness RS, Gery I, Kinoshita JH. Effects of lipid
peroxidation products on the rat lens in organ culture: a possible
mechanism of cataract initiation in retinal degenerative disease.
Arch Biochem Biophys. 1983;225:149 –156.
16. Mitchell P, Smith W, Attebo K, Wang JJ. Prevalence of age-related
maculopathy in Australia: the Blue Mountains Eye Study. Ophthal-
mology. 1995;102:1450 –1460.
17. Attebo K, Mitchell P, Smith W. Visual acuity and the causes of
visual loss in Australia: the Blue Mountains Eye Study. Ophthal-
mology. 1996;103:357–364.
18. Klein R, Klein BE, Linton KL, De Mets DL. the Beaver Dam Eye
Study: visual acuity. Ophthalmology. 1991;98:1310 –1315.
19. Mitchell P, Smith W, Wang JJ. Iris color, skin sun sensitivity, and
age-related maculopathy: the Blue Mountains Eye Study. Ophthal-
mology. 1998;105:1359 –1363.
20. Klein BE, Klein R, Linton KL, Magli YL, Neider MW. Assessment of
cataracts from photographs in the Beaver Dam Eye Study. Oph-
thalmology. 1990;97:1428 –1433.
21. Zeger SL, Liang KY, Albert PS. Models for longitudinal data: a
generalized estimating equation approach [published correction
appears in Biometrics. 1989;45:347]. Biometrics. 1988;44:1049 –
1060.
22. Liang KY, Zeger SL. Longitudinal data analysis using generalized
linear models. Biometrika. 1986;73:13–22.
23. Hiller R, Sperduto RD, Ederer F. Epidemiologic associations with
nuclear, cortical, and posterior subcapsular cataracts. Am J Epide-
miol. 1986;124:916 –925.
24. McCarty CA, Mukesh BN, Fu CL, Taylor HR. The epidemiology of
cataract in Australia. Am J Ophthalmol. 1999;128:446 – 465.
25. Delcourt C, Cristol JP, Tessier F, Leger CL, Michel F, Papoz L. Risk
factors for cortical, nuclear, and posterior subcapsular cataracts:
the POLA study: Pathologies Oculaires Liees a l’Age. Am J Epide-
miol. 2000;151:497–504.
26. Leske MC, Wu SY, Nemesure B, Li X, Hennis A, Connell AM.
Incidence and progression of lens opacities in the Barbados Eye
Studies. Ophthalmology. 2000;107:1267–1273.
27. Cumming RG, Mitchell P. Alcohol, smoking, and cataracts: the
Blue Mountains Eye Study. Arch Ophthalmol. 1997;115:1296 –
1303.
28. West SK, Duncan DD, Munoz B, et al. Sunlight exposure and risk
of lens opacities in a population-based study: the Salisbury Eye
Evaluation project. JAMA. 1998;280:714 –718.
29. Goodrich ME, Cumming RG, Mitchell P, Koutts J, Burnett L. Plasma
fibrinogen and other cardiovascular disease risk factors and cata-
ract. Ophthalmic Epidemiol. 1999;6:279 –290.
30. Klein BE, Klein R, Lee KE. Diabetes, cardiovascular disease, se-
lected cardiovascular disease risk factors, and the 5-year incidence
of age-related cataract and progression of lens opacities: the Bea-
ver Dam Eye Study. Am J Ophthalmol. 1998;126:782–790.
31. Cumming RG, Mitchell P. Use of inhaled corticosteroids and the
risk of cataracts. N Engl J Med. 1997;337:8 –14.
32. Cumming RG, Mitchell P, Lim R. Iris color and cataract: the Blue
Mountains Eye Study. Am J Ophthalmol. 2000;130:237–238.
33. Cruickshanks KJ, Klein BE, Klein R. Ultraviolet light exposure and
lens opacities: the Beaver Dam Eye Study. Am J Public Health.
1992;82:1658 –1662.
34. Wensor M, McCarty CA, Taylor HR. Prevalence and risk factors of
myopia in Victoria, Australia. Arch Ophthalmol. 1999;117:658 –
663.
35. Wu SY, Nemesure B, Leske MC. Refractive errors in a black adult
population: the Barbados Eye Study. Invest Ophthalmol Vis Sci.
1999;40:2179 –2184.
36. Lee KE, Klein BE, Klein R. Changes in refractive error over a 5-year
interval in the Beaver Dam Eye Study. Invest Ophthalmol Vis Sci.
1999;40:1645–1649.
37. McCarty CA, Nanjan MB, Taylor HR. Attributable risk estimates for
cataract to prioritize medical and public health action. Invest
Ophthalmol Vis Sci. 2000;41:3720 –3725.
3632 Younan et al.
38. Smith GD, Egger M. Socioeconomic differentials in wealth and
health (editorial). BMJ. 1993;307:1085–1086.
39. Leske MC, Chylack LT Jr, Wu SY. The Lens Opacities Case-Control
Study: risk factors for cataract. Arch Ophthalmol. 1991;109:244 –
251.
40. Leske MC, Chylack LT Jr, He Q, et al. Risk factors for nuclear
opalescence in a longitudinal study. LSC Group: Longitudinal
Study of Cataract. Am J Epidemiol. 1998;147:36 – 41.
41. Dandona R, Dandona L, Naduvilath TJ, Srinivas M, McCarty CA,
Rao GN. Refractive errors in an urban population in Southern
India: the Andhra Pradesh Eye Disease Study. Invest Ophthalmol
Vis Sci. 1999;40:2810 –2818.
42. Harding JJ, Harding RS, Egerton M. Risk factors for cataract in
Oxfordshire: diabetes, peripheral neuropathy, myopia, glaucoma
and diarrhoea. Acta Ophthalmol Copenh. 1989;67:510 –517.
43. Boscia F, Grattagliano I, Vendemiale G, Micelli-Ferrari T, Altomare
E. Protein oxidation and lens opacity in humans. Invest Ophthal-
mol Vis Sci. 2000;41:2461–2465.
44. Kamei A. Glutathione levels of the human crystalline lens in aging
and its antioxidant effect against the oxidation of lens proteins.
Biol Pharm Bull. 1993;16:870 – 875.
IOVS, December 2002, Vol. 43, No. 12
45. Bosch-Morell F, Sanz A, Diaz-Llopis M, Romero FJ. Lipid peroxida-
tion products in human subretinal fluid. Free Radic Biol Med.
1996;20:899 –903.
46. Mitchell P, Hourihan F, Sandbach J, Wang JJ. The relationship
between glaucoma and myopia: the Blue Mountains Eye Study.
Ophthalmology. 1999;106:2010 –2015.
47. Curtin B. Basic science and clinical management. In: The Myopias:
Basic Science and Clinical Management. Philadelphia: Harper
and Row; 1985;39 –59.
48. Tay MT, Au Eong KG, Ng CY, Lim MK. Myopia and educational
attainment in 421,116 young Singaporean males. Ann Acad Med
Singapore. 1992;21:785–791.
49. Matsumura H, Hirai H: Prevalence of myopia and refractive
changes in students from 3 to 17 years of age. Surv Ophthalmol.
1999;44(suppl 1):S109 –S115.
50. Familial aggregation and prevalence of myopia in the Framingham
Offspring Eye Study: the Framingham Offspring Eye Study Group.
Arch Ophthalmol. 1996;114:326 –332.
51. Rose K, Smith W, Morgan I, Mitchell P. The increasing prevalence
of myopia: implications for Australia. Clin Exp Ophthalmol. 2001;
29:116 –120.