Is There an Association Between Cataract

Surgery and Age-related Macular

Degeneration? Data From Three
Population-based Studies
ELLEN E. FREEMAN, MSc, BEATRIZ MUNOZ, MSC, SHEILA K. WEST, PHD,
JAMES M. TIELSCH, PHD, AND OLIVER D. SCHEIN, MD, MPH

PURPOSE: To determine whether cataract surgery is

associated with an increased prevalence of age-related
macular degeneration (AMD) in three independent population-based data sets.
DESIGN: Cross-sectional study.
METHODS: Data were used from the Salisbury Eye
Evaluation (2,520 subjects from Salisbury, Maryland,
aged 65 to 84 years), the Proyecto VER (4,774 Hispanic
subjects from Arizona aged 40 years and older), and the
Baltimore Eye Survey (4,396 subjects from Baltimore,
Maryland, aged 40 and older). The main outcome measure was AMD as determined by retinal photographs or
clinical examination.
RESULTS: A history of cataract surgery was associated
with an increased prevalence of late AMD in all three
data sets after adjusting for age, race, sex, and smoking,
but odds ratios (OR) were not individually statistically
significant. The OR for the combined analysis was 1.7
(95% confidence interval: 1.12.6). Having a severe
cataract in the eye was also associated with a slightly
higher prevalence of late AMD, although the combined
OR was not statistically significant (OR 1.4; 95%
confidence interval, 0.8%2.4). Overall, increasing time
since cataract surgery was not associated with late AMD.

Accepted for publication Dec 10, 2002.

InternetAdvance publication at ajo.com Dec 12, 2002.
From the Dana Center for Preventive Ophthalmology (E.E.F., B.M.,
S.K.W., J.M.T., O.D.S.), Wilmer Eye Institute, Johns Hopkins University
School of Medicine, and the International Health Department (J.M.T.),
Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland.
This research was supported in part by an unrestricted grant from
Alcon Laboratories, Inc, Fort Worth, Texas; NEI Grants K24EY00395,
EY11283; and NIA Grant AG16294. Doctor West is a Research to
Prevent Blindness Senior Scientific Investigator. Doctor Schein is a
consultant to Alcon Research, Ltd.
Inquiries to Oliver Schein, MD, 116 Wilmer Bldg, Johns Hopkins
Hospital, 600 N Wolfe St, Baltimore, MD 21287-9019; fax: (410)
614-9651; e-mail: oschein@jhmi.edu
0002-9394/03/$30.00
doi:10.1016/S0002-9394(02)02253-5

2003 BY

CONCLUSIONS: A history of cataract surgery may be

associated with an increased prevalence of late AMD.
However, having a severe cataract in the eye may also be
associated with a higher prevalence of late AMD. Additional research is needed to investigate whether a
causal relationship exists between cataract surgery and
AMD or whether this relationship is due to residual
confounding or bias. (Am J Ophthalmol 2003;135:
849 856. 2003 by Elsevier Inc. All rights reserved.)

HE MOST COMMON CAUSE OF VISION LOSS IN THE

elderly in the United States is age-related cataract,1,2 whereas the most common cause of blindness
in this group is age-related macular degeneration
(AMD).3,4 The visual prognosis for patients with cataract
is very good. In fact, 96% have improved vision 4 months
after surgery.5 In contrast, the prognosis for patients with
AMD is less optimistic, as there is only proven treatment
for the less common neovascular form.
Cataract surgery is the most common procedure performed in the Medicare population, with more than 1
million cataract surgeries performed each year in the
United States.6 Several large epidemiologic studies have
assessed the question of a possible association between
cataract surgery and late AMD.710 The Beaver Dam Eye
Study, a population-based prospective study of 3,684 adults
aged 40 years and older, found that persons who had
previous cataract surgery had a higher 10-year risk of
developing incident late age-related maculopathy (odds
ratio [OR] 3.8; 95% confidence interval [CI], 1.9 7.7).8
However, the Blue Mountain Eye Study, a populationbased survey of 3,654 adults aged 49 years and older, did
not find an association.9 In addition, the Visual Impairment Project did not find an association between cataract
surgery and late AMD after adjusting for other variables.10
Some studies have found an association between the
presence of a lens opacity in the eye and late AMD,

ELSEVIER INC. ALL

RIGHTS RESERVED.

849

macular zone. Subjects who had late AMD were excluded

from analyses of early AMD. Presence of severe cataract
was defined as nuclear sclerosis grade 4 (highest grade),
cortical opacification of 8/16 or greater, or any posterior
subcapsular opacification as graded by lens photographs
using the Wilmer grading scheme.14,15 Persons excluded
from the analyses were those who did not have photographs taken owing to camera failure, inability to dilate the
pupil, refusal of the participant, significant corneal opacity,
inability to maintain fixation, or photographs that were
not of gradable quality for the specific outcome under
study. Of 5,040 eyes, there were 4,641 eyes with gradable
photographs for late AMD, 4,022 for early AMD 1, and
4,489 for early AMD 2.
The PVER is a population-based study of 4,774 Hispanic
persons over age 40 years living in either Pima County or
Santa Cruz County in Arizona. Of those subjects identified
as eligible, 4,774 of 6,668 (72%) completed both the home
questionnaire and the clinical examination. Further details
about the sample and recruitment have been previously
described.16 Data were collected on history of previous
cataract surgery by clinical examination. Date of cataract
surgery was not collected. Subjects were identified in a
clinical examination as having choroidal neovascularization or geographic atrophy. In addition, data were collected on whether late AMD was a cause of blindness
(vision 20/200 or worse). Early AMD was not assessed.
Information was also collected on age, sex, and smoking
habits through an interviewer-administered questionnaire.
For this analysis, late AMD was defined as choriodal
neovascularization or geographic atrophy present as assessed
by clinical examination, or if late AMD was determined to be
a cause of blindness (20/200 or worse). Presence of severe
cataract was defined as nuclear sclerosis grade 4, cortical
opacification 8/16 or greater, or any PSC opacification as
graded clinically.14 Late AMD was clinically assessed in 9,478
of 9,548 eyes. Reasons for not being able to assess late AMD
included inability to dilate the pupil, refusal of the participant, and significant corneal opacity.
The Baltimore Eye Survey (BES), also population-based,
is a study of ocular disease and vision loss among 5,308
African American and Caucasian residents of Baltimore,
Maryland, aged 40 years or older. Of subjects selected to
participate, 5,308 of 7,265 (73%) completed the screening
examination. All were then offered stereoscopic fundus
photography, and 4,396 persons who agreed to have
photography had at least one gradable photograph that was
included in this analysis. Further details have been previously described.17 History of cataract surgery was obtained
by clinical examination. Those who had previous cataract
surgery (n 138) were asked to report their age at cataract
surgery. Because we knew their age at the baseline examination, the difference in the two ages provided us with the
time since cataract surgery. Photographs of the retina were
reviewed in a masked fashion and graded for signs of
choroidal neovascularization, geographic atrophy, and

suggesting that cataract and late AMD may share one or

more common risk factors.7,11 Why cataract surgery and
late AMD may be associated is still unknown, although
some have suggested that surgery may increase the risk of
late AMD in susceptible eyes.11,12 Therefore, to evaluate
whether cataract surgery and late AMD are associated, we
examined existing data from three independent population-based studies to maximize the number of subjects with
late AMD.

DESIGN
THE THREE POPULATION-BASED STUDIES TO BE USED FOR

this cross-sectional analysis, the Salisbury Eye Evaluation

(SEE), Proyecto VER (PVER), and the Baltimore Eye
Survey (BES), were conducted according to the guidelines
established in the Declaration of Helsinki. In addition, the
projects were approved by the appropriate Internal Review
Boards of The Johns Hopkins University. All subjects
provided written informed consent.

METHODS
THE SEE PROJECT IDENTIFIED A POPULATION-BASED RAN-

dom sample of elderly persons between the ages of 65 and

84 years living in Salisbury, Maryland, in 1993. Details
about sample recruitment and methods are described
elsewhere.13 A total of 65% of those invited to participate
completed both the home questionnaire and the medical
examination, resulting in 2,520 participants. Data were
collected on history of previous cataract surgery by clinical
examination. At the clinical examination, photographs of
the retina were taken and then assessed in a masked
fashion by trained graders for signs of AMD such as
choroidal neovascularization, geographic atrophy, hyperpigmentation, nongeographic atrophy, and drusen. Information on age, sex, race, and smoking habits were also
collected by interviewer-administered questionnaire.
There were 471 participants who had cataract surgery on
718 eyes. Information on when cataract surgery occurred
was obtained after the initial examination by review of
medical records (232 subjects, or 49%) or by telephone
calls to participants (98 subjects, or 21%). Date of cataract
surgery could not be obtained for 141 participants (123
eyes, 17%) because of inability to obtain medical records,
lack of information in records of cataract surgery date, or
inability to reach the participant by telephone.
Late AMD was defined for this study as the presence of
signs of choroidal neovascularization or geographic atrophy. Two definitions of early AMD were used: definition 1
(AMD 1), drusen 64 m or greater, hyperpigmentation, or
nongeographic atrophy in the central macular zone; and
definition 2 (AMD 2), drusen 125 m or greater, hyperpigmentation, or nongeographic atrophy in the central
850

AMERICAN JOURNAL

OF

OPHTHALMOLOGY

JUNE 2003

signs of early AMD such as drusen, hyperpigmentation, or

nongeographic atrophy. Also, during the clinical examination, AMD was assessed as a cause of blindness. During
the personal interview, information was collected on age,
race, sex, and smoking habits.
The definition of late AMD for this study was choroidal
neovascularization, geographic atrophy, or if AMD was
diagnosed by an ophthalmologist as a cause of blindness
(20/200 or worse). The two definitions of early AMD were
used (AMD 1: drusen 64 m or greater, hyperpigmentation, or nongeographic atrophy; AMD 2: drusen 125 m or
greater, hyperpigmentation, or nongeographic atrophy in
the macular zone). Those who had late AMD were
excluded from analyses of early AMD. Presence of a severe
cataract was defined if visual impairment or blindness was
attributed to cataract in the clinical examination. Eyes
excluded from the analyses were those whose photographs
were not of gradable quality for the specific outcome under
study. Of 8,792 eyes, 7,428 were gradable for late AMD,
8,734 eyes were gradable for early AMD 1, and 8,717 eyes
were gradable for early AMD 2.
The general characteristics of the three study populations were examined and compared. Next, the characteristics of those who had cataract surgery in at least 1 eye were
compared with those who did not for the three populations.
Age- specific and race-specific prevalence rates of having late
AMD in at least 1 eye were calculated. Next, regression
analyses were performed for each of the individual study
populations and then for all the data sets combined.
The primary outcome for this study was late AMD.
Associations with early AMD were also assessed in SEE
and BES. Indicator variables were created for the various
cataract exposures with neither cataract surgery nor severe
cataract at the time of the clinical examination as the
reference level, presence of a severe cataract as the next
level, and previous cataract surgery or time since cataract
surgery as the final level. Regression analyses were done at
the eye level because cataract surgery and AMD status may
differ by eye. The standard errors were adjusted using
generalized estimating equations to account for the correlation between eyes.18
Age-adjusted analyses were done using logistic regression to determine the association of cataract surgery and
AMD for all eyes. Next, multiple logistic regression was
used to determine the association of cataract surgery and
AMD adjusting for the following other covariates besides
age: race (SEE and BES only), sex, smoking (never, past,
current), and study (combined analyses only). In the combined analyses, race was coded as Caucasian vs non-Caucasian to avoid colinearity between Hispanic race and study
population, as all the PVER participants were Hispanic.
A fixed effects model that assumed no heterogeneity in
the true odds ratios across studies was assumed. With only
three studies, it was not possible to reliably estimate the
across-study component of variance for the log odds ratios
as would be done in a random-effects model.19 However,
VOL. 135, NO. 6

CATARACT SURGERY

AND

the combined odds ratios under different assumptions about

the among-study variance were estimated. The moment
estimate of the among-study variance was 0.0, because the
variance among the three estimated log odds ratios was
smaller than the average statistical variance across the studies.
Hence, the best indication available from the data was that
the fixed effects model was appropriate.
Interactions were assessed by stratification and by the
addition of interaction terms into the regression models.
All analyses were done using the SAS statistical software
(SAS Institute, Cary, North Carolina, USA).

RESULTS
THE GENERAL CHARACTERISTICS OF THE THREE STUDY POP-

ulations are presented in Table 1. The SEE population was

more likely to be Caucasian, to have had more formal
education, to be current smokers, and because they were older
(selected to be aged 65 to 84 years), they were more likely to
have had previous cataract surgery and late AMD compared
with the PVER and BES populations. The proportion of
African Americans in BES was high (42%) owing to the
sampling strategy. Baltimore Eye Survey participants had the
highest proportion of current smokers, at 39%. Late AMD
was rare in the three studies with 73 cases in SEE, 53 in
PVER, and 48 in BES.
The characteristics of those who had cataract surgery in
at least one eye compared with those who did not for the
three studies are shown in Table 2. Subjects who had
undergone cataract surgery were older and were more likely to
be Caucasian. In addition, those who did not have cataract
surgery in SEE were more likely to have never smoked.
The age-specific and race-specific late AMD prevalence
rates for the study populations are presented in Table 3.
The prevalence of late AMD tended to increase with age
in the Caucasian population. This was not obviously so in
the African-American population, although there were
very small numbers of cases in the older age groups.
African-Americans were less likely to have late AMD
compared with Caucasians and Hispanics.
The three datasets for cataract surgery, severe cataract, and
AMD consistently showed that previous cataract surgery was
associated with an increased prevalence of late AMD, although they differed in the magnitude of the point estimates
and none of them were individually statistically significant
(Table 4). An analysis of the three data sets combined
together revealed an odds ratio (OR) for previous cataract
surgery of 1.7, which was statistically significant (95% confidence interval [CI], 1.1 2.6; Table 4). In addition, having a
severe cataract in the eye at the time of the clinic examination was also associated with a higher odds of AMD in all
three studies, although these estimates were not statistically
significant (combined OR 1.4; 95% CI, 0.8 2.4; Table 4).
In the two datasets (SEE and BES) that had information
on time since cataract surgery (Table 5), the prevalence of

AGE-RELATED MACULAR DEGENERATION

851

TABLE 1. General Characteristics of the Study Populations

Characteristic

Sample size
Age, mean
Sex, %
Male
Female
Race, %
Caucasian
African American
Hispanic
Education, %
12 years
Smoking, %
Never
Past
Current
Severe cataract,* at least 1 eye, %
Previous cataract surgery, at least
1 eye, %

Late AMD in at least 1 eye, %

4049 years
5059 years
6064 years
6569 years
70 years
Overall

Salisbury Eye
Evaluation

Proyecto VER

Baltimore Eye
Survey

2,520
73.5

4,774
56.9

4,396
58.9

42.1
57.9

38.8
61.2

40.4
59.6

73.6
26.4
0.0

0.0
0.0
100.0

57.3
41.9
0.7

51.5

65.1

69.6

38.8
46.2
15.0
9.0
18.8

52.0
27.6
20.4
8.0
8.2

34.6
26.5
38.9
1.0
3.1

NA
NA
NA
1.8
3.4
2.9

0.2
0.5
0.2
1.5
4.1
1.1

0.1
0.6
0.7
0.9
4.1
1.1

*Severe cataract defined as grade 4 or higher on photographs in SEE and PVER, and by the clinical
determination of cataract as a cause of visual impairment or blindness in BES.

Late age-related macular degeneration (AMD) defined as choroidal neovascularization, geographic

atrophy, or as a cause of blindness as determined by photographs in the Salisbury Eye Evaluation
(SEE) and Baltimore Eye Survey (BES), and by clinical examination in Proyecto VER (PVER).

was also attenuated and no longer statistically significant (OR 1.5; 95% CI, 0.8 2.6). Therefore, it does
not appear that the prevalence of late AMD increases
with time since cataract surgery.
A higher prevalence of early AMD 1 (the definition that
includes drusen of smaller size) was found in those who had
cataract surgery 5 or more years ago in SEE (OR 1.7;
95% CI, 1.12.4; Table 6). The prevalence of early AMD
1 was no different in those who were missing dates of
cataract surgery compared with those who had information
on date of cataract surgery (46% vs 44%). Thus, it is
unlikely that the missing data would have changed these
results. In the BES study and in the combined analysis,
though, previous cataract surgery was not associated with
early AMD 1. Severe cataract was also not associated with
early AMD 1 in SEE or in BES.
Neither cataract surgery nor severe cataract was associated with the odds of early AMD 2 in either SEE or BES
(Table 7), as none of the estimates were statistically
significant and no consistent trends were apparent.

late AMD increased as time since surgery increased.

Results from the combined analysis showed that those who
reported surgery 5 or more years earlier had 2.1 times the
odds of late AMD (95% CI, 1.0 4.6; Table 5). Those who
reported surgery less than 5 years earlier had modestly
elevated odds of late AMD, although it was not statistically significant (OR 1.4; 95% CI, 0.72.6; Table 5).
Sensitivity analyses were done to see how the eyes in SEE
that were missing dates of cataract surgery might bias the
results. Those eyes that were missing date of cataract
surgery were less likely to have AMD (0.8% vs 2.2%). A
minimum odds ratio was estimated by assuming that
those eyes that had cataract surgery but were missing the
date of cataract surgery were in the group that had
surgery 5 or more years earlier. The odds ratio was
dramatically reduced and was no longer statistically
significant, suggesting that surgery 5 or more years
earlier was likely not associated with late AMD in the
SEE data set (OR 1.2; 95% CI, 0.52.8). The
combined odds ratio for surgery 5 or more years earlier
852

AMERICAN JOURNAL

OF

OPHTHALMOLOGY

JUNE 2003

TABLE 2. Characteristics of Subjects Who Had Cataract Surgery in at Least One Eye Compared With Those Who Did Not Have
Cataract Surgery

Characteristics

Age category, %
4049 years
5059 years
6064 years
6569 years
7079 years
80 years
Sex, %
Male
Female
Race, %
White
Black
Education, %
12 years
Smoking, %
Never
Past
Current

SEE
Cataract Surgery
n 471

SEE
No Cataract Surgery
n 2,037

PVER
Cataract Surgery
n 390

PVER
No Cataract Surgery
n 4,362

BES
Cataract Surgery
n 138

BES
No Cataract Surgery
n 4,256

NA
NA
NA
15
59
26

NA
NA
NA
35
54
11*

4
8
7
13
41
27

36
30
11
9
11
2*

4
7
11
19
45
14

25
27
15
14
15
3*

36
64

57
43

37
63

39
61

39
61

40
60

85
15

71

29

NA
NA

NA
NA

76
24

57

43

48

52

83

64

76

69

35
50
15

40
45
15

54
36
10

52
27
21

46
28
26

34
26
40

BES Baltimore Eye Survey; NA not applicable; PVER Proyecto VER; SEE Salisbury Eye Evaluation.
*P .05 chi-square test.

P .05 after age adjustment using multiple logistic regression.

TABLE 3. Age- and Race-specific Prevalence Rates for Late AMD* in at Least One Eye for
Study Populations

Age Strata

SEE Whites
N (%)

BES Whites
N (%)

PVER Hispanics
N (%)

SEE Blacks
N (%)

BES Blacks
N (%)

4049 years
5059 years
6064 years
6569 years
70 years
Overall

NA
NA
NA
9/521 (1.7)
50/1,254 (4.0)
59/1,775 (3.3%)

0/478 (0.0)
4/539 (0.7)
2/373 (0.5)
4/377 (1.1)
29/508 (5.7)
39/2,275 (1.7%)

3/1,587 (0.2)
7/1,357 (0.5)
1/518 (0.2)
7/463 (1.5)
34/827 (4.1)
52/4,752 (1.1%)

NA
NA
NA
4/214 (1.9)
5/400 (1.4)
9/614 (1.5%)

1/536 (0.2)
2/544 (0.4)
2/244 (0.8)
1/204 (0.5)
1/218 (0.5)
7/1,746 (0.4%)

AMD age-related macular degeneration; BES Baltimore Eye Survey; N number; NA not
applicable; PVER Proyecto VER; SEE Salisbury Eye Evaluation.
*Late age-related macular degeneration (AMD) defined as choroidal neovascularization, geographic
atrophy, or as a cause of blindness as determined by photographs in SEE and BES and by clinical
examination in PVER.

DISCUSSION
THESE ANALYSES BASED ON DATA FROM THREE INDEPEN-

dent, population-based studies provide some evidence that

previous cataract surgery may be associated with a higher
prevalence of late AMD. In all studies, the odds ratios for
previous cataract surgery and late AMD were greater than
1, although they varied in magnitude from 1.3 to 2.6. The
VOL. 135, NO. 6

CATARACT SURGERY

AND

reasons for this variation are not clear, but they may be due
to differences in the population that could not be controlled for. They could also be due to the small number of
cases of late AMD, which can lead to unstable measures of
association. In addition, the assessment of AMD differed in
PVER, compared with BES and SEE. When data from the
three populations were pooled, the combined odds ratio
was 1.7, which was statistically significant.

AGE-RELATED MACULAR DEGENERATION

853

TABLE 4. Analysis of Late AMD* Associated With Previous Cataract Surgery or Presence of Severe Cataract at Time of Clinical

Examination for SEE, BES, PVER

SEE n 4,627 eyes

BES n 7,364 eyes

PVER n 9,477 eyes

Late AMD

Late AMD

Late AMD

Combined n 21,460
Late AMD

Variable

OR

95% CI

OR

95% CI

OR

95% CI

OR

95% CI

Neither cataract nor surgery

Severe cataract present

Previous cataract surgery

1.0
1.3
1.3

Reference
0.62.8
0.72.4

1.0
2.1
2.6

Reference
0.317.8
0.97.5

1.0
1.3
1.9

Reference
0.52.9
0.94.1

1.0
1.4
1.7

Reference
0.82.4
1.12.6

AMD age-related macular degeneration; BES Baltimore Eye Survey; CI confidence interval; OR odds ratio; PSC posterior
subcapsular; PVER Proyecto VER; SEE Salisbury Eye Evaluation.
*Late AMD defined as choroidal neovascularization, geographic atrophy, or as a cause of blindness as determined by photographs in SEE
and by clinical examination in PVER.

Individual studies adjusted for age, sex, smoking, race (SEE and BES only). Combined analyses adjusted for age, sex, smoking, race
(Caucasian, non-Caucasian), and study.

Severe cataract: nuclear 4, cortical 8/16, or any PSC opacity as determined from photographs for SEE and PVER and by visual
impairment or blindness attributed to cataract as determined by clinical examination in BES.

TABLE 5. Analysis of Late AMD* Associated With Time Since Cataract Surgery or Presence of Severe Cataract at Clinical

Examination for SEE, BES

SEE n 4,504 eyes

BES n 7,363 eyes

Late AMD

SEE and BES n 11,867

Late AMD

Late AMD

Variable

OR

95% CI

OR

95% CI

OR

95% CI

Neither severe cataract nor surgery

Presence of severe cataract

Surgery 5 years ago
Surgery 5 years ago

1.0
1.3
1.4
1.9

Reference
0.62.8
0.72.9
0.74.9

1.0
2.1
1.8
3.5

Reference
0.317.8
0.47.5
0.815.0

1.0
1.3
1.4
2.1

Reference
0.62.6
0.72.6
1.04.6

AMD age-related macular degeneration; BES Baltimore Eye Survey; CI confidence interval; OR odds ratio; SEE Salisbury Eye
Evaluation.
*Late AMD defined as choroidal neovascularization, geographic atrophy, or as a cause of blindness as determined by photographs in SEE
and BES.

Adjusted for age, sex, smoking, race, and study (SEE and BES combined). Generalized estimating equations were used to estimate the
standard errors to account for the correlation between eyes.

Missing data for 123 eyes on time of cataract surgery. Sensitivity analyses showed that the association disappears when all eyes missing
date of cataract surgery are placed in surgery 5 years ago (SEE results: OR 1.2, 95% CI, 0.52.8; (combined results: OR 1.5, 95% CI,
0.8 2.6).

Severe cataract defined as grade 4 or higher on photographs in SEE and PVER and by the clinical determination of cataract as a cause
of visual impairment or blindness in BES.

In addition, at first glance it appeared that increasing

time since cataract surgery was associated with a higher
prevalence of late AMD. In the BES study, those who had
cataract surgery 5 or more years earlier had an increased
prevalence of late AMD, although this was not statistically
significant. The SEE study showed a similar trend, although sensitivity analyses showed that this trend was due
to a bias caused by the missing dates of surgery.
Our overall findings are in agreement with the prospective Beaver Dam results, which found previous cataract
surgery to be associated with a higher 10-year incidence of
late AMD (RR 3.8; 95% CI, 1.9 7.7).8 They also found
854

AMERICAN JOURNAL

previous cataract surgery to be associated with a greater

progression of early AMD (OR 2.0; 95% CI, 1.3 3.0),
but not with the incidence of early AMD.8
There are several possible reasons that might explain,
either individually or in concert, the association between
cataract surgery and late AMD that we and others have
observed. One possibility is that cataract and AMD simply
share one or more common risk factors. In our analyses, the
finding that severe cataract was also associated with a
modestly increased prevalence of late AMD indicates that
it may not be the cataract surgery but perhaps factors
leading to cataract that potentially have a causal relationOF

OPHTHALMOLOGY

JUNE 2003

TABLE 6. Analysis of Early AMD 1 Associated With Time Since Cataract Surgery or
Presence of Severe Cataract at Clinical Examination for SEE, BES*
SEE

BES

Early AMD 1
n 3,912

Early AMD 1
n 8,656

SEE and BES

Early AMD 1
n 12,568

Variable

OR

95% CI

OR

95% CI

OR

95% CI

Neither severe cataract nor surgery

Presence of severe cataract

Surgery 5 years ago
Surgery 5 years ago

1.0
1.2
1.2
1.7

Reference
0.91.7
0.91.6
1.12.4

1.0
0.5
0.7
0.7

Reference
0.21.4
0.41.4
0.41.4

1.0
1.1
1.2
1.3

Reference
0.81.5
0.91.5
1.01.7

AMD age-related macular degeneration; BES Baltimore Eye Survey; CI confidence interval;
OR odds ratio; SEE Salisbury Eye Evaluation.
*Adjusted for age, sex, smoking, race, and study (SEE and BES combined). Generalized estimating
equations were used to estimate the standard errors to account for the correlation between eyes.

Early AMD 1: drusen 64 mm, hyperpigmention, or nongeographic atrophy.

Severe cataract defined as grade 4 or higher on photographs in SEE and PVER and by the clinical
determination of cataract as a cause of visual impairment or blindness in BES.

TABLE 7. Analysis of Early AMD 2 Associated With Time Since Cataract Surgery or
Presence of Severe Cataract at Clinical Examination for SEE, BES*
SEE

BES

Early AMD 2
n 4,356

Early AMD 2
n 8,639

SEE and BES

Early AMD 2
n 12,995

Variable

OR

95% CI

OR

95% CI

OR

95% CI

Neither severe cataract nor surgery

Presence of severe cataract
Surgery 5 years ago
Surgery 5 years ago

1.0
0.6
0.9
1.4

Reference
0.31.1
0.61.5
0.82.5

1.0

1.0
0.9

Reference

1.0

1.0
1.2

Reference

0.52.0
0.41.9

0.71.4
0.81.8

AMD age-related macular degeneration; BES Baltimore Eye Survey; CI confidence interval;
OR odds ratio; SEE Salisbury Eye Evaluation.
*Adjusted for age, sex, smoking, race, and study (SEE and BES combined). Generalized estimating
equations were used to estimate the standard errors to account for the correlation between eyes.

Early AMD 2: drusen 125 m, hyperpigmention, or nongeographic atrophy.

Severe cataract defined as grade 4 or higher on photographs in SEE and PVER and by the clinical
determination of cataract as a cause of visual impairment or blindness in BES.

ship with late AMD. The association between cataract

surgery and late AMD persists after controlling for the
known common risk factors of age and smoking. There
may be other risk factors (environmental or genetic) that
have not been identified, however, or cataract and late
AMD may both be markers for accelerated aging.
A second possibility that could explain the association
between cataract surgery and AMD is that we are less
likely to detect late AMD in persons who may still have a
lens opacity in the eye. That should not be an issue in our
study, however, because persons with severe cataract are
not included in the reference population. Therefore, there
should not have been a problem in detecting retinal
pathology in the reference population. Likewise, an ophVOL. 135, NO. 6

CATARACT SURGERY

AND

thalmologist may look more carefully for signs of AMD in

an eye that has had cataract surgery in the past. Nevertheless, retinal photographs were used for two of the
studies and the graders had no knowledge of whether the
person had cataract surgery or not.
A third possibility is that cataract surgery in some way
physically predisposes an eye to develop AMD, perhaps
through inflammatory mechanisms.12 The epidemiologic
data cannot directly address this possibility. Our data
indicate that the prevalence of late AMD does not
increase with time from cataract surgery, suggesting that
such an effect (if it exists) likely occurs soon after surgery.
A fourth explanation for the observed association between cataract surgery and late AMD is the light hypoth-

AGE-RELATED MACULAR DEGENERATION

855

2. Hyman L. Epidemiology of eye disease in the elderly. Eye

1987;1:330 341.
3. Hyman L. Epidemiology of advanced AMD: In: Hampton
RG, Nelson PT, editors. Age-related macular degeneration:
principles and practice. New York: Raven Press, 1992:135.
4. 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:
819 825.
5. Steinberg EP, Tielsch JM, Schein OD, et al. National study
of cataract surgery outcomes. Variation in 4-month postoperative outcomes as reflected in multiple outcome measures.
Ophthalmology 1994;101:11311140.
6. Steinberg EP, Javitt JC, Sharkey PD, et al. The content and
cost of cataract surgery. Arch Ophthalmol 1993;111:1041
1049.
7. Klein R, Klein BEK, Wang Q, Moss SE. Is age-related
maculopathy associated with cataracts. Arch Ophthalmol
1994;112:191196.
8. Klein R, Klein BEK, Wong TY, Tomany SC, Cruickshanks
KJ. The association of cataract and cataract surgery with the
long-term incidence of age-related maculopathy. Arch Ophthalmol 2002;120:15511558.
9. Wang JJ, Mitchell P, Cumming RG, Lim R. Cataract and
age-related maculopathy: the Blue Mountains Eye Study.
Ophthalmic Epidemiol 1999;6:317326.
10. McCarty CA, Mukesh BN, Fu CL, Mitchell P, Wang JJ,
Taylor HR. Risk factors for age-related maculopathy: the
Visual Impairment Project. Arch Ophthalmol 2001;119:
14551462.
11. Liu IY, White L, LaCroix AZ. The association of age-related
macular degeneration and lens opacities in the aged. Am J
Public Health 1989;79:765769.
12. van der Schaft TL, Mooy CM, de Bruijn WC, Mulder PG,
Pameyer JH, de Jong PT. Increased prevalence of disciform
macular degeneration after cataract extraction with implantation of an intraocular lens. Br J Ophthalmol 1994;78:441
445.
13. West SK, Mun oz B, Rubin GS, et al. Function and visual
impairment in a population-based study of older adults. The
SEE project. Salisbury Eye Evaluation. Invest Ophthalmol
Vis Sci 1997;38:7282.
14. Taylor HR, West SK. The clinical grading of lens opacities.
Aust N Z J Ophthalmol 1989;17:8186.
15. West SK, Rosenthal F, Newland HS, Taylor HR. Use of
photographic techniques to grade nuclear cataracts. Invest
Ophthalmol Vis Sci 1988;29:7377.
16. West SK, Klein R, Rodriguez J, et al. Diabetes and diabetic
retinopathy in a MexicanAmerican population: Proyecto
VER. Diabetes Care 2001;24:1204 1209.
17. Tielsch JM, Sommer A, Witt K, Katz J, Royall RM. Blindness and visual impairment in an American urban population. The Baltimore Eye Survey. Arch Ophthalmol 1990;
108:286 290.
18. Zeger SL, Liang K-Y. Longitudinal data analysis for discrete
and continuous outcomes. Biometrics 1986;42:121130.
19. Diggle P, Heagerty P, Liang K-Y, Zeger SL. Analysis of
longitudinal data. 2nd ed. Oxford, Oxford University Press,
2002.
20. Taylor HR, Mun oz B, West S, Bressler NM, Bressler SB,
Rosenthal FS. Visible light and risk of age-related macular
degeneration. Trans Am Ophthalmol Soc 1990;88:163178.

esis.7,11 In this hypothesis, removal of the cataract newly

exposes the retina to certain wavelengths of light (or at
greater intensity), damaging the retina and increasing the
risk of AMD.20 As most of the intraocular lenses currently
in use have ultraviolet-B blockers, the critical wavelengths
are likely to be in the blue light region. Further investigations of this hypothesis would best be assessed by a
randomized controlled clinical trial in which critical or
putative wavelengths are blocked.
Some eyes (1,803/23,380) could not be evaluated for
AMD status, which could cause concern for selection bias.
When we examined the characteristics, however, eyes
without information on AMD status were from older
subjects (66 vs 61 years old) and the eyes were more likely
to have had cataract surgery (8.1% vs 6.6%). Given the
older age, it is likely that the AMD rates of those eyes missing
AMD status information would have been higher than the
rates seen in this study. As we suspect the AMD rates to be
higher and we know the rates of cataract surgery were higher
in eyes missing AMD status, including these eyes in the
analyses would have likely only strengthened the results.
These analyses are cross-sectional in nature. Therefore,
we are limited to identifying an association between
cataract surgery and AMD. Because we are unable to
determine the temporal relationship of cataract surgery
and signs of AMD, we cannot declare that cataract surgery
increases the risk of AMD. Another limit to our analyses is
the small number of cases of AMD, particularly in the
younger populations. For this reason, it is important to
examine the overall results and not just focus on the results
that were statistically significant. Finally, the PVER study
assessed AMD using a clinical examination whereas the
other two studies assessed AMD with photographs. We
attempted to control for this by adjusting for study in the
combined analyses.
Using three large population-based studies, we found
modest evidence that cataract surgery may be associated
with a greater prevalence of late AMD. We also saw that
having a severe cataract in the eye was associated with a
slightly higher prevalence of late AMD. Thus, it is difficult
to determine from this type of study design whether the
higher rates of late AMD are due to the cataract surgery, to
the cataract itself, or to a shared risk factor. It would be
helpful if additional studies could clarify the nature of this
association.

REFERENCES
1. 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;103:17211726.

856

AMERICAN JOURNAL

OF

OPHTHALMOLOGY

JUNE 2003