Christen WG, Glynn RJ, Ajani UA, Schaumberg DA, Buring JE, Hennekens CH, Manson JE. Smoking Cessation and Risk of Age-Related Cataract in Men. JAMA. 2000;284(6):713-716. doi:10.1001/jama.284.6.713
Author Affiliations: Division of Preventive Medicine (Drs Christen, Glynn, Ajani, Schaumberg, Buring, and Manson) and Channing Laboratory (Dr Manson), Department of Medicine, and Department of Ambulatory Care and Prevention (Dr Buring), Harvard Medical School and Brigham and Women's Hospital, and Departments of Biostatistics (Dr Glynn) and Epidemiology (Drs Buring and Manson), Harvard School of Public Health, Boston, Mass; and Departments of Medicine, Epidemiology, and Public Health, University of Miami School of Medicine, Miami, Fla (Dr Hennekens).
Context Although cigarette smoking has been shown to be a risk factor for age-related
cataract, data are inconclusive on the risk of cataract in individuals who
Objective To examine the association between smoking cessation and incidence of
Design Prospective cohort study conducted from 1982 through 1997, with an average
follow-up of 13.6 years.
Setting and Participants A total of 20,907 US male physicians participating in the Physicians'
Health Study I who did not have a diagnosis of age-related cataract at baseline
and had reported their level of smoking at baseline.
Main Outcome Measures Incident age-related cataract defined as self-report confirmed by medical
record review, diagnosed after study randomization and responsible for vision
loss to 20/30 or worse, and surgical extraction of incident age-related cataract,
in relation to smoking status and years since quitting smoking.
Results At baseline, 11% were current smokers, 39% were past smokers, and 50%
were never smokers. Average reported cumulative dose of smoking at baseline
was approximately 2-fold greater in current than in past smokers (35.8 vs
20.5 pack-years). Two thousand seventy-four incident cases of age-related
cataract and 1193 cataract extractions were confirmed during follow-up. Compared
with current smokers, multivariate relative risks (RRs) of cataract in past
smokers who quit smoking fewer than 10 years, 10 to fewer than 20 years, and
20 or more years before the study were 0.79 (95% confidence interval [CI],
0.64-0.98), 0.73 (95% CI, 0.61-0.88), and 0.74 (95% CI, 0.63-0.87), respectively,
after adjustment for other risk factors for cataract and age at smoking inception.
The RR for never smokers was 0.64 (95% CI, 0.54-0.76). The reduced risk in
past smokers was principally due to a lower total cumulative dose (RR of cataract
for increase of 10 pack-years of smoking, 1.07; 95% CI, 1.04-1.10). A benefit
of stopping smoking independent of cumulative dose was suggested in some analyses.
Results for cataract extraction were similar.
Conclusion These prospective data indicate that while some smoking-related damage
to the lens may be reversible, smoking cessation reduces the risk of cataract
primarily by limiting total dose-related damage to the lens.
Cigarette smoking has been shown to be an important independent risk
factor for development of age-related cataract.1- 11
However, risk of cataract in individuals who quit smoking is uncertain. Some
studies have found that risk of cataract remains elevated for many years following
smoking cessation.4,9 Others have
reported that the risk approaches the level of never smokers several years
after quitting smoking, suggesting that damage to the lens may be reversible.5,6,10
In a report from the Physicians' Health Study I, based on the first
5 years of follow-up, we showed that current smokers of 20 or more cigarettes
per day, compared with never smokers, had a 2-fold increased risk of cataract,
while past smokers had a 15% increased risk of cataract that was not statistically
significant.8 In this article, we extend these
findings by including cataracts diagnosed during more than 13 years of follow-up
and by examining the relationship of time since quitting smoking with risk
of cataract following smoking cessation.
The Physicians' Health Study I was a randomized, double-blind, placebo-controlled
trial of low-dosage aspirin and β-carotene in the prevention of cardiovascular
disease and cancer among 22,071 US male physicians aged 40 to 84 years in
1982. At baseline, participants completed mailed questionnaires that included
information on history of and risk factors for cataract and whether they had
ever smoked cigarettes regularly (grouped as never, past, or current). Current
smokers were asked how many cigarettes per day, on average, they smoked. On
the 60-month follow-up questionnaire, subjects were asked to give a detailed
smoking history, including their age at starting and quitting smoking. The
60-month questionnaire also requested information on amount of cigarettes
smoked for both current and past smokers. Subjects were asked, "When you smoke
(or smoked), on average how many cigarettes per day do (did) you smoke (<1
pack per day, 1 pack per day, 1-2 packs per day, or ≥2 packs per day)?"
Information on new occurrence of cataract and cataract extraction was ascertained
on yearly follow-up questionnaires.
This investigation includes the 20,907 participants who had no diagnosis
of cataract and provided information about cigarette smoking at baseline.
Following a report of a cataract diagnosis or extraction and receipt
of written consent to obtain medical records pertaining to cataract, treating
ophthalmologists or optometrists were contacted by mail to obtain information
about presence of lens opacities, date of diagnosis, visual acuity loss, cataract
extraction, other ocular abnormalities that could explain visual acuity loss,
cataract type, and etiology.
End points were incident cataract and extraction of incident cataract.
Cataract was defined as a self-report confirmed by medical record review initially
diagnosed after randomization, age-related in origin, with best-corrected
visual acuity of 20/30 or worse attributable to cataract. This article includes
all available data through December 1997.
Cox proportional hazards models were used to assess the effect of smoking
cessation on risk of cataract while simultaneously controlling for other cataract
risk factors. Models were fit using current smokers as the reference category,
as recommended in the 1990 surgeon general's report on smoking cessation.12 In initial analyses, subjects were classified as
never, past, or current smokers, based on data reported at baseline. We then
examined the risk of cataract in past smokers who quit less than 10 years,
10 to less than 20 years, or 20 or more years before study entry in models
that controlled for other cataract risk factors and for age at starting smoking
or number of cigarettes consumed per day. Finally, we examined the independent
contributions of total cumulative dose and smoking status (past vs current)
to risk of cataract in ever smokers. The significance of variables was tested
using the likelihood ratio test. To calculate total cumulative dose, subjects
were classified by pack-years of smoking. We used baseline data on amount
smoked for current smokers and 60-month follow-up questionnaire data on amount
smoked for past smokers (since information on amount smoked was collected
only for current smokers at baseline) to calculate pack-years of smoking at
baseline. We defined pack-years as the number of years of smoking times the
number of packs of cigarettes smoked per day.
Cigarette smoking is associated with an increased risk of age-related
macular degeneration (AMD) in this population,13
and subjects with cataract may have been identified because of presence of
AMD. Therefore, we also conducted analyses in which we included diagnosis
of AMD as a time-varying covariate. Relative risk (RR) estimates derived from
these models, however, were not materially different from estimates derived
from models that were unadjusted for diagnosis of AMD (data not shown).
At baseline, 11% of the study participants were current smokers, 39%
were past smokers, and 50% were never smokers. Compared with current smokers,
past smokers were older and, after adjusting for age, tended to report less
alcohol use, diabetes, parental history of myocardial infarction, and multivitamin
use, but more physical activity (Table 1). Mean age at starting smoking was similar in past and current
smokers, but total pack-years of smoking at baseline were almost 2-fold greater
in current smokers than in past smokers (35.8 vs 20.5 pack-years) (Table 1).
During an average of 13.6 years of follow-up, there were 2074 age-related
cataract diagnoses and 1193 cataract extractions confirmed by medical record
review. Risk of cataract in past smokers was intermediate between current
and never smokers. Compared with men who continued to smoke, past smokers
had a statistically significant 23% reduced risk of cataract diagnosis (RR,
0.77; 95% confidence interval [CI], 0.66-0.88) and 28% reduced risk of cataract
extraction (RR, 0.72; 95% CI, 0.60-0.86), after adjustment for other risk
factors for cataract. The RR in never smokers compared with men who continued
to smoke was 0.64 (95% CI, 0.54-0.76) for cataract diagnosis and 0.65 (95%
CI, 0.53-0.79) for cataract extraction.
Compared with men who continued to smoke, men who had quit less than
10 years before study entry had an approximately 20% reduced risk of cataract
diagnosis after adjustment for other cataract risk factors and average number
of cigarettes smoked per day or age at starting smoking (Table 2). There was little additional reduction in risk in men who
had quit 10 or more years before study entry. For cataract extraction, there
was an approximately 25% reduced risk in men who had quit less than 10 years
before study entry vs those who continued to smoke. Similarly, there appeared
to be little additional reduction in risk in men who quit smoking 10 or more
years before study entry. For both end points, risk of cataract in long-term
quitters appeared to remain slightly (but not significantly) elevated compared
with risk in never smokers.
When we examined the independent contributions of total cumulative dose
and smoking status to risk of cataract among ever smokers, the best-fitting
multivariate model included only a term for total cumulative dose (RR, 1.07;
95% CI, 1.04-1.10). Thus, there was a 7% increased risk of cataract associated
with a 10-pack-year increase in smoking exposure. Addition of a 2-level categorical
term for smoking status (past vs current, RR, 0.85; 95% CI, 0.73-1.00) was
suggestive of an independent benefit associated with being a past smoker,
but did not significantly improve the fit of the model based on the likelihood
ratio test (P = .08). For cataract extraction, the
best-fitting model did include a term for smoking status (past vs current,
RR, 0.80; 95% CI, 0.65-0.97) in addition to a continuous term for total pack-years
of smoking (for a 10-pack-year increase, RR, 1.05; 95% CI, 1.01-1.09), suggesting
a benefit for past smokers that was independent of total cumulative dose.
There were no significant interactions between smoking status and cumulative
dose for either end point.
In this cohort of US male physicians, the risk of cataract in past smokers
appeared to be intermediate between the risks in continuing smokers and never
smokers. Compared with those who continued to smoke, a lower risk of cataract
in past smokers was apparent in men who had quit smoking less than 10 years
before study entry, with little additional reduction in risk associated with
longer time since quitting smoking. The lower risk in past smokers was due
primarily to their lower total cumulative dose of smoking, although there
was evidence of a benefit of quitting smoking that was independent of cumulative
dose, suggesting that some smoking-related damage in the lens may be reversible
on smoking cessation.
Several previous studies have examined the risk of cataract in former
smokers. A population-based, cross-sectional survey of 838 Maryland watermen
indicated that the risk of pure nuclear opacities decreased during the first
10 years following smoking cessation and continued to decrease with longer
time since quitting.5 A more recent analysis
of prospective data from that cohort showed that past smokers and never smokers
had similar progression rates of nuclear sclerosis during 5 years of follow-up.10 In the Lens Opacities Case-Control Study, there was
an increased risk of nuclear sclerosis among current smokers (RR, 1.7; 95%
CI, 1.0-2.8), but none was reported for past smokers.6
The results of 2 other studies suggest that the risk of cataract, especially
among heavy smokers, decreases little, if at all, following smoking cessation.
In the City Eye Study, a cross-sectional analysis of baseline data showed
that past heavy smokers (≥25 cigarettes per day) had a risk of nuclear
sclerosis (RR, 2.6; 95% CI, 1.4-5.0) comparable in magnitude with the risk
in current heavy smokers (RR, 2.9; 95% CI, 1.4-5.9).4
There was no increased risk of nuclear sclerosis in past moderate (15-24 cigarettes
per day) or past light (1-14 cigarettes per day) smokers in that study. In
the Nurses Health Study, the overall RR for past smokers vs never smokers
was 1.0 (95% CI, 0.8-1.2). However, the risk of cataract in past smokers who
had smoked at least 35 cigarettes per day (RR, 1.8; 95% CI, 1.1-2.9) was similar
to the risk in current heavy smokers (RR, 1.6; 95% CI, 1.0-2.7), and these
risks remained comparable after 10 or more years following smoking cessation.9
Our findings in the Physicians' Health Study I are in broad agreement
with previous studies that indicate a reduced risk of cataract in past smokers
compared with current smokers, due primarily to lower total cumulative dose
in past smokers. We found no evidence to indicate that even the most heavily
exposed men in our population did not benefit from smoking cessation. However,
our data show that compared with men who have never smoked, past smokers appear
to have a slightly elevated risk of cataract that may persist for years following
Several possible limitations of the study should be considered. The
prospective study design reduces the possibility of bias in reports of cigarette
smoking or other potential risk factors according to disease outcome. Random
misclassification of cigarette exposure is a possibility, but would tend to
underestimate any true association between cigarette smoking and cataract.
Random misclassification of cataract would also underestimate any true effect
of smoking, but was minimized by the use of medical records to confirm the
self-reports. Nonrandom misclassification of cataract is unlikely since medical
records were reviewed without knowledge of participants' exposure status.
Morbidity follow-up was more than 99% complete (through December 1997) and
medical records were obtained for 92% to 94% of current, past, and never smokers
who reported cataract. Thus, bias due to incomplete follow-up is not likely
to distort these results.
Mechanisms linking cigarette smoking and cataract have been described,
including a direct effect on the lens,14,15
as well as indirect effects on antioxidant levels16- 20
and levels of endogenous proteolytic enzymes, thought to be important for
removal of damaged protein from the lens.21
Stopping smoking may alleviate further direct damage to lens proteins and,
perhaps, allow reversal of some of the early deleterious effects of smoking.
Cataract is a leading cause of visual impairment in the United States
and represents a major drain on health care resources.22
Approximately 1.35 million cataract operations are performed yearly in the
United States at an estimated cost of $3.5 billion.22
Given these considerations, recognition that smoking is an important, avoidable
cause of age-related cataract can be expected to have major public health
implications. The data presented extend previous findings by demonstrating
that smoking cessation reduces the risk of cataract primarily by limiting
total smoking-related damage to the lens. The data also indicate that some
damage in the lens may not be reversed with smoking cessation, underscoring
the importance of early cessation of smoking and, preferably, the avoidance
of smoking altogether.