Wu R, Wang JJ, Mitchell P, Lamoureux EL, Zheng Y, Rochtchina E, Tan AG, Wong TY. Smoking, Socioeconomic Factors, and Age-Related CataractThe Singapore Malay Eye Study. Arch Ophthalmol. 2010;128(8):1029-1035. doi:10.1001/archophthalmol.2010.147
To describe the relationship of smoking, sex, and socioeconomic factors with age-related cataract in Malay adults in Singapore.
In a population-based study, 3280 Malay individuals aged 40 to 80 years participated (78.7% response rate). All had interviews, systemic examination, and laboratory investigations. Lens opacity was graded from slitlamp and retroillumination photographs using the Wisconsin Cataract Grading System. Smoking-cataract associations were compared with the Blue Mountains Eye Study in Australia.
Of 2927 participants (89.2%) with gradable lens photographs, 1338 (45.7%) had cataract. After adjusting for age, sex, body mass index, hypertension, and diabetes, current smokers had a higher prevalence of nuclear cataract (odds ratio [OR], 2.06; 95% confidence interval [CI], 1.46-2.98), cortical cataract (OR, 1.33; 95% CI, 1.02-1.74), posterior subcapsular cataract (OR, 1.39; 95% CI, 1.02-1.91), or any cataract (OR, 1.48; 95% CI, 1.10-1.99). These associations were not seen in the Blue Mountains Eye Study. Primary or lower education (OR, 1.67; 95% CI, 1.06-2.64) and low monthly income (OR, 1.43; 95% CI, 1.09-1.87) were both associated with nuclear cataract, while small-sized public housing was associated with posterior subcapsular cataract (OR, 1.70; 95% CI, 1.28-2.25). Among men, 43.5% currently smoked compared with only 3.2% of women. The population attributable risk of nuclear cataract due to smoking was estimated to be 17.6% in men.
Smoking and indicators of low socioeconomic status were associated with cataract in Malay persons, with 1 in 6 nuclear cataract cases in men attributable to smoking. Smoking-cataract associations were stronger in Malay than in white persons.
Cataract is a major cause of avoidable blindness and visual impairment worldwide, accounting for about half the global prevalence of blindness.1 A growing body of research has investigated risk factors associated with the early development and progression of age-related cataract.2- 22
Previous studies show that lifestyle factors such as smoking and alcohol consumption and indicators of socioeconomic status (SES) such as education, household income, and housing are associated with both higher prevalence and/or progression of cataract.4- 8,10- 13,16,18- 20 However, these data have mainly been derived from white populations in the United States,4,9,19 Europe,7,8,13,18 and Australia.6 While data are now emerging from India15,16,21 and China,11 few have been conducted in urban environments that are representative of the major population trends in Asia.23 Furthermore, there are no studies among persons of Malay ethnicity, the third largest racial/ethnic group in Asia and 5% of the world's population.24
In this population-based study of adult Malay individuals in Singapore, we describe the associations of smoking and SES risk factors with age-related nuclear, cortical, and posterior subcapsular (PSC) cataract. We further assessed the proportion of cataract cases in this Malay population that were attributable to smoking. Finally, we had the unique opportunity to compare the associations of smoking with cataract between our study sample and white persons from the Blue Mountains Eye Study (BMES) in Australia, given the same standardized methods used to assess cataract from lens photographs in the 2 studies.
The Singapore Malay Eye Study was a population-based, cross-sectional epidemiological study of 3280 Malay adults aged 40 to 80 years who lived in Singapore. The design, sampling plan, methods, and baseline characteristics have been described elsewhere.25- 27 In brief, an age-stratified random sampling of all Malay adults aged 40 to 80 years residing in 15 residential districts in the southwestern part of Singapore was performed. Of the 4168 eligible individuals, 3280 participated in the study (78.7% response rate). Demographic characteristics of Singapore Malay Eye Study participants are described elsewhere.28
All examinations were conducted from August 2004 to June 2006 after obtaining informed consent at the Singapore Eye Research Institute, a clinical research facility located centrally in Singapore. The study adhered to the Declaration of Helsinki, and ethics approval was obtained from the Singapore Eye Research Institute institutional review board.
At the study clinic, all participants had extensive and standardized ocular examination including visual acuity testing, slitlamp biomicroscopy, stereoscopic fundus examination, and ocular imaging (lens and retina). Clinical assessment of lens status and the presence of aphakia or pseudophakia were determined at the slitlamp. Photographs of the lens were taken after pharmacologic (tropicamide, 1%, and phenylephrine hydrochloride, 2.5%) dilation. Slitlamp photographs (Topcon SL-7e camera; Topcon Optical Co, Tokyo, Japan) were taken to grade the degree of nuclear sclerosis, and retroillumination photographs were taken with the Neitz CT-R camera (Neitz Instruments Co, Tokyo, Japan) to determine the presence and severity of cortical and posterior subcapsular cataract.
Cataracts were assessed from lens photographs using the Wisconsin Cataract Grading System.29 Slitlamp photographs were compared against a set of 4 standards, and the opacity of the nucleus was graded using a 5-step severity scale. Nuclear cataract was defined as opacity greater than standard 3. Cortical and PSC cataract were assessed by laying a grid over the anterior and posterior photographs and estimating the proportion of lens areas involved with each opacity. Cortical cataract was defined if 5% or more of the total lens area was involved and PSC cataract if any such opacity was present. Any cataract was defined to include any cataract of the 3 types. All nuclear and PSC cataract cases were adjudicated by 1 or both senior researchers (J.J.W. and P.M.).
A person was considered to have a specific cataract subtype if it was present in either eye; if a person had different cataract subtypes in each eye, he or she was classified as a case in the analyses of each subtype.
A structured, interviewer-administered questionnaire was used to collect socioeconomic, demographic, and medical information including age, sex, marital status, level of education obtained, occupation, income, and family history of diabetes and hypertension. We assessed SES variables based on education, income, and housing type. Educational level was recorded as the highest number of years of schooling completed and was categorized into 2 groups: (1) primary school level or lower (≤6 years) and (2) secondary school level or higher (≥7 years, including university education). Income level was defined as individual monthly income in Singapore dollars (S $) and was divided into 2 categories: (1) low, less than S $1000; and (2) middle and high, S $1000 or greater. Housing was categorized into 2 categories: (1) 1- to 2-room public housing and (2) 3- to 5-room public or private housing. Smoking history was used to determine whether participants were current smokers, past smokers, or nonsmokers. If they were past or current smokers, the number of packs smoked per week was recorded.30 Alcohol consumption was divided into drinker (irrespective of quantity) and nondrinker.
In the current study, we examined education, income, and housing as SES indicators. Occupation was not included, as it was not available for all individuals.
Several laboratory investigations were conducted to assess other metabolic risk factors. Nonfasting serum glucose and serum total cholesterol levels were measured from venous blood samples collected from participants. Diabetes mellitus was defined as a random glucose level of 200 mg/dL or higher (to convert to millimoles per liter, multiply by 0.0555), use of diabetic medication, or self-report of physician diagnosis. Hyperlipidemia was defined as a total cholesterol level of 239.4 mg/dL (to convert to millimoles per liter, multiply by 0.0259) or use of lipid-lowering drugs.
Height and weight were measured to determine body mass index (calculated as weight in kilograms divided by height in meters squared). Blood pressure was measured on 2 occasions (5 minutes apart). If the blood pressure levels differed by more than 10 mm Hg (systolic) and 5 mm Hg (diastolic), a third measurement was made. The blood pressure of the individual was then taken as the mean between the 2 closest readings. Hypertension was defined as systolic blood pressure of 140 mm Hg or higher, diastolic blood pressure of 90 mm Hg or higher, or current use of antihypertensive medications.
Descriptive analyses were performed for all variables, and differences between current smokers, past smokers, and nonsmokers were initially assessed using t test or χ2 test, as appropriate. The prevalence of specific types of cataract and any cataract were evaluated initially by contingency tables. Age-, sex-, and multivariable-adjusted odds ratios (ORs) and 95% confidence intervals (CIs) were estimated for associations of specific types of cataract or any cataract with smoking status and socioeconomic factors using logistic regression models. All reported P values were based on 2-sided tests and compared with a significance level of 5%. All analyses were performed using SPSS version 15 (SPSS Inc, Chicago, Illinois). The population-attributable risk of cataract from smoking was estimated using the Levin formula in which the prevalence risk ratio was calculated from the multivariate adjusted OR.31
A total of 3280 participants were examined; after excluding persons with cataract surgery and ungradable lens photographs, data from 2927 (89.2%) persons were available for analysis. Cataract was identified in at least 1 eye in 1338 participants (45.7%). Participant characteristics stratified by the presence or absence of cataract are shown in Table 1. When compared with those without cataract, participants with any cataract were older and had higher prevalence of hypertension, diabetes mellitus, hyperlipidemia, primary or lower education level, and low monthly income but lower prevalence of current smoking or alcohol intake.
The prevalence of specific types of cataract and any cataract by smoking and SES is listed in Table 2. In general, never or past smoking, primary or lower education, smaller housing type, lower individual monthly income, and no alcohol consumption were associated with higher prevalence of all 3 types of cataract.
Table 3 summarizes the relationship between potential risk factors (including smoking and SES) and cataract. Increasing age was strongly associated with all types of cataract. Female sex was associated with nuclear, cortical, and presence of any cataract. After adjusting for age, sex, body mass index, hypertension, and diabetes, the following associations were noted: (1) current smoking was associated with increased prevalence of each of the 3 cataract types or any cataract; (2) primary or lower educational levels and low individual monthly income were associated with higher prevalence of nuclear cataract; and (3) living in smaller housing units (1-2 rooms) was associated with higher prevalence of PSC cataract. No association was identified between alcohol consumption and any particular type of cataract or any cataract.
In addition, current smoking was associated with nuclear cataract in a dose-dependent manner (test for trend, P = .002) (Table 4). Specifically, persons who smoked at least 5 cigarette packs per week had higher odds (OR, 2.51) of nuclear cataract than those who smoked between 1 and 5 cigarette packs per week (OR, 1.62) compared with those who did not smoke. A similar trend with borderline nonsignificance was also evident for any cataract (test for trend, P = .09).
Current smokers were, on average, younger (mean [SD] age, 55.5 [10.4] years) than nonsmokers (mean [SD] age, 62.2.5 [11.2] years) and past smokers (mean [SD] age, 58.7 [10.9] years) in the study sample. Among current smokers, those with nuclear cataract had a mean (SD) age of 69.2 (6.1) years, while in nonsmokers or past smokers, those with nuclear cataract had a mean (SD) age of 70.8 (6.0) years. Similarly, the mean (SD) ages of persons with cortical and PSC cataract were 63.5 (8.1) and 65.2 (8.7) years, respectively, among current smokers and 66.2 (8.5) and 68.0 (8.4) years, respectively, among the combined group of nonsmokers and past smokers. These findings suggest that current smokers appear to develop lens opacity, particularly nuclear cataract, at a younger age.
The associations of current smoking with cataract prevalence by sex are presented in Table 5. In our study sample, 43.5% of men were current smokers compared with only 3.2% of women. In men, 17.6% (95% CI, 6.4%-21.7%) of nuclear cataract cases were attributable to current smoking behavior, while in women, the corresponding proportion of nuclear cataract cases was 4.1% (95% CI, 0.1%-7.7%). In men and women combined, the population-attributable risk from current smoking was 12.4% for nuclear (95% CI, 6.2%-19.3%), 4.1% for cortical (95% CI, 0.3%-8.1%), 6.0% for PSC (95% CI, 0.3%-12.3%), and 4.3% for any cataract (95% CI, 1.1%-7.6%).
When we stratified nonsmokers and past smokers into never and past smokers, no difference was observed in the association with cataract prevalence. Only current smoking was found to be associated with cataract prevalence (data not shown).
When we compared our findings with those from the BMES, the association of current smoking with nuclear cataract was marginally nonsignificant in the BMES sample, and no dose-dependent trend was evident (Table 5).
In this urban Malay population in Singapore, we have made 3 main conclusions. First, current smoking was significantly associated with an increased prevalence of all types of age-related cataract, independent of age, sex, body mass index, hypertension, and diabetes. The association of current smoking with nuclear cataract prevalence suggested a dose-dependent pattern, with increasing risk associated with greater pack-years of smoking. Second, current smokers with either nuclear, cortical, or PSC opacity were younger, on average, than nonsmokers or past smokers with the same types of lens opacity. Third, among men (most smokers), 17.6% of nuclear cataract cases could be attributable to smoking. Finally, while controlling for smoking, low SES was also associated with higher prevalence of nuclear cataract, while small-sized public housing was associated with higher prevalence of PSC cataract.
Although smoking is unquestionably a major contributor to ill health and death, during the past few decades the incidence of smoking has increased exponentially in many Asian countries and is leading to an equally rapid increase in the incidence of smoking-induced morbidity and mortality in these countries.32 While the prevalence of smoking has been reduced to less than 30% in developed countries such as England and Australia as a result of comprehensive antitobacco measures,33 an estimated 50% to 70% of adult men are reported to be current smokers in some Asian countries,34 which is responsible for a subsequently increased risk of cancer, cardiovascular disease, and respiratory illness.35 In addition, the effect of smoking on ocular disorders such as cataract, diabetic retinopathy, and age-related macular degeneration needs to be well recognized in Asian countries where the smoking prevalence is still at high levels. As in our previous article, current smoking was recorded in more than 40% of Singaporean Malay men and was associated with a high risk of age-related macular degeneration,30 retinal arteriolar emboli,36 and thyroid-associated ophthalmopathy.37
While many studies have reported an association of cigarette smoking with both the prevalence of cataract and/or its progression, these studies have been mainly conducted in white populations in the United States, Europe, and Australia.4,6,8,10,11,13,15,19- 21 Studies from India and China also show that cigarette smoking has been a risk factor for nuclear,10,16,20,28,38 cortical,10,20,38 and any cataract.10,16,20,28,38 The association of other types of tobacco such as cigar,16,39 pipe,15,40 smokeless tobacco,10 or indoor cooking15 with cataract has also been demonstrated previously. In our current study, the fact that cataract is strongly age-related explains the observation that the crude prevalence of cataract was higher in nonsmokers or past smokers (Table 2) owing to their older mean age (59.0 years) than in current smokers, who were younger (55.3 years). However, after adjusting for age or age and other confounding risk factors, current smoking was found to be independently associated with a higher prevalence of all cataract types. Our results, broadly representative of persons living in an urban city environment typical of larger cities elsewhere in Asia, support the notion that smoking is a risk factor for earlier-onset age-related cataract and is consistent with the current knowledge of aging effects from smoking.41- 43
Our study, conducted in Malay individuals in Singapore, and the BMES, conducted in white individuals in Australia, used the same protocol and graders to assess cataract from lens photographs. We therefore took the unique opportunity to examine whether the major associations with age-related cataract were similar between the 2 ethnic groups. While the findings were generally similar between our study and the BMES, we demonstrate a dose-response effect of smoking on the presence of nuclear cataract that was not observed in the BMES study sample (Table 4). Furthermore, association of current smoking and nuclear cataract prevalence was stronger in women than men in our study, with an odds ratio doubled in female current smokers (Table 5). In contrast, the association of current smoking with nuclear cataract prevalence in women was not significant in the BMES (Table 5). The sex differences in these associations across ethnicity samples, which could be owing to random variation in study samples, will need confirmation.
Low SES is now recognized as a risk factor in the development of many chronic eye diseases including age-related macular degeneration, glaucoma, diabetic retinopathy, and cataract.28 Increasing evidence has shown that low SES levels are associated with both the prevalence and progression of cataract.16,20,21,44- 48 In our study, consistent with previous articles, significant associations were observed between nuclear cataract and either low primary education or low monthly income, and between PSC cataract and poorer housing conditions. It could be speculated that people with lower SES have lower rates of cataract surgery, thus accounting for higher cataract prevalence. However, as shown in our previous article,44 no association of any SES factors with cataract surgery was evident in this Malay population. Health care services, including cataract surgery, are readily accessible to and affordable for Singapore residents, with service provision throughout the country to all population sectors. We therefore speculate that other factors are likely to be associated with low SES that are also associated with the development of cataract such as low diet quality with insufficient intake of antioxidants and vitamins.49- 51 Diets consumed by persons with lower SES status tend to be higher in fat and lower in other healthy nutrients such as magnesium, vitamin C, folate, calcium, and iron.49 Protective effects of antioxidants on lens tissue50 including supplements of vitamin C and other antioxidants, have also been found to be associated with a decreased risk of cataract in many observational studies.51,52 Another possible explanation of the association between low SES and the development of cataract is that people with lower SES may have had greater sunlight exposure because they are more likely to have outdoor occupations. As widely recognized, UV-B damage is an established modest contributor to age-related cataract.5 A significant association between outdoor occupations and lower SES level (such as only primary or lower education) was found in our study population (data not shown). When we investigated the association between outdoor occupations and cataract, however, we found no association of any specific cataract type or cataract with occupations (data not shown). Consistent with other studies,5 this absence of association suggests that outdoor occupation may be a poor surrogate for UV-B exposure.
In other epidemiological studies, alcohol consumption, strongly related to smoking, has also been found to be associated with an increased prevalence of nuclear cataract but a decreased prevalence of cortical cataract.39,53 We could not demonstrate any association of alcohol consumption with cataract prevalence. Alcohol consumption is likely underreported by Muslim Malay populations.
Strengths of our study include its population-based sample, the use of a standardized lens grading system to assess lens opacity, and ability to compare data with the BMES. Limitations include its cross-sectional nature and the need to exclude the possibility of a chance finding.
In summary, in this population-based sample of Singaporean Malay individuals, we confirmed the previously reported association of smoking with prevalent cataract, consistent with findings from other Asian populations. Our findings strongly support the implementation of smoking cessation campaigns in urban Asian countries, as the prevalence of smoking remains relatively high in these countries, particularly in men, despite the well-known detrimental effects of smoking on health.
Correspondence: Tien Yin Wong, FRCS, Singapore Eye Research Institute, 11 Third Hospital Ave, Singapore 168751 (firstname.lastname@example.org).
Submitted for Publication: August 4, 2009; final revision received December 22, 2009; accepted January 8, 2010.
Financial Disclosure: This study was funded by the National Medical Research Council grant 0796/2003 and the Biomedical Research Council grant 501/1/25-5, with support from the Singapore Prospective Study Program and the Singapore Tissue Network, A*STAR.
Financial Disclosure: None reported.