Purpose
To assess whether alcohol consumption is associated with the long-term incidence of cataract or cataract surgery.
Design
Population-based prospective cohort study.
Methods
A total of 3654 persons aged 49+ years were examined at baseline and 2564 were re-examined after 5 and/or 10 years. Lens photographs were taken at each visit and assessed using the Wisconsin Cataract Grading System by masked graders. An interviewer-administered questionnaire was used to collect information on alcohol consumption.
Results
No significant associations were observed between alcohol consumption and long-term risk of nuclear, cortical, and posterior subcapsular cataract. However, after adjusting for age, gender, smoking, diabetes, myopia, socioeconomic status, and steroid use, total alcohol consumption of over 2 standard drinks per day was associated with a significantly increased likelihood of cataract surgery, when compared to total daily alcohol consumption of 1 to 2 standard drinks (adjusted odds ratio [OR] 2.10, 95% confidence interval [CI] 1.16-3.81). Abstinence from alcohol was also associated with increased likelihood of cataract surgery when compared to a total alcohol consumption of 1 to 2 standard drinks per day (adjusted OR 2.36, 95% CI 1.25–4.46).
Conclusion
A U-shaped association of alcohol consumption with the long-term risk of cataract surgery was found in this older cohort: moderate consumption was associated with 50% lower cataract surgery incidence, compared either to abstinence or heavy alcohol consumption.
Cataract and cataract surgery represent a significant public health burden in aging populations. Although surgical techniques and subsequent outcomes have greatly improved in recent years, the economic cost of cataract surgery remains substantial. Recognizing modifiable risk factors could potentially reduce this burden. Improved understanding of risk factors for cataract could also help to identify high-risk groups and assist in eye health care planning.
It is estimated that 4% of the total global burden of human diseases is attributable to alcohol consumption. Alcohol has been shown to increase the risk of more than 60 different medical conditions, including mouth and oropharyngeal cancer, esophageal cancer, liver cancer, breast cancer, unipolar major depression, epilepsy, hypertensive disease, hemorrhagic stroke, and liver cirrhosis. On the other hand, mild to moderate levels of alcohol consumption have been shown to have beneficial influences on coronary heart disease, stroke, and diabetes.
The association between alcohol consumption and cataract formation, however, remains uncertain. Alcohol has been shown to disrupt calcium homeostasis in the lens, augment processes such as membrane damage, alter protein-protein interactions, and produce pro-oxidant molecules when metabolized in the liver. It is therefore biologically plausible that alcohol may interfere with normal lens physiology and lead to cataract development.
Several cross-sectional and case-control studies have examined this relationship, with inconsistent findings to date. As most lens opacities develop slowly over an extended period, long-term follow-up studies are needed to establish associations between preceding exposures and the subsequent development of cataract. Very few studies have examined the longitudinal association between alcohol and cataract development and the Beaver Dam Eye Study (BDES) has been the only population-based long-term follow-up study to examine this association. It reported a possible protective association between moderate alcohol consumption and the long-term risk of posterior subcapsular (PSC) cataract. We aimed in this study to assess associations between consumption of various types of alcohol and the long-term incidence of both cataract and cataract surgery in the Blue Mountains Eye Study (BMES) cohort.
Material and Methods
Study Population
Details of the BMES population and its methods are reported elsewhere. In brief, the BMES is a population-based cohort study of vision and common eye diseases in an urban older population comprising 2 postcode areas in the Blue Mountains region, west of Sydney, Australia. This geographically well-defined area has a stable population that is reasonably representative of Australia in socioeconomic status and other measures. All residents of these 2 postcode areas aged 49 years or older were eligible and were invited to participate in the survey.
At baseline examinations (1992–1994), 4433 eligible residents were identified, of whom 3654 (82.4%) were interviewed and examined. Baseline differences between participants and nonparticipants were previously reported. All surviving participants were invited for re-examination after 5 (1997–1999) and 10 years (2002–2004), with 2335 (75.1% of survivors) and 1952 (75.6% of survivors) returning for re-examinations at these times, respectively. Altogether, 2564 participants were followed at least once since their baseline examinations.
Comparisons of participants and nonparticipants at each follow-up examination have previously been reported. Nonparticipants were significantly younger ( P < .0001) and were more likely to be current smokers ( P < .0001) and to have been diagnosed with diabetes ( P = .049). They were also more likely to report lower job prestige index ( P = .041) and were less likely to live in their own home ( P = .0006) than those who participated in follow-up examinations.
Procedures
An interviewer-administered questionnaire was used to collect detailed demographic and medical history data at each visit. Alcohol intake was assessed by questions about the frequency of consuming alcoholic drinks (days per week), the usual number of drinks per day when alcohol was consumed, and the usual type of alcohol (beer, wine, or spirits). Port drinkers were too few to list separately. Drinking patterns were classified into 4 categories (none, ≤1, >1 but ≤2, and >2 drinks per day). The categories were slightly different for spirits (none, ≤1, and >1 drinks/day) because there were no participants in the “>1 but ≤2 drinks/day” category. These categories were formulated based on the recent Australian National Health and Medical Research Council (NHMRC) recommendations of up to 2 standard drinks a day.
All participants underwent detailed eye examinations. At the 5-year and 10-year follow-up visits, participants were re-examined in approximately the same order as that at baseline, using the same procedures and equipment. Slit-lamp lens photographs of each eye were taken using Ektachrome 200 color film (Kodak, Rochester, New York, USA) on a Topcon SL-7E photograph slit-lamp camera (Topcon, Tokyo, Japan) to assess presence of nuclear cataract. Retro-illumination lens photographs were taken using a Neitz CT-R cataract camera (Neitz Instruments, Tokyo, Japan) to assess presence of cortical and PSC cataract. The Wisconsin Cataract Grading System, first developed in 1990 for use in the BDES, was closely followed in performing masked grading of all lens photographs taken at each visit. Inter-grader and intra-grader reproducibility of the lens photograph grading was assessed using quadratic weighted kappa statistics and was shown to be within an acceptable range in our study.
Incident cataract was defined as the appearance of nuclear, cortical, or PSC cataract subtypes in bilaterally phakic participants, in which the corresponding cataract subtype was not present in either eye at baseline. Similarly, incident cataract surgery was defined as cataract surgery performed in either eye of participants who were bilaterally phakic at baseline.
Statistical Analysis
SAS software (SAS Institute, Cary, North Carolina, USA) was used for data analysis. Discrete logistic models estimated associations between consumption of alcoholic drinks at baseline and the 10-year incidence of cortical, nuclear, and PSC cataract or cataract surgery. Moderate alcohol consumption (1-2 standard drinks per day, recommended by the NHMRC guidelines ) was used as the reference level of alcohol consumption. Findings are presented as odds ratios (OR) with 95% confidence intervals (CI). As age and gender are strong predictors of incident cataract and cataract surgery, these 2 variables were included in the initial model. If any significant associations were found after adjusting for age and sex, further adjustments were made for other potential confounders identified from previous studies. These potential confounders were diabetes, steroid use, myopia, socioeconomic status, and smoking.
Results
Tables 1, 2, and 3 show the associations between consumption of various types of alcoholic drinks and the 10-year incidence of nuclear cataract, cortical cataract, and PSC cataract, respectively. After adjusting for age and gender, the consumption of less than 1 standard drink of white wine per day was associated with an increased risk of PSC cataract when compared with consumption of 1 to 2 drinks of white wine per day (OR 2.65, 95% CI 1.06–6.61). This association remained significant after further adjustment for smoking, diabetes, myopia, socioeconomic status, and steroid use (OR 2.75, 95% CI 1.01–7.41) ( Table 3 ). No other significant associations were found between consumption of beer, spirits, and white or red wine and the long-term incidence of the 3 cataract subtypes. Similarly, no associations were found between total alcohol consumption and the long-term incidence of these cataract subtypes.
| Number of Drinks/Day | Number of Incident Cases/Number at Risk | Age- and Gender-Adjusted Odds Ratio a (95% Confidence Interval) |
|---|---|---|
| Beer | ||
| 0 | 62/203 | 1.00 (0.54–1.84) |
| >0 to ≤1 | 75/229 | 1.14 (0.64–2.05) |
| >1 to ≤2 | 19/80 | Referent |
| >2 | 58/223 | 1.25 (0.69–2.27) |
| Spirits | ||
| 0 | 62/203 | 0.82 (0.58–1.18) |
| >0 to ≤1 | 91/308 | Referent |
| >1 | 75/223 | 0.90 (0.64–1.28) |
| Red wine | ||
| 0 | 62/203 | 1.24 (0.55–2.81) |
| >0 to ≤1 | 92/308 | 1.53 (0.69–3.38) |
| >1 to ≤2 | 8/41 | Referent |
| >2 | 32/129 | 1.37 (0.59–3.20) |
| White wine | ||
| 0 | 62/203 | 0.80 (0.46–1.42) |
| >0 to ≤1 | 123/391 | 1.02 (0.60–1.73) |
| >1 to ≤2 | 22/67 | Referent |
| >2 | 77/276 | 0.92 (0.53–1.59) |
| Total alcohol | ||
| 0 | 62/203 | 0.93 (0.57–1.54) |
| >0 to ≤1 | 59/193 | 1.19 (0.72–1.95) |
| >1 to ≤2 | 28/83 | Referent |
| >2 | 188/618 | 1.13 (0.73–1.76) |
| Number of Drinks/Day | Number of Incident Cases/Number at Risk | Age- and Gender-Adjusted Odds Ratio a (95% Confidence Interval) |
|---|---|---|
| Beer | ||
| 0 | 68/305 | 1.69 (0.94–3.02) |
| >0 to ≤1 | 60/304 | 1.58 (0.89–2.80) |
| >1 to ≤2 | 17/106 | Referent |
| >2 | 66/314 | 1.74 (1.00–3.04) |
| Spirits | ||
| 0 | 68/305 | 0.92 (0.67–1.26) |
| >0 to ≤1 | 93/412 | Referent |
| >1 | 61/307 | 0.84 (0.61–1.15) |
| Red wine | ||
| 0 | 68/305 | 1.37 (0.78–2.45) |
| >0 to ≤1 | 82/412 | 1.25 (0.71–2.21) |
| >1 to ≤2 | 11/65 | Referent |
| >2 | 28/173 | 1.17 (0.62–2.21) |
| White wine | ||
| 0 | 68/305 | 0.91 (0.55–1.51) |
| >0 to ≤1 | 128/571 | 0.89 (0.55–1.43) |
| >1 to ≤2 | 23/99 | Referent |
| >2 | 71/369 | 0.91 (0.55–1.50) |
| Total | ||
| 0 | 68/305 | 0.80 (0.53–1.20) |
| >0 to ≤1 | 57/255 | 0.84 (0.55–1.28) |
| >1 to ≤2 | 34/132 | Referent |
| >2 | 175/844 | 0.76 (0.53–1.10) |
| Number of Drinks/Day | Number of Incident Cases/Number at Risk | Age- and Gender-Adjusted Odds Ratio a (95% Confidence Interval) | Multivariable-Adjusted Odds Ratio b (95% Confidence Interval) |
|---|---|---|---|
| Beer | |||
| 0 | 23/350 | 0.71 (0.30–1.65) | NI |
| >0 to ≤1 | 28/355 | 0.88 (0.39–2.01) | NI |
| >1 to ≤2 | 8/112 | Referent | Referent |
| >2 | 19/328 | 0.64 (0.27–1.52) | NI |
| Spirits | |||
| 0 | 23/350 | 0.79 (0.46–1.34) | NI |
| >0 to ≤1 | 33/456 | Referent | Referent |
| >1 | 21/347 | 0.85 (0.50–1.44) | NI |
| Red wine | |||
| 0 | 23/350 | 1.27 (0.41–3.99) | NI |
| >0 to ≤1 | 36/452 | 1.89 (0.63–5.71) | NI |
| >1 to ≤2 | 4/72 | Referent | Referent |
| >2 | 15/186 | 1.52 (0.47–4.87) | NI |
| White wine | |||
| 0 | 23/350 | 1.78 (0.68–4.68) | 2.19 (0.77–6.23) |
| >0 to ≤1 | 56/642 | 2.65 (1.06–6.61) | 2.75 (1.01–7.41) |
| >1 to ≤2 | 4/107 | Referent | Referent |
| >2 | 21/394 | 1.45 (0.54–3.84) | 1.28 (0.44–3.76) |
| Total alcohol | |||
| 0 | 23/350 | 0.60 (0.31–1.18) | NI |
| >0 to ≤1 | 21/308 | 0.66 (0.33–1.31) | NI |
| >1 to ≤2 | 13/154 | Referent | Referent |
| >2 | 60/915 | 0.65 (0.36–1.19) | NI |
a Age was used as a continuous variable. Gender, smoking, diabetes, steroid use, socioeconomic status, and myopia were used as binary variables.
b Adjusted for age, gender, smoking, diabetes, socioeconomic status, steroid use, and myopia.
Associations between alcohol consumption and the long-term risk of cataract surgery are shown in Table 4 . After adjusting for age and gender, no significant associations were found between consumption of beer, spirits, white wine, or red wine and the long-term incidence of cataract surgery. Nevertheless, participants who consumed more than 2 standard drinks of total alcohol per day had an increased likelihood of cataract surgery compared to those who drank 1 to 2 standard drinks of total alcohol per day (OR 2.10, 95% CI 1.21–3.65). This association persisted after further adjustment for smoking, diabetes, myopia, socioeconomic status, and steroid use (OR 2.10, 95% CI 1.16–3.81) ( Table 4 ). Similarly, persons who abstained from alcohol also had an increased likelihood of cataract surgery when compared to those who drank 1 to 2 standard drinks of total alcohol per day (OR 2.13, 95% CI 1.18–3.84). This association also persisted after further adjustment for age, gender, smoking, diabetes, myopia, socioeconomic status, and steroid use (OR 2.36, 95% CI 1.25–4.46) ( Table 4 ).
| Number of Drinks/Day | Number of Incidence Cases/Number at Risk | Age- and Gender-Adjusted Odds Ratio a (95% Confidence Interval) | Multivariable-Adjusted Odds Ratio b (95% Confidence Interval) |
|---|---|---|---|
| Beer | |||
| 0 | 75/430 | 1.02 (0.55–1.89) | NI |
| >0 to ≤1 | 66/435 | 0.89 (0.49–1.65) | NI |
| >1 to ≤2 | 18/129 | Referent | |
| >2 | 39/398 | 0.86 (0.45–1.64) | NI |
| Spirits | |||
| 0 | 75/430 | 1.20 (0.81–1.78) | NI |
| >0 to ≤1 | 68/536 | Referent | |
| >1 | 76/424 | 1.26 (0.85–1.87) | NI |
| Red wine | |||
| 0 | 75/430 | 1.12 (0.53–2.36) | NI |
| >0 to ≤1 | 78/542 | 1.10 (0.53–2.29) | NI |
| >1 to ≤2 | 12/82 | Referent | |
| >2 | 18/213 | 0.69 (0.29–1.64) | NI |
| White wine | |||
| 0 | 75/430 | 1.63 (0.83–3.21) | NI |
| >0 to ≤1 | 115/767 | 1.46 (0.76–2.79) | NI |
| >1 to ≤2 | 14/120 | Referent | |
| >2 | 69/468 | 1.77 (0.90–3.51) | NI |
| Total alcohol | |||
| 0 | 75/430 | 2.13 (1.18–3.84) | 2.36 (1.25–4.46) |
| >0 to ≤1 | 54/368 | 1.82 (0.98–3.39) | 1.77 (0.90–3.48) |
| >1 to ≤2 | 16/171 | Referent | Referent |
| >2 | 163/1109 | 2.10 (1.21–3.65) | 2.10 (1.16–3.81) |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
