Cataract as a global health issue

Cataract is one of the most common causes of preventable blindness in the older worldwide. The term age-related cataract is used to distinguish lens opacification associated with old age from opacification associated with other causes, such as congenital and metabolic disorders or trauma. By far, the most common of these disorders are referred to as senile or age-related cataracts. The prevalence of the disease increases dramatically after age 60. In Spain, cataract extraction is the most frequently performed ophthalmic operation. In the decade up to 1980 the number of cataract surgeries performed in people 65 years old and over doubled1 and in England and Wales2 there was an increased of two-thirds in the decade before 1985.

Over the last 30 years, the elderly population of developed countries has shown an unprecedented increase, with developing countries showing similar trends3,4. In Spain, the percentage of elderly people above 65 years of age has doubled in just 45 years, from 6% in 1950 to 16% in 1998. The predictions indicate that people of 65 years of age and over will experience the highest demographic growth comparing to other groups3. As the average lifespan in the population increases, the incidence of cataract will increase as well.

Although surgical treatment to remove cataract is very effective, cataract surgery carries risks. Moreover, in developing countries cataracts are more common and develop earlier, more than 90% of the cases of blindness and visual impairment are due to cataract5. The incidence of cataract is so high, that new cases will always outstrip the capacity of surgical programs. A preventive approach appears to be essential to the global problem of cataract2.

The etiology and natural history of cataract is unknown. It is considered that cataract is a multi-etiological process. Laboratory investigations suggest that age-related cataract might result from oxidative stress after sunlight exposure6,7. Animal and observational studies suggest that a diet low in antioxidant micronutrients may increase the risk of lens opacification. The possible role of antioxidants as a potentially modifiable risk factor has drawn considerable interest with major public health relevance. Epidemiological studies have also suggested that associated risk factors may include age, sex, race, low socioeconomic status, smoking and alcohol consumption, diabetes mellitus, hypertension, and use of some drugs.

A basis for preventive strategies depends upon adequate understanding of the causes of lens opacification. Strategies to prevent or delayed cataract have not been identified.

This article reviews the epidemiological information about the relation between cataract and antioxidant vitamins and minerals for several important reasons: a) antioxidants might be a plausible risk factor from a biological point of view; b) results from epidemiological studies about the role of different antioxidants are inconsistent, and c) if antioxidants are important factors on cataract prevention, the level of these could be increased through the diet, according to the presumed needs of the organism for antioxidant protection.

In this context, this narrative review examines the evidence from epidemiological studies about the extent to which existing knowledge on the relationship between antioxidant micronutrients and lens opacification can account for the disease in the elderly.

A little about the lens

The primary function of the eye lens is to collect and focus light on the retina. To do so, it must remain clear throughout life. The lens is an encapsulated organ, located posterior to the cornea and receives nutriture from the aqueous humor5. This capsule completely isolates a compact mass of epithelial cells, which form a single layer on the anterior surface of the lens. At the equatorial region the epithelial cells elongate migrate posteriorly, and transform into fiber cells, a process that continues throughout the life of the organism. These fibers elaborate the proteins of the lens called crystallins. Therefore, new cells are formed throughout life, but older cells usually are not lost. Instead, they are compress into the nucleus of the lens5. The nucleus is the region of the lens, which lacks protein turnover opposite to the single layer of epithelial cells, which is incessantly rejuvenated by molecular renewal8. The cellular maintenance and repair of the nucleus depends upon the synthesis and transport of antioxidant enzymes and micronutrients from the cortical area into the nucleus to maintain lens integrity9.

Possible mechanism of lens opacification

As the lens ages, or on stress due to oxidative insults such as exposure to ultraviolet light10,11, environmental pollutants, cigarette smoke, car exhaust fumes, alcohol and certain drugs12 the proteins of the lens are damaged, aggregate and form opacities. It is generally agreed that oxidation of the lens is an important and probable mechanism of cataract genesis in humans13. Oxidation is a biochemical process of loss of electrons associated with another of reception call reduction. Oxidative stress appears when oxidation is excessive14. This mechanism would generate the production of oxygen species, the free radicals, which carry an unpaired electron. These species, which are highly reactive and very unstable, are capable of damaging biologically relevant molecules such as the lens crystallins to depart from their unstable state. In the lens, the free radicals would impair the function of proteolytic enzymes, which normally eliminate the damaged proteins. These proteins, the lens crystallins, would aggregate and precipitate forming lens opacities12. Depending where the opacification is located the cataracts will be cortical, posterior subcapsular, nuclear, or mixed cataracts (multiple locations).

The antioxidant defense system in the eye

To counteract these potentially damaging stresses, the lens has an elaborate antioxidant defence system including vitamin C (ascorbic acid), vitamin E (tocopherol), and carotenoids13. They could interrupt the chain reaction of free radicals if free radical generation is not so great as to overwhelm the defence network. The important presence of some of these antioxidant micronutrients in the lens is well documented. Vitamin A and the major lens carotenoids, lutein and zeaxanthin, are also present in the cataract lens. The findings that oxidation reactions might be important factors in cataractogenesis, and that antioxidants would help to ameliorate that risk implies that antioxidant micronutrients might be expected to prevent or retard the cataract process.

Material and methods

The computerized bibliographic database MEDLINE was used as a source to search for published epidemiological studies on cataract and antioxidants. The search strategy included: a) the key words: «cataract AND antioxidant micronutrients» and covered the year period from 1970 up to the end of 2001. The screening of the articles was done firstly from the titles, secondly from the abstracts and finally from the complete article, and b) citation of other published studies in the different articles found through the MEDLINE database was also used as a source of reference for more articles to be included.

With this strategy I have tried to cover all articles published during the last thirty years. However, although efforts were made to search for all articles published during that period, I cannot guarantee that may be some relevant old or new publications could have been missed.

The present review first focus on observational studies identifying the controversial issues and then, on interventional trials examining the effects of different antioxidant on cataract prevention or development.

Evidence from observational epidemiological studies in the elderly

Though the explanation of high cataract prevalence in one country need not necessarily be the same as that in another, the worldwide distribution of increased risk in both sexes suggests that there may be a common underlying mechanism. Laboratory studies have shown an important role for a protective effect of some antioxidant micronutrients against cataract. However, epidemiological evidence is less consistent. Although some epidemiological studies suggest that low levels of antioxidant micronutrients are associated with an increased risk of cataract, there is considerably less agreement on the role of individual antioxidants15,16. Epidemiological evidence is provided primarily by case-control and cohort studies and more recently by cross-sectional studies and intervention trials. Most epidemiological studies have been conducted in the US and far less is known from other populations such as the European populations with different diet and lifestyles patterns. Across Europe, there is considerable variation in dietary intakes, especially in the consumption of fresh fruits and vegetables, main food sources of antioxidant micronutrients, with Northern European populations having lower intakes compared with «Mediterranean» diets.

Vitamin C

Vitamin C is considered to be the most prominent water-soluble molecule with antioxidant properties. Animal experiments have shown an important role for a protective effect of vitamin C against cataract17,18. The high concentrations of ascorbate present in the normal lenses, cornea, and aqueous humor with 20-30 fold the level found in human plasma suggests that ascorbic acid might be of relevant importance to the eye's health13. The concentration of vitamin C in the lens has been found to be related to vitamin C supplementation. Two studies clearly demonstrated that supplementation with vitamin C significantly increased plasma, aqueous humour, and lens levels19,20. On the other hand, levels of ascorbic acid have been found to be low or absent in cataractous lenses and increased ascorbate free radicals are found in human lens with progression of cataracts21.

Epidemiological studies have investigated the association between dietary or plasma vitamin C, or vitamin C supplements and cataract. Most studies do not give percentile cut-offs to define the dietary vitamin C categories nor blood vitamin C levels but rather categories of high, medium or low. Table 1 shows a selection of the most relevant studies about vitamin C and cataract and discussed in this section.

Several studies have examined the relation between serum vitamin C levels and risk of cataract. One small North-American case-control study found that levels lower than 0.80 µmol/l of ascorbic acid were associated with an increased risk for cataract (OR = 11.3; p < 0.10). However, this study had methodological problems, numbers were too small (77 cases and 35 controls) and confidence intervals were not given22. Two studies found an inverse association. In the National Health and Nutrition Examination Survey II (NHANES II)23 a strong inverse association between serum ascorbic acid and self reported cataract was found whereas in the Nutrition and Vision Project24 a marginal inverse association was found with prevalence of nuclear cataract. In contrast, in the Baltimore cross-sectional study25 and the population-based POLA study26 no association was found for blood ascorbic acid levels and any type of cataract. On the other hand, a study, which was conducted in India, found that high levels of ascorbic acid were associated with an increased risk of cataract (OR = 1.87; 95%CI, 1.29, 2.69)27. In a case-control study28 conducted in a Spanish Mediterranean region, blood levels of ascorbic acid above 49 mmol/l were found to be associated with a 64% reduced risk of cataract (OR = 0.34; 95%CI, 0.23, 0.50; p < 0.0001).

Several studies have reported results for dietary vitamin C from FFQs. The lens opacities case-control study found a 52% reduction of risk of nuclear cataract for those having the highest intake of vitamin C compared with those having the lowest intake. Intake was grouped as low, medium, and high using the lowest quintile, the three middle quintiles, and the highest quintile. ORs compared high vs. low using the latter as the reference group29. Jacques and Chylack (1991)22 found that low vitamin C intake was associated with an increased risk of cortical cataract (OR = 3.7; p < 0.10) and posterior subcapsular cataract (OR = 11.0; p < 0.05). However, the sample size was too small and the associations observed did not reach statistical significance.

Four studies showed no association between dietary vitamin C and cataract30-33, whereas in the Beaver Dam Cohort Study the ORs for men for a protective effect of high levels of vitamin C intake (fifth quintile vs. lowest quintile, median 78 mg/day) were 0.71, although only significant when combined with the vitamin C supplements34. The prospective Nurses' Study found no association with dietary vitamin C but supplement use of vitamin C for more than 10 years was associated with a 70% reduction in cataract extraction31. Later on, in the Nutrition and Vision Project24, the relationship between newly diagnosed nuclear cataract and usual nutrient intake was examined. A significant inverse association was seen between nuclear opacities and use of vitamin C supplement for more than 10 years. This result corroborates earlier work from the same study31,35. These results are consistent with the results from the Beaver Dam Eye study where 5-year risk for nuclear and cortical cataract was 40 and 60% lower among persons who reported the use of any supplement containing vitamin C for more than 10 years36. In contrast, no effect of vitamin C supplement was observed in the large prospective cohort Physicians' Study37 and the longitudinal study of cataract38,39.

In summary, the evidence from epidemiological studies using food frequency questionnaires is insufficient. They either have found an association between vitamin C intake and cataract22,29 or a weak one28,34 or did not find any association30,33.

Results on blood ascorbic acid are also inconsistent. Three studies found an association22,23,28 with no association reported in some studies25,26 while a study in India27 found an increased risk with higher intake. However, none of the other large cohort studies31,36 have analyzed blood vitamin C. There is some evidence that long-term supplement use of vitamin C is associated with lower nuclear cataract risk24,29,36, although no effect was seen in other studies37-39.

Vitamin E

Vitamin E, a term that encompasses a small group of related tocopherols is also found in the lens but only in concentrations similar to that of plasma40. It is believed that most of the vitamin E in the lens is in the epithelial cell membranes and could help to prevent peroxidative damage to these membranes41. It has been suggested that peroxide damage to cellular membranes not only accompanies its development but is the initiatory cause of cataracts42. Vitamin E is one of the antioxidant liposoluble vitamins found in human lenses20,43. Experimental studies in animals have demonstrated that vitamin E is able to reverse cataract formation to some degree suggesting a protective role for vitamin E44,45.

Evidence for an inverse association of dietary vitamin E with cataract from epidemiological studies is inconsistent. Table 2 presents a summary of the main studies which found a positive association between antioxidant nutrients and cataract or type of cataract.

Food frequency questionnaires have been used by those studies investigating a role for dietary vitamin E in the cataract process. Three studies showed an association with cataract. In a small case-control study subjects with intakes up to 35.7 mg/day of vitamin E had an increased risk of 2.7 for cortical cataract, which was not statistically significant22. In the LOCS study29 high levels vs. low levels of vitamin E were protective for cortical and mixed cataract (OR = 0.59, 95%CI, 0.35-0.99; OR = 0.58, 95%CI, 0.37-0.93). In an Italian case-control study a reduction of 50% in the risk of cataract extraction was seen for those in the fifth quintile of vitamin E intake vs. those in the lowest quintile group32. Two studies found no associations between dietary vitamin E and cataract27,31, whereas two other studies could not examine vitamin E intake because the food databases used were either insufficient25 or did not provide vitamin E content of foods30. The relation with nuclear cataract was also investigated. No association was found between vitamin E intake and incident of nuclear opacities in the Beaver Dam cohort46 whereas in the Nutrition and Vision Project there was a decreased risk of nuclear cataract that was non-significant24.

Epidemiological studies show some evidence for a benefitial role of supplements of vitamin E on cataract risk. In a small case control study supplements with vitamin E during the preceding 5 years showed a 55% reduction in risk of cataract extraction47. In the Beaver Dam Cohort study use of multivitamin supplements containing vitamin E for more than 10 years was inversely associated with nuclear cataract36 and in the longitudinal study of cataract vitamin E supplements were associated with lower risk of increased nuclear opacification during a 5-year follow-up38. However, the Physicians' Health Cohort Study37 showed no association between users of vitamin E supplements and risk of cataract.

Results for blood alpha-tocopherol levels show some evidence for a beneficial role of vitamin E. Two studies found a protective effect of plasma levels of alpha-tocopherol and risk of cataract. A US case-control study48 found that people with high levels of alpha-tocopherol (42 µmol/l) had a 56% reduced risk of nuclear cataract. In that study however, analyses were not adjusted for cholesterol. In the Baltimore cross-sectional study25, middle (from 18.57 to 29.72 µmol/l) and high levels (> 29.720 µmol/l) of plasma alpha-tocopherol were associated with a decreased risk of nuclear cataract of 45% and a decreased of 48% for cortical cataract. During the 5-year follow-up Leske and co-workers38 reported a decreased risk of nuclear opacities with higher plasma vitamin E concentrations. Inverse associations were found between plasma concentration of vitamin E and nuclear cataract in the Beaver Dam cohort46 and the Nutrition and Vision Project study24. A Finish case-control study showed a marginal association with risk of cataract49, whereas two epidemiological studies did not find any association between plasma alpha-tocopherol levels and risk of cataract27,30.

Vitamin A

Although the presence of retinol has been confirmed in human lenses43 results from epidemiological studies are controversial. Two epidemiological studies did not find a protective effect for plasma retinal22,49. In the Baltimore study25 serum retinol levels above 1.3 µmol/l (at both middle and high levels) were associated with an increased risk, approximately of two fold, for cortical cataract whereas in the POLA study26 strong inverse association with cataracts was found. Dietary vitamin A in other studies was collected using food frequency questionnaires either self-administered or through interviews. Four studies investigated dietary vitamin A but did not analyze plasma retinol30-32,48. Of these, two were associated with a decreased risk and two were not. In the Lens Opacities Case-Control Study29 dietary vitamin A was associated with a decreased risk of 55% for nuclear and 40% for mixed cataract. In the Nurses Health Study31 dietary vitamin A was associated with a decreased risk of 39% for cataract extraction, whereas no association was seen in two different Italian studies30,32.

Carotenoids

Carotenoids are also lipid-soluble antioxidants that accumulate in cellular membranes and like vitamin E are thought to play a role in maintaining cell membrane integrity, although it has yet to be demonstrated in the lens50.

While as many as 42 carotenoids appear to be available in the diet, usually only six of the carotenoids present in most populations are assessed in human serum51. These are: beta-carotene, alpha-carotene, beta-cryptoxanthin, lycopene, zeaxanthin and lutein.

Experimental studies give support for a beneficial effect of carotenoids as antioxidants, in particular beta-carotene, as being highly effective quenchers of singlet oxygen, especially at low partial pressures of oxygen (15 mmHg)52,53.

So far, studies showed that lutein and zeaxanthin are the only carotenoids present in the human lens43. There are few epidemiological studies which have investigated the role of individual carotenoids intake in the cataract process, mainly due to the lack until recent years of food composition tables, which contained information on individual carotenoid content of foods.

β-carotene

Blood beta-carotene levels accounts for 10-20% of total serum carotenoids and its concentration reflect recent intake54. Evidence for a protective role of beta-carotene in the development of cataract has been shown in experimental studies both in vitro and in vivo55,56. However, no detectable beta-carotene concentrations have been found in human lens extracts20.

So far, epidemiological evidence for a role of beta-carotene is scarce and insufficient. To date, very few epidemiological studies have studied the relation between dietary and/or blood beta-carotene and cataract. No associations have been found between dietary beta-carotene and cataract extraction in the Nurses' Health Study and in a cross-sectional study conducted in northern Italy31,32.

Few epidemiological studies have examined the relation between levels of blood beta-carotene and cataract. In the Baltimore Longitudinal Study of Ageing25, blood levels of beta-carotene were not associated with risk of nuclear or cortical cataract, whereas in a case-control study in Finland, levels lower than 0.095 µmol/l in men and 0.114 µmol/l in women were found to increase risk of cataract. The OR for the lowest thirds of the distribution relative to the higher thirds was 1.7 (0.8-3.8)49. In the Nutrition and Vision Project study24 a decreased risk for nuclear opacities were found which was not significant.

Lycopene

Lycopene, one of the major carotenoids in human serum and other tissues, has not been detected in human cataractous or normal lenses20,43. In vitro, studies have shown that lycopene is as a very effective circulating singlet-oxygen quencher57.

Very few studies have investigated the role of lycopene in the aetiology of cataract. In the Beaver Dam Eye Cohort study34, lycopene intake was measured by means of a food frequency questionnarie. Intake of energy-adjusted lycopene was related to nuclear sclerosis: women in the highest quintile had an increased risk of 51%. This relation was similar in direction in men and women though significantly only in women. Quintiles cut-off points were not shown. The authors also observed adverse associations with the intake of tomato products. In the Spanish case-control study moderately high levels of blood lycopene (> 0.30 mmol/l) were associated with a 46% increased odds of cataract (p = 0.04)28. High lycopene levels associated with risk of cataract could also reflect other non-nutritive aspects of diet. Carotenoids other than those with provitamin A activity, food additives, food substitutes, pesticides or products formed in the processing of tomatoes, as well as other lifestyle factors, might have been measured through the blood lycopene levels. High serum levels of this carotenoid were also related to more severe nuclear sclerosis in a cross-sectional analysis in the same population.

Lutein and zeaxanthin

Lutein and zeaxanthin have been the only carotenoids to be detected in human lenses20,43.

In the Nurses' Health Study31, spinach, which is rich in lutein and zeaxanthin, was consistently associated with a decreased risk of cataract extraction.

In the Beaver Dam Eye cohort46 there was a strong inverse association between past lutein intake and incidence of nuclear cataract and a 30 to 40% reduction in risk of incident nuclear opacities for person with serum levels of lutein in the highest tertile relative to those in the lowest tertile. Results from the Nutrition and Vision Project study24 were consistent with those from the Beaver Dam study. However, in both studies this association was not clearly independent of the relation between cataract and vitamin C intake. In a prospective study58 a modest decreased of risk of cataract extraction was seen in men in the highest quintile of lutein and zeaxanthin intakes comparing to the lowest fifth. In the Nutrition and Vision Project study24, there was a decreased risk of prevalence of nuclear cataract in the highest quintile of lutein/zeaxanthin intake. However, no association was found in other studies32,34.

Selenium

Although animal experiments have shown evidence for a protective role of selenium on cataract formation59 and selenium levels seem to be lower in serum as well as in aqueous humour of cataractous patients60, no association have been found in epidemiological studies so far28,30,48,49.

Alpha-carotene, Beta-cryptoxanthin and zinc

Alpha-carotene, beta-cryptoxanthin and zinc are considered to be part of the eye antioxidant defence system. Very few epidemiological studies have been performed to examine the role of alpha-carotene, beta-cryptoxanthin and zinc on caratact risk. One study showed a significant inverse association of serum alpha-carotene and beta-cryptoxanthin and risk of nuclear cataract for men who smoked34. However, no other studies have found a positive association between these antioxidants and risk of cataract.

From the studies reviewed in this section the strongest evidence for a possible beneficial role should be for blood alpha-tocopherol levels24,25, 38,46,48,49 and long term users of multivitamin supplements, in particular vitamin C24,29,36 and vitamin E36,38,47.

Weaker associations have found between dietary vitamin C and/or E and cataract. High levels of dietary vitamin E were found to be a protective antioxidant in several studies22,29,32 but no association was found in others27,31,33. Dietary vitamin C was not associated with risk of cataract in several studies30-33, while women in the highest quintile group of lycopene had an increased risk of nuclear opacities34. High levels of dietary intake of some carotenoids have been found to be protective against cataract. Spinach, a source of lutein-zeaxanthin, was associated with a decreased risk for cataract extraction in the Nurses' study31. In the Beaver Dam33 and the Nutrition and Vision Project24 studies, high past intakes of lutein were associated with lower risk of nuclear cataract however, no association was found in other studies32,34.

Finally, some studies have computed an antioxidant index either in plasma or a dietary antioxidant index or both. Because oxidative mechanisms are thought to play a role in cataract formation, the nutrients and enzymes in the eye natural antioxidant defence system, may act synergistically27. The majority of these studies have found an inverse association with an antioxidant index and cataract22,27,29 but not all30.

Intervention trials

Problems caused by confounding and bias in epidemiological studies are of less concern in randomized clinical trials. Moreover, in observational epidemiological studies, the high degree of correlation between different nutrients make it difficult to identify which nutrient might be important in any observed association.

So far, only limited and inconsistent data are available from such trials. Several other ongoing large-scale randomized trials are underway and will provide complementary information to the total body of evidence on the benefits and risks of supplementation with vitamins. Clinical trials are expected to give relevant information on the role of single antioxidants micronutrients on the prevention of cataract.

The Alpha-tocopherol Beta-carotene Cancer Prevention (ATBC) study61, was a randomized placebo-controlled clinical trial originally designed to study the effect of alpha-tocopherol and beta-carotene on the incidence of lung and other cancers. A cross-sectional end-of-trial eye examination of 1,828 participants from age 50 to 69 years was conducted to assess the effect of the supplements on cataract prevalence. No effect was found of either vitamin E or beta-carotene on cataract prevalence after a median supplementation of 6.6 years. The strength of the conclusions in this study is limited due to the lack of information on cataract status for each participant at the start of the supplementation. The other trial was conducted in China in a nutritionally deprived population. A beneficial effect for nuclear cataract was noted after 5 or 6 years of supplementation with multivitamins and minerals. However, generalization of the results from this study is made difficult due to several reasons: the nutritionally deficient nature of the study population; a formal eye examination was performed only at the end of the 6 year-trial, no examinations were performed at the beginning of the trial to ensure that all individuals were cataract free and very large number of statistical comparisons were made, hence compromising the alpha level62.

A large randomized trial of 15,000 US physicians aged 55 years and older63 noted no effect on cataract incidence or cataract extraction after 13 years of beta-carotene use. This trial has been designed to test alternate day beta-carotene, alternate day vitamin E, daily vitamin C, and a daily multivitamin, in the prevention of total and prostate cancer, cardiovascular disease, and cataract and macular degeneration.

Recent results of the Age Related Eye Disease (ARED) study64 found no effect after an average treatment of 6.3 years with high doses of vitamins C and E, beta-carotene, and/or zinc on the development or progression of age-related cataracts or cataract surgery. The ARED study is a US multicentre factorial double-blind, placebo-controlled trial enrolling 5,000 patients from age 55 to 84 years old. Interpreting the ARED study results also requires a consideration of some limitations. A considerable number of ARED study participants had a grade of opacity at baseline (52% had some cortical opacities, 15% nuclear and 10% PSC). It may be that the ARED study intervention was started too late in the process for it to be effective and that cataracts had probably started to develop. The ARED study participants will be followed up for at least another 5 years.

So far, clinical trials have provided little support for a beneficial effect on cataract development. Ongoing clinical trials will provide additional data about whether these antioxidants or other nutrients are important in the cataract process.

Discussion

So far, the available epidemiological evidence does not provide a basis for recommending dietary change for the general population. Errors in the assessment methods could account for the lack of consistency seen among the reviewed epidemiological studies. Moreover, because of the long-term and multifactorial nature of cataract, it is difficult to predict the influence of nutrients on the cataract process when diet is measured at only one time point. Epidemiological studies have assessed the antioxidant status of individuals by measuring intake using food frequency questionnaires and/or by measuring blood antioxidant's levels. In general, the results from the food frequency questionnaires data showed a smaller effect of antioxidants on cataract risk compared with the blood data analyses. The smaller magnitude of associations based on the food frequency questionnaires antioxidant levels may have resulted from the error in the measurement of «usual» dietary intake with the questionnaire, while data obtained from blood analyses are likely to have been less susceptible to error. The resulting error in measuring antioxidant intakes from the food frequency questionnaire be expected to lead to a dilution of the magnitude of association. Food frequency questionnarie's responses are based on memory and the participants may have had difficulty in accurate classification of intakes across a wide range of food products. Of more concern is «recall bias» where errors in reporting food intakes occur because of biased reporting due to knowledge of having cataract. This is of increasing concern given the current widespread publicity about diet in relation to certain diseases. However, this situation seems unlikely. So far doctors do not give dietary advice for patients with cataract. There is no obvious population awareness of possible dietary influences on cataract risk. Other errors might result from inaccurate food composition data and/or incomplete listing of foods in the food frequency questionnaire leading to errors in estimates of the average individual caloric and nutrient intake. For instance, it has only been recently that a few food composition tables have started to incorporate individual carotenoids content of foods. Most of the food composition tables provide data on total vitamin A activity or total carotenoid content of foods and tend to overestimate total carotenoid content and subsequently, vitamin A activity of plant foods.

Observational studies are subject to confounding and design biases. The majority of epidemiological studies reviewed here took account of main potential confounders such as smoking and alcohol consumption, sun exposure, body mass index, iris colour, education energy intake, and other antioxidant micronutrients. However, confounding bias may remain if these variables were measured with error or were classified into broad categories or there may be other unknown factors that protect for lens opacification. On the other hand, some of the studies reviewed here were lacked of statistical power and some real associations could be missed.

Differences in the definition of case patients could also account for the different results found in this review. In some studies, cases were subjects with any grade of lens opacities; in others, they had more severe cataract that required extraction.

Finally, more data on the potential oxidative activity of these antioxidants is needed to better understand the relationship between antioxidants and cataract. It is known from experimental studies, that some of these antioxidants might be acting as oxidants under specific circumstances. Some data have suggested potential adverse effects of vitamin C in the lens65.

Conclusions

Given the scarcity of available clinical trial data, and the inconclusive results from observational studies about the individual role of the antioxidant micronutrients on cataract risk, no recommendations on supplements should be made. However, epidemiological studies suggest that a diet low in antioxidants might be associated with an increased risk of cataract. It seems therefore prudent to recommend diets rich in fruits and vegetables with high consumption of vitamins C, E and carotenoids. Clinical trials of sufficient size and duration are needed before such recommendations are made. Moreover, more data on the potential oxidative effect of these antioxidants in the lens must be gathered until we confidently could recommend dietary changes or the use of supplements for the prevention of cataract. Other population studies and clinical trials are under way which may elucidate the possible role of these antioxidants on risk of cataract.

Acknowledgments

I am most grateful to Professor Astrid Fletcher, from the London School of Hygiene and Tropical Medicine, for scientific advice.