Our results did not show an association between occupational sunlight exposure when all types of cataract were combined making them similar to the findings of other studies [11
]. When examining the risk of cataract separately by type, we found an association between occupational years of outdoor exposure and nuclear cataract. In particular, uncorrected outdoor exposures above 1.42 years (i.e. the top fifth) showed an increase in odds, greater than threefold (OR = 3.68; 95% CI = 1.50–9.01). Similarly, corrected exposure above 0.80 years (i.e. the top fifth) was found to increase the odds by a factor of about 3 (OR = 3.19; 95% CI = 1.24–8.21).
Our results are not consistent with an association of sunlight exposure and cortical cataract [6
] or with the negative association with nuclear cataract reported by some [6
We are only aware of three studies that have suggested an association between sunlight exposure and nuclear cataract. A French study found a weak association of this type in a coastal Mediterranean population, in some respects similar to ours [16
]. The methodology used in the French-study to measure UVR exposure was comparable to our study but included meteorological data on ambient solar radiation. Although we did not have this information we think it unlikely that it would have altered our results in any significant way. Most our participants (97%) lived all their entire life in Valencia and therefore are not expected to show much heterogeneity in UVR exposure other than that provided by occupational and leisure activities.
In Nambour, Australia, the presence of nuclear cataract later in life was strongly associated with occupational sun exposure between ages 20 and 29 (OR = 5.9; 95% CI = 2.1–17.1). The authors reported that the strength of the association between sun exposure during the fourth and subsequent decades of life and cataracts was greatly reduced after adjusting for sun exposure between ages 20 and 29 [13
]. In a Japanese study, lifetime UVB exposure was investigated in relation to type of lens opacities. Lifetime cumulative UVB exposure and exposure after the teenage years correlated with the presence of nuclear opacities later in life in females [14
Obtaining accurate measurements of personal past sunlight exposure is not a trivial task. One limitation of our study lies in the design of the questionnaire itself, particularly, when used to collect information from certain specific groups of participants. Recall was particularly difficult for those who did not have a type of work defined by a fixed schedule, for example housewives. The majority of the women participating in this study were housewives (87%), and most had been housewives all their life. However, some evidence of the validity of our questionnaire was provided by the positive association between levels of education and type of occupation, and the measures of years of outdoor exposure. For instance, farmers showed the highest median sunlight exposure (n = 86, median = 2.10 years), while people working indoors such as receptionists or shop assistants had the lowest (n = 42, median = 0.89 years). Further, higher educational level was inversely associated with years of exposure to sunlight: illiterate people had the highest median exposures (n = 132, median = 2.0 years) while those who went to high school or university had the lowest (n = 33; median = 0.93 years).
Factoring in ocular protection factors did not alter the original results. In general, the practice of wearing protective devices was quite similar between the cases and the controls. Thus, 52% of our cases, as compared to 48% of our controls, reported having ever worn spectacles, 53% of cases and 47% of controls, a hat, and 47% of cases and 53% of controls, sunglasses. Several studies have shown that wearing a hat or sunglasses reduces the amount of light entering the eye [21
]. However, in Spain, there is no medical advisory from either ophthalmologists or public-education campaigns indicating the possibility that sunlight could contribute to the onset of cataracts. If there were a misclassification of years of outdoor exposure, there would be no reason to believe that this would be greater in people with cataracts. Inaccuracy in reporting is likely to have similarly occurred in both cases and controls, leading to a reduction in the possible magnitude of the odds ratio.
The association between sunlight exposure and nuclear cataract is difficult to explain solely on the basis of direct absorption, since the damage would be expected to appear in the cortical area first, where most of the UV rays reaching the lens would be absorbed by UV filters [31
]. A combination of internal (generated within the nucleus) and external oxidative factors (outside the nucleus), with aging being by far the major risk [32
], might contribute to the aetiology of nuclear cataract. The lens is a tissue that maintains every fiber cell ever formed and continues to grow throughout life with the old fibers compressed into the center of the nucleus and the cortical area composed of the newly formed fibers [31
We found that early occupational exposure to sunlight, from 25 to 45 years of age, increased the risk of nuclear cataract later in life. UV filters concentrations decreases at approximately 12% per decade [32
]. It may be that lifetime cumulative sunlight exposure in conjunction with internal factors, e.g. the development of an internal barrier in middle age, may play an important role in the development of nuclear cataract in adult life. Our findings showed no increased risk of cortical cataract as a result of outdoor exposure. Another possibility involves dietary vitamin C intake, which among our study population was high (157 mg/day) [18
]. This high intake has also been observed in another Spanish population from the same geographical area [33
] and exceeds the intakes of other populations [34
]. High vitamin C intake might especially have protected the cortical area of the lens from UVR oxidative stress. Indeed, our antioxidant analyses showed that plasma ascorbate levels greater than 49 μmol/L were associated with a 64% reduction in the risk of cataracts and this protective effect was consistent across all types [18
We did not find evidence for any interaction between levels of outdoor exposure and antioxidant intake, but the power of our study to detect such interactions was very low. Finally, the possibility of residual confounding and chance are alternative explanations for our findings. The small numbers of cortical (40) and posterior (62) cases could have increased the risk for Type II error, though less so for nuclear cases (100).