Effects of Ultraviolet Light on the Eye : Role of Protective Glasses by

Global atmospheric changes such as depletion of ozone in the stratosphere are thought to lead to increased levels of ultraviolet radiation on earth. This can have adverse effects on human health, and long-term effects of ultraviolet light on the eye are of increasing concern. Ultraviolet light eposure to the eye has been associated with cataract formation and retinal degeneration. In both cases, it is hypothesized that ultraviolet light can initiate formation offree radicals, which can cause protein modification and lipid percxidation. Several procedures can be recoInended to prevent ultraviolet light damage to the eye, such as the use of suitable protective glasses when outdoors.


Introduction
Several effects of exposure of the eye to excessive light have been well established, including photokeratitis, erythema ofthe eyelid, cataract, solar retinopathy, and retinal damage (1,2). This paper discusses the role of ultraviolet (UV) light exposure in cataract formation and retinal degeneration.
Global atmospheric changes such as stratospheric ozone depletion may increase the level of UV radiation on earth (3). This would contribute to enhanced chronic exposure of the skin and eye tissues of humans to UV light (1,3). Hazardous effects ofUV light on the eye were already recognized in 1920 by van

der
Hoeve (4). Biochemical changes that take place in eye tissues during light exposure are discussed, with emphasis on alterations in the lens and the retina.
One mechanism that is thought to play an important role involves generation of oxygen radicals, which can cause lipid peroxidation and protein modification. This process is also thought to be related to aging and many human diseases, including cancer, atherosclerosis, and arthritis (5,6).
Lipid peroxidation has received an enormous amount ofattention, and Esterbauer illustrated the enormous increase in the numbers ofpapers published in this field between 1972 and 1985 (7). Unfortunately, most of the research exploring the role of lipid peroxidation in human disease states is based on the measurement of malondialdehyde (MDA) by its reaction with thiobarbituric acid (TBA). This method, often called the TBA test, is nonspecific and not very sensitive (8).
During previous studies on lipid peroxidation in retinal degeneration, highly sensitive and specific gas chromatography-mass spectrometry methods were developed for detecting phospholipid peroxides and aldehydic products of lipid peroxidation (9)(10)(11)(12). It was found that these products of lipid peroxidation are present in vitamin E-deficient and -supple-*Department of Chemistry and Biochemistry, 108 Gaines Hall, Montana State University, Bozeman, MT 59717. mented rat retinal tissues (13)(14)(15). Oxidized fatty acids in membrane phospholipids significantly alter the permeability of cell membranes, which may contribute to malfunction of cells and eventually to cell death (16). Vitamin E is an antioxidant known to inhibit lipid peroxidation, and vitamin E deficiency can cause structural changes in membranes (17). In addition, aldehydic products of lipid peroxidation, especially 4-hydroxyalkenals, are very reactive and combine with amino and sulfhydryl groups of proteins (18,19). Inactivation ofproteins due to modification by 4-hydroxyalkenals is another mechanism by which retinal damage may occur through lipid peroxidation (13).

Cataract
A cataract is an opacity of the lens, which can vary in degree ofdensity and is usually associated with aging (senile cataract) (20). There are three major types of senile cataract: a) nuclear cataract in the nucleus ofthe lens; b) cortical cataract in the surrounding cortex; and c) posterior subcapsular cataract, which occurs beneath the posterior capsule ofthe lens. Other causes of cataract are associated with systemic diseases (21), such as diabetes, galactosemia (22), microwave or ionizing radiation (23), trauma, the use of photosensitizing drugs (24), and UV light (1,20,25).
It is thought that reactive oxygen species and free radicals (26) may play a role in the formation ofcataracts by oxidation oflipids (27) and modification ofproteins (28). Lipid peroxidation can be inhibited by various antioxidants such as vitamin C and vitamin E, the glutathione system, superoxide dismutase, and catalase (29). These antioxidants deplete in the lens with increasing age (30), which is thought to contribute to cataract formation due to increased oxidative stress (31). It is known that antioxidants are decreased (vitamin C) or absent (glutathione) in the cataractous lens. Recently, Jacques et al. reported that the oxidant systems may play a role in cataractogenesis (32,33). They found that individuals with high plasma levels of two or more vitamins that influence the antioxidant status (vitamin C, vita-min E, carotenoids) appear to have a reduced risk of cataract (33). There appeared to be no difference in the levels of erythrocyte superoxide dismutase, glutathione peroxidase, and glucose-6-phosphate dehydrogenase enzymes between subjects with and without cataracts (33). These results suggest that the level of lens antioxidants could perhaps be influenced by the plasma levels of antioxidants.
The decrease of antioxidant levels in the human lens could lead to an increase of lipid peroxidation products. Bhuyan et al. (27) found an increase of malondialdehyde in the human lens with age. However, Borchman et al. found no changes with age in the concentration ofa phospholipid malondialdehyde adduct, which was also measured as an indicator of lipid peroxidation (34). They argue that hydrolysis of the adduct may explain its constant levels. Perhaps the specific gas chromatography-mass spectrometry methods for measurement of lipid peroxidation products (9-12) may provide a better tool to test the hypothesis that lipid peroxides increase in the human lens with age.
Alteration of lens proteins is another biochemical mechanism that has been proposed following UV radiation (28,35,36). These biochemical changes can occur through a direct effect of UV light or by virtue of photosensitivity (35,37). Studies on animals and humans have shown several effects. A 90% decrease in Na+K+-ATPase activity was found following 20 hr of in vitro exposure to UV radiation of 365 nm at 5 mW/cm2. Chronic exposure to UV light was found to lead to abnormal lens epithelial cell differentiation and enhancement of aggregated proteins in mice lenses within several months, and cortical cataracts developed after 1.5 years (38). Grey squirrel lens have also been employed because they have a yellow near-UV absorbing pigment similar to that present in the human lens. Following UV exposure invivo, aggregation of soluble proteins and enhancement of low molecular weight peptides was found. In invivo studies, near UV radiation enhanced levels of insoluble protein. Fluorescence properties changed because of a decrease of the usual tryptophan, and an increase of nontryptophan fluorescence was found. Goosey et al. reported theformation of singlet oxygen as a major mechanism by which lens proteins were aggregated (39), and Jedziniak et al. found that singlet oxygen altered the properties of the crystallin component of the human lens, in addition to damage to several associated enzymes (40). Dillon et al. identified the crosslinks of insoluble proteins from brunescent cataracts as tryptophan oxidation products (41).
During recent years, there has been increasing evidence showing that UV radiation could be a risk factor in the formation of human cataracts. The mechanisms are unknown, but could depend on protein aggregation due to free radical reactions as discussed above. It has been proposed that chronic exposure to ambient UV radiation results in generation of a series of fluorescent chromophores. These chromophores are associated with a deepening yellow color of the lens nucleus (37). Furthermore, acute exposure of the lens to higher levels of UV light than normally present in the environment has been implicated in cortical opacities (25). A relationship between sunlight exposure and cataracts has been suggested, based on basic chemical studies (25,26,35).
Epidemiological studies supported the hypothesis that persons who live in geographic areas where there are long periods of sunlight (42) have an increased incidence of cataract (20,43-45).
Zigman et al. reported that exposure to sunlight specifically enhances the development of dark brown cataracts in humans (43). Recendy, more detailed epidemiological surveys have been performed, and the relation between UV light exposure and the development of senile cataract was quantified (46,47). These studies were based on a protocol to measure the dose of UV radiation in different occupational groups (48).
UV light is usually divided into three portions, UV-C (200-290 nm), UV-B (290-320 nm), and UV-A (320-400 nm). An association between cumulative exposure to UV-B radiation after age 15 and cortical cataract formation was recently reported by Taylor et al. (46). They did not find an association between nuclear cataract and UV-B exposure or between any type of senile cataract and UV-A exposure (46). Since relatively few subjects with posterior subcapsular cataracts participated, no attempt was made in their study to correlate this type of senile cataract to UV light exposure. However, Bochow et al. reported that a history of high exposure to UV-B is associated with an increased risk for development of posterior subcapsular cataract (47).
Treatment of cataract has been dramatically improved with modem surgical techniques. The lens is typically removed from the eye (lens extraction) and is usually replaced with an intraocular lens. Since cataracts can be treated with a high success rate, research efforts on the biochemical pathways behind cataract may have been reduced. Although cataracts are now effectively treated in Western societies, cataract is still the leading cause of blindness in third world countries (20). Surgical treatment is not available worldwide and operations for cataracts have nearly doubled during the past decade in wealthy countries using an increasing proportion of resources for health (20). Therefore, research efforts on finding the cause of cataract should be maintained, this work could eventually lead to effective treatment plans or recommendations for lifestyle changes that may preventthis disease worldwide.

Retina Degeneration
The retina is sensitive to light and can be damaged mechanically, thermally, and photochemically (49). Mechanical damage results from short-term high irradiance levels such as obtained with YAG laser photodisrupters. Thermal damage occurs when absorbed light raises the temperature in the retina by 10 to 20(C and also depends on short-term exposure. Photochemical damage is induced by relatively long-term exposure to lower levels of light in the UV and blue regions of the spectrum. To understand why photochemical reactions may occur in the retina following light exposure, it is helpful to review the structure of the retina. The retinal photoreceptors convert light into an electrical signal, which is sent to the occipital cortex. The photoreceptor outer segments consist of a plasma membrane that surrounds a pile of disc membranes. The discs are lipid bilayers containing the lightsensitive protein rhodopsin (52). The phospholipids of the disc membranes have a unique fatty acid composition. About 50 mol % of the fatty acid content of the disc consist ofdocosahexaenoic acid (22:603) (53), which is the most highly polyunsaturated fatty acid (PUFA) known. Susceptibility to lipid peroxidation increases in proportion to the amount of double bonds (54). Light exposure may also be a factor that plays a role in promoting lipid peroxidation because light of appropriate wavelength may trigger photooxidation reactions (55,56). Furthermore, there is an excellent oxygen supply to the photoreceptors through the choroid and retinal vessels causing a high oxygen tension in the retinal pigment epithelium (RPE) and photoreceptors (49).
Since high PUFA and oxygen are often combined with abundant light in the retina, Noell et al. first suggested a role for lipid peroxidation in photochemical retinal damage in 1966 (57). Noell later showed that light damage to the rat retina is enhanced when body and eye temperature are elevated and that the extent of damage is also related to history of light intensity of the animal before acute light exposure (58). Noell's work initiated many studies investigating the role ofantioxidants in retinal degeneration, in which vitamin E-deficient and -supplemented animals were most often used. These studies were reviewed by Handelman and Dratz (29), and they showed microscopic degeneration of the photoreceptors after exposure to light when animals were reared in the dark or in low-level cyclic light. The morphological changes after light exposure were usually enhanced in vitamin E-deficient animals as compared to vitamin Esupplemented animals (29). Biochemical changes investigated showed that antioxidant-deficient rats accumulated fluorescent pigment thought to be characteristic of lipid peroxidation in the retinal pigment epithelium (59), the conjugated diene content in the retina doubled (60), and the polyunsaturated fatty acids in the retina decreased (61). These studies all suggested a role for lipid peroxidation in retinal degeneration. However, to prove that lipid peroxidation products are formed in retinas of vitamin Edeficient rats, gas chromatography-mass spectrometry (GC-MS) assays were developed and employed (9)(10)(11)(12). Using these specific and sensitive methods, it was possible to show the presence ofoxidized fatty acids and 4-hydroxyalkenals in rat retinas (13)(14)(15). The levels of these lipid peroxidation products were higher in vitamin E-deficient as compared to vitamin E-supplemented retinas, giving more support for the involvement of lipid peroxidation in retinal degeneration (13,14), although firm evidence is not yet available on the effect of light.
Evidence that UV light exposure has a variety of adverse effects on the retina is accumulating (49). Different sources of damage from excessive light are possible, including sunlight (62,63), arcs of electrical current (64), lasers (65,66), and operating microscopes (67,68). It has been proposed that lipid peroxidation is an important factor in solar radiation contributing to age-related macular degeneration (ARMD) (62). ARMD deserves attention because it is the major cause ofblindness in the United States and Western Europe in the elderly and is likely to become a greater public health problem as the number of persons in older age groups continues to increase (69). ARMD is characterized by histopathological features of pigmentary disturbances, drusen, thickening of Bruch's membrane, and basal laminar deposits (70). It is apparently influenced by a large number of variables such as nutrition, age, family history, iris pigmentation, and solar radiation (71).
The role ofsolar radiation in ARMD was recently reviewed by Young (62), and the major concepts ofthe radiation hypothesis proposed by Young can be summarized as follows. Damaging effects ofradiation are dependent on the wavelength. Because the energy content is highest at lower wavelengths, this higher energy light is more damaging to various eye tissues (72,73). A critical point appears to be the wavelength of 510 nm. Light below this wavelength has a photon energy of > 2.43 eV, and a sudden rise occurs in the amount ofdamage produced in the retina as a function of exposure to light of different wavelengths (62,72,73). Therefore, the blue portion of the visible region (400-500 nm) is much more hazardous than the green and red portion (500-700 nm). This phenomenon is referred to as blue light hazard (62).
The lens transmits light oflower wavelengths to a different extent at various ages (74)(75)(76)(77). Lerman reported that after birth, nearly all ofthe UV light is transmitted by the lens (77). During childhood, lens transmittance decreases, and by the age of25, the lens absorbs UV light almost completely, as shown in Figure IA. It is interesting to correlate this result with the data on lipofuscin accumulation in the retinal pigment epithelium as a function of age shown in Figure IB. Wing et al. demonstrated that there is a steady increase of lipofuscin content during the first three decades, after which it does not change significantly until the eighth decade (Fig. I) (78). Comparing these data provoke the hypothesis that the accumulation of lipofuscin in the RPE at younger ages is associated with exposure of the retina to UV radiation transmitted by those lenses. As lipofuscin is thought to be an end product of lipid peroxidation, this mechanism could have an important contribution to biochemical changes mediated by UV light. These data are consistent with the hypothesis by Weiter et al. that cumulative exposure to UV light before the age of 25 is a risk factor in ARMD (79).
Several epidemiological studies have recently been performed studying the role of light in development of retinal disease. Liu et al. showed an association between occurrence ofARMD and lens opacities in the aged (80). They found that the odds for having ARMD may be increased by 50% among persons with cataracts, 80% in persons with opacities that do not decrease visual acuity, and 200% among aphakic persons (80). They hypothesize that the weak correlation between cataracts and ARMD in their group of subjects could be due to the fact that cataracts decrease the transmission of light to the retina, providing some protection from light damage (80). On the other hand, the strong increase in occurrence of ARMD in aphakic persons, where the UV absorbing lens is not present, suggests that UV light exposure could be an important risk factor for ARMD (80). West et al. found no association between ARMD and cumulative exposure to either UV-A or UV-B light (81). This study was conducted on 838 Maryland watermen from whom cumulative exposure to UV-A and UV-B light after age 15 was well documented. A weakness ofthis study is perhaps that the cumulative exposure to UV light before age 15 was not established; however, it could be practically impossible to obtain such data. As indicated above and in Figure IA, the human retina may receive most of its cumulative UV exposure during the first 15 years of life (77). The results ofthe two studies evaluating light as a risk factor for ARMD described above do not agree, and further studies are warranted to resolve the question of whether cumulative UV light exposure during life is a risk factor for ARMD. Recently, Munoz et al. reanalyzed the data obtained from the study (81) on 838 Maryland watermen (82) to determine whether ocular exposure to blue or visible light is associated with ARMD. They found that established cases of ARMD had significantly higher exposure to both blue and visible light over the last 20 years and concluded that high levels of exposure to blue and visible light late in life may be an important factor in ARMD (82). Interestingly, these results differ from those by Weiter (83)(84)(85). These include goggles with narrow slits as used by the Eskimos; hats with large brims like the stetson, the pith helmet, and the sombrero; and protective glasses. Rosenthal et al. performed an interesting study on light blocking effects using mannilin headforms containing UV-B sensitive film, fitted with brimmed baseball caps (84). They showed a 22 to 95 % reduction in ocular exposure to light depending on the angle of the hat brim on the forehead. In addition, they found that the amount of UV-B exposure was independent of the cloud cover (84). This study demonstrates that a habit ofwearing a hat with a brim outdoors greatly reduces the amount ofsunlight exposure to the eye.
Protective glasses are commonly recommended to protect the eye from harmful effects ofUV radiation (1,3). However, several studies showed the occurrence of UV windows in commercial sunglasses and commonly used clip-on sunglasses (85)(86)(87)(88)(89)(90). UV transmitting windows are not desirable because tinted sunglasses may cause pupillary dilatation, which would increase the exposure ofthe retina to UV light when the sunglasses used have a UV window (88). On the other hand, it has been proposed that visible light energy in a sunny outdoor environment penetrates dark tinted glasses sufficiently to produce significant pupillary constriction (87). It is also reported that darker lenses do not assure sufficientblocking of low wavelength light (88). Protection ofthe retina against UV light exposure is a concept that has also received attention from ophthalmic surgeons. Intraocular lenses implanted during surgery for cataracts now have chromatophores incorporated that block UV light. This will help to counteract a sudden increase of exposure of the retina to UV light, which was blocked by the cataract before surgery.
Although several publications in professional journals have shown problems with UV transmission of sunglasses (86)(87)(88)90) and standards have been proposed (88,91), these problems are not yet eliminated. My recent experiences are illustrated in Figure 2, which shows transmission curves of several different glasses. Curve I is the transmission ofa clear lens from corrective prescription glasses which are used continuously throughout the day. A UV filter was requested for these lenses when purchased, and it was said by the sales person that UV filters were present in any lens sold inthat store. When the new glasses were picked up several weeks later, it was noticed that no yellow tint was present, which is often seen when a UV coat is applied. The sales manager s asked ifa UV coat was present, and assurance was given that the lenses (purchased for $70 each) had a new type of transparent UV coating. The glasses were purchased and several months later a transmission curve was obtained with a Perkin Elmer 576 UV-Visible spectrophotometer. It was found that the UV-B light is blocked, but that a substantial fraction of UV-A light is transmitted (Fig. 2, curveI). This was reason to approach several local "eye care" stores, and various samples of clear lenses were borrowedthat were said to have a UV absorbing coat. There was a substantial difference in UV lighttransmittance properties as reported before for sunglasses (86)(87)(88)90) and tinted lenses (89). Some lenses blocked all of the UV light (< 400 nm), while others showed transmittance similar to curve I, Figure 2.
Curve II in Figure 2 is an average transmittance curve of several lenses coated with blue blocking filter, and it shows that some ofthe UV-A light is notblocked and that 10 to 20% ofthe blue light is still transmitted. The transmittance is equal to or higher than the 10% blue light transmittance recommended by Silver (88).
The third curve in Figure 2 represents the transmission of prescription sunglasses that were recently purchased. A UV filter was added to a dark colored lens, and it was agree( transmission curve could be obtained before payment (I lens). The results are shown in Figure 2, curve HI, and d strate that essentially all of the UV light is blocked and t transmission of blue light was reduced to 1 to 3 %. In ad transmission was highest at the long wavelength of the region at about 700 nm and decreased in the infrared i Because infared radiation may enhance thermal light dar the retina, reduction of infrared radiation is also desirabl These UV protective glasses serve well for driving and ot sports activities.
The question remains what should be recommended public as far as reduction of ocular exposure to hazardoi is concerned. Should education start with children? Receti U.S. Council on Scientific Affairs recommended that cl should be taught to use sunscreen protection early in life b skin damage accumulates over an entire lifetime (1), recommendations were made about the use ofglasses or protect the eyes when outdoors. To develop recommendati children, several factors should be taken into consideral The ocular lens ofchildren transmits UV light (77), an UV light damage to the eye may occur in the earlier years Weiter et al. showed that the incidence of ARMD is Ic myopic subjects, who had been forced to use glasses at a age, as compared to hyperopic subjects who did not star glasses early in life (79). Since the glasses filter a part ofI light, Weiteret al. suggested an association between devel( ofARMD and UV light exposure in the first 25 years ofli results ofthis study favor a recommendation to teach child use of protective glasses or hats. On the other hand, Ler al. reported that ambient low levels of UV light ass development of UV-absorbing chromophores in the you] (36), which contribute to making the lens an efficient U' by age 20 to 30 (Fig. IA) (77). This study suggests that soi bient UV exposure to eyes of children is desirable. F evaluation of these factors is needed to establish an appr4 recommendation for protection of children's eyes agai cessive UV-light exposure.

100%
Second, it is desirable that consensus be obtained among professionals on what wavelengths oflight UV protective sunglasses should block/transmit. Unfortunately, there is considerable 80 % disagreement on this subject (46,47,62,83,92). Recommendations to date vary from the use of ordinary glasses with plastic lenses to reduce UV-B exposure (46,47) to the use ofglasses that 60X block UV and blue light (62,83,92). In addition, there are differences in opinion about the transmittance characteristics of Tr clear lenses and their usefulness to protect against exposure to 40X UV light (85,(93)(94)(95)(96)(97). This is surprising because several studies have shown that there are great differences in the transmittance characteristics of tinted lenses, clip-on sunglasses, and regular 20% sunglasses (86)(87)(88)(89)(90). These different opinions lead to different recommendations, which have caused confusion among scientific colleagues (98) and must be even more confusing to the public. o MThe differences in opinions about transmittance characteristics of plastic lenses are in part due to the fact that the wavelength 1) Trnsrange of UV-B light is defined differently by different inittanceof vestigators. Table 1 summarizes which wavelength ranges are attibuted to UV-A, UV-B, and UV-C in studies cited in this paper.
The definitions vary substantially, and the ranges used in this d that a paper are those most commonly used in other studies. Surpris-l65 per ingly, all these wavelength ranges differ from those as defined by A plastic lens can therefore block nearly all the UV-B light nge to when 310 nm is defined as the cutoff wavelength ofUV-B (85,89), le (83). but the same lens can transmit substantial UV-B light when 320 utdoors nm is defined as the cutoff (88). Agreement on the wavelength range ofUV-B light is desirable to establish more valid and con-1 to the sistent transmittance characteristics of lenses. us light Upon reviewing the recent literature, it can be concluded that idy, the recent epidemiological studies showed an association between hildren cumulative UV-B light exposure and the development ofnuclear xecause cataract (46) and posterior subcapsular cataract (47). Scientific but no evidence is also accumulating that there is an association between hats to solar radiation and ARMD, also referred to as the blue light ions for hazard (62). Epidemiologic surveys performed to date, studying tion.  (81). Further murk is needed to evaluatetheeffects ofchronicandcumulativeUV andbluelightexposuretotheretina. Sincethis issue is notyet resolved, a conservative recommendation for UV protective glasses wouldbe to use lenses that have at least a UV filter and blue blocker (e.g., curve El[, Fig. 2), as this may prevent lens and retinal disease. This has been suggested before (62,83) and is also in agreement with the standards suggested by Silver, who noted that it has never been suggested that intermittent wearing ofany lens can actually cause damage to the eye (88). Furthermore, recent reports on retinal damage caused by excessivevisible light (51,82) indicatethatreductionoftransmittance ofvisible light should be considered as well. A general reduction oftransmittance oflight ofall wavelengths including UV, blue and visible (e.g., curve mU, Fig. 2) is a recommendation that could be considered after further research on this topic.
Rosenthal et al. showed that the amount of attenuation of ocular exposure to UV mdiation by sunglasses is highly variable and is affected by size, shape, and wearing position (85). Placing the sunglasses a small distance from the forehead can triple the ocular exposure (85). This is important information that can help to improve education on the proper use of protective sunglasses.