A pair of snow goggles carved from walrus ivory was found in an early Inuit house on Baffin Island. It was made sometime between the years 1000 and 1600, and it now belongs to the Canadian Museum of History, which holds more than 100 pairs. For a teaching kit, the museum asked Ross Kayotak, a carver from Igloolik, to make new ones. He had no caribou antler at the time, so he carved them from the rib of a bowhead whale. "Hunters were the main people using them," he said, "those that would walk distances or wait for seals to come up, when it was bright outside." 1
Much of the glare those hunters shaded their eyes from came up off the ground. Fresh snow can reflect as much as 80 percent of the ultraviolet that falls on it, and dry beach sand about 15 percent. Without goggles, the museum notes, people risk snow blindness, "a painful condition likened to a sunburn on the eyes." Around 400 BC the Greek writer Xenophon described soldiers falling behind on a march through snow and cold, "those whose eyes had been blinded by the snow," and noted that a man was protected "if he marched with something black in front of them." 2; 1; 3; 4
Besides the skin, the eyes are the only major organs that sunlight reaches directly. Looking straight at the sun can burn a small blind spot into the retina, the layer of light-sensing cells at the back of the eye. A day on bright snow can burn the cornea, the clear window at the front, which replaces its outer cells in about two days. Among 838 watermen who worked on Chesapeake Bay, those whose estimated ultraviolet dose to the eye was higher had more cortical cataract, one of the three common kinds of clouding in the lens. In a three-year trial in Guangzhou, 30.4 percent of 853 children given an extra 40-minute class outdoors every school day became nearsighted, against 39.5 percent of 726 children at schools that kept their usual routine. 5; 6; 7; 8
Our catalog of the questions people type into search engines holds 470 about sunlight and the eyes. Of the 784 studies our library keeps for deep analysis, 27 are tagged as eye studies, and 13 of those study the eye. None of the 27 measured how much sunlight reached an eye. The Chesapeake study, the Guangzhou trial, and most other large studies on this page sit in our evidence map but did not pass our rule for the library, so this page cites them from their journals.
This page covers snow blindness, sun damage to the retina, the eye's dose, cataract, pterygium, macular degeneration, nearsightedness in children, sunglasses, eye color, and the 27 library studies. Each study is named as real sun, a lamp, a product put on animals or cells, or an estimate built from how people lived.
Key facts
- Among 838 Chesapeake Bay watermen, doubling the estimated lifetime dose of ultraviolet B to the eye went with 1.6 times the risk of cortical cataract; no link turned up with nuclear cataract or with ultraviolet A. 7
- In a three-year trial at 12 schools in Guangzhou, 30.4 percent of children given an extra 40-minute outdoor class each school day became nearsighted, against 39.5 percent at schools without it. 8
- Of 29 eyes burned by the sun in a London series after the 1966 eclipse, 15 ended at 6/9 or worse, and 11 of those had looked through sunglasses, smoked or stained glass, filters, or old negatives. 9
- The World Health Organization estimates that 15 million people are blind from cataract and that some 10 percent of those cases may be due to ultraviolet. 10
- Of the 27 eye-tagged studies in our library, 13 study the eye, and none measured the sunlight that reached one.
Snow blindness burns the cornea and heals in about two days
Snow blindness, which welders call arc eye, is the common name for photokeratitis, an inflammation of the cornea caused by ultraviolet, mostly by UVB, the shorter-wave band that also causes sunburn. It starts after a delay of half an hour to a day, usually 6 to 12 hours, with the feeling of sand in the eyes, tears, and lids that squeeze shut. The acute symptoms last 6 to 24 hours and the discomfort usually passes within 48, about the time the cornea takes to replace its outer layers. It rarely leaves lasting damage, according to the International Commission on Non-Ionizing Radiation Protection, which sets exposure limits for ultraviolet. 6; 11
Our library holds no study of snow blindness in people, only animals and cells under lamps. In 1990 a team at the Doheny Eye Institute in Los Angeles set white rabbits, three at a time, 15 centimeters from an 80-watt fluorescent UVB lamp. Before the lamp, each rabbit got a drop of octinoxate, a sunscreen chemical, in the right eye and a drop of castor oil in the left. Fifteen minutes reliably inflamed the cornea of the castor-oil eye. The treated eyes came through 30 and 60 minutes without visible inflammation, which the authors counted as "a sun protection factor of at least 4." 11
A myopia team in Tokyo later met the same corneal burn in chicks. Under a UVB lamp at 305 nanometers, a measure of wavelength, the birds "kept both eyes closed all day because the UVB exposure caused severe corneal keratitis." 12
At the University of California, Davis, in 2006, human corneal cells grown in dishes took 50 millijoules per square centimeter of UVB from a lamp, a dose in the range the authors equated to about 20 minutes of sunlight on the cornea, and 20 hours later about 20 percent were alive. Cells first treated with arsenite, a compound of arsenic that switches on protective stress proteins, survived better. In 2013 a team shone lamps on donated human eyes and found that damage from UVB and the even shorter UVC stayed in the outer skin of the cornea, while UVA, the longer-wave band, reached through the whole cornea into the first layers of the iris. The authors noted that "no UV-related neoplasm has been reported in this ocular structure." 5; 13
Staring at the sun burns the retina with light, not heat
On the morning of 20 May 1966 a partial eclipse of the sun lasted two hours in London, from 9:20 to 11:20, in weather that "permitted an excellent view." At Moorfields Eye Hospital, P. A. MacFaul saw 18 people with retinal burns from it, and one more from sunbathing. They were 13 males and 6 females, all aged 9 to 24 except one woman of 45, and 11 of the 18 eclipse cases were schoolchildren. Eight had looked straight at the sun, and the other eleven used sunglasses, stained glass, smoked glass, "filters", old photographic negatives, a pinhole, or partly closed lids, and "one relied on transient cirrus cloud formation to reduce the light intensity." The series counted only people who reached the hospital, and MacFaul wrote that "there must be many mild cases of solar retinopathy which clear up rapidly with no lasting ill-effects." 9
The burns sat at the macula, the small patch at the center of the retina that is used for reading and for faces, an injury doctors call solar retinopathy. In Britain sight is written as a fraction: 6/6 is normal, and 6/9 means a person sees at 6 meters what a normal eye sees at 9. An 18-year-old watched the eclipse three days before his national school exams, fell to 6/9, and found that the blind spots at the center of his vision "interfered considerably with his ability to write his examination papers." He recovered fully in four weeks. 9
After 6 to 12 months of follow-up, 15 of the 29 burned eyes were at 6/9 or worse, and 11 of those 15 had used one of the "so-called protective devices." A 22-year-old in a psychiatric hospital who "watched the eclipse constantly throughout its duration" never improved from 6/24 and was placed on the partially sighted register. A 45-year-old barrister "whose hobby was astronomy" watched through smoked glass, and after two weeks of steroids her sight returned to 6/5 in each eye, slightly better than normal, with a small blind spot still in each. The 1966 eclipse was the second to injure the eyes of an 18-year-old carpenter; four years earlier, the first had caused only a passing disturbance. 9
MacFaul explained the burn as heat, with visible and infrared light focused on the retina and absorbed by its pigment layer "with the liberation of heat." In a 2023 safety report for the American Astronomical Society, the optometrist B. Ralph Chou writes that the sun's image warms the retina of an unaided eye by less than 4 degrees Celsius, and that the main harm is photochemical, a chain of reactions set off in the light-sensing cells mostly by blue light. The retina has no pain receptors, and the loss may not show for several hours. Even with 99 percent of the sun covered, the crescent left over can injure the retina while daylight looks like twilight. 9; 14
Safety limits allow about 0.6 seconds of staring at an overhead sun and 2.5 seconds at a sun 10 degrees above the horizon, with a large margin built in. MacFaul concluded that "the only way to avoid solar retinopathy is to refrain from looking directly at the sun." For eclipse watchers he described two pieces of stiff cardboard, one with a small hole in it, held with the back to the sun so that the image falls on the other. 14; 9
Sun-gazing rituals caused lasting loss of vision; most eclipse burns in recent series healed
A 1973 report of sun-gazing under the influence of LSD listed earlier cases among anti-aircraft lookouts, "military recruits hoping to obtain discharge from service," people watching the sun "as a religious ritual," and sunbathers. It passed on a suggestion that the temporary blindness of Saul of Tarsus "may have been the result of solar retinopathy." In 1988 three ophthalmologists reported four people who had gazed at the sun during religious rituals: "All four patients suffered irreversible visual loss." 15; 16
Eye doctors in the hills of Himachal Pradesh in India, 1,300 to 1,500 meters up, described three people in 2021 whose retinal scans showed sun damage at the center. None admitted staring at the sun. A 42-year-old farmer had sunbathed for long hours over three days, "not a typical habit," which "he did to combat the severe cold." A 40-year-old farm supervisor habitually sunbathed for long hours "lying supine with closed eyes directly under the sun." The authors tied the damage to sunbathing through the patients' histories and by ruling out other causes. 17
After a partial eclipse over Hawaii in 1962, 27 of 52 cases examined regained normal acuity. After the eclipse of August 1999, ophthalmologists across the United Kingdom reported 70 people with lost vision to a three-month national survey. When first seen, 84 percent had a visible change at the macula; at six months none reported continued loss or symptoms. 9; 18
After the eclipse of April 2024 over the southern and eastern United States, a count of emergency department visits for eye injuries found 853 before it and 921 after, a difference within chance. In a US database of 1,042,995 eye patients from 2012 to 2024, 555 were newly diagnosed with solar retinopathy; eclipse years averaged 51.5 new cases and other years 37.3, again a difference within chance. The median patient was 54, and 32 percent had an anxiety disorder and 10 percent a psychotic disorder. 19; 20
At sunset the disk sitting on the horizon is safe to look at, Chou writes, "though not necessarily comfortable," because nearly all its blue light has been scattered out of the image. 14
The eye's dose depends on the horizon more than the overhead sun
The UV Index on a forecast measures the ultraviolet falling on a flat, level surface, the way it falls on a beach towel. An eye faces forward under a brow, and the upper lid and the brow ridge block most light from overhead. The commission that sets ultraviolet limits wrote in 2004 that the index "is not very predictive of ocular exposure." 6
In 1999 D. H. Sliney, working out the geometry of the eye's exposure, wrote that the opening of the lids limits the eye to rays arriving near the horizon, and that clouds scatter overhead ultraviolet toward the horizon sky. As a result, "the highest UVR exposure occurs during light overcast where the horizon is visible and ground surface reflection is high." In a high mountain valley with green foliage, the eye's dose is much lower. 21
In Kanazawa, Japan, in 2011, a team mounted two mannequin heads with built-in ultraviolet sensors on a motorized stand that tracked the sun, one face toward it and one away. In September the sun-facing eye's UVB dose rose and fell twice during the day, a pattern unlike the ultraviolet in the open air and unlike a sensor on top of the head. The authors called the UV Index "invalid" as a guide to eye risk and wrote that eye protection "may be warranted throughout the year." 22
The 27 eye studies in our library contain one number for how much ultraviolet reaches the eye, and it is a citation. The 2006 Davis corneal cell paper states, citing a 1988 source, that "it is known that 2–17% of solar UV can reach the eyes." 5
Chesapeake watermen with higher estimated ultraviolet doses to the eye had more cortical cataract
A cataract is a clouding of the lens, the clear disk behind the pupil that focuses light onto the retina. Cortical cataract forms in the lens's outer layers, nuclear cataract in its center, and posterior subcapsular cataract under its back surface. In an intact adult eye the lens absorbs nearly all the ultraviolet A that gets past the cornea. 6; 14
In the 1980s Taylor's group examined 838 watermen who worked on Chesapeake Bay in Maryland, with a mean age of 53. For each man they calculated a yearly ultraviolet dose to the eye from age 16, combining a detailed history of his working life with measurements of sun exposure in the laboratory and in the field. Some cortical cataract was found in 111 of them (13 percent) and nuclear cataract in 229 (27 percent). 7
Doubling a man's cumulative dose of ultraviolet B went with 1.6 times the risk of cortical cataract, and the true effect could be anywhere from barely any increase to about two and a half times. The men in the top quarter of yearly dose had 3.3 times the risk of the bottom quarter, and the true effect could be anywhere from slightly lower risk to about ten times as much. No link turned up with nuclear cataract, or with ultraviolet A. 7
In Salisbury, Maryland, from 1993 to 1995, a group led by West, who had worked on the Chesapeake studies, examined 2,520 townspeople aged 65 to 84, 26.4 percent of them African American, and gave each an estimated lifetime ultraviolet B dose to the eye in a unit they called "Maryland sun-years." The odds of cortical clouding rose modestly with the dose, an odds ratio of 1.10, which means 10 percent higher odds, and the true effect could be anywhere from 2 to 20 percent higher. The link held for women and for African Americans, in whom, the authors wrote, it had not been reported before. They found "no vulnerable age group" after age 30, which they took to mean that the damage builds up over a lifetime. 23
Among 43- to 84-year-olds in Beaver Dam, Wisconsin, men with higher average yearly ambient ultraviolet B were 1.36 times as likely to have more severe cortical clouding. Women showed no link, although they had more cortical cataract, and the authors wrote that "other factors may be more important." 24
In the Blue Mountains west of Sydney, graders split photographs of 3,270 lenses into eight sectors and estimated the clouding in each. The lower half nearest the nose held four times the cortical clouding of the upper half nearest the temple, at every age and in both sexes. Earlier reports had tied that pattern to sunlight entering past the shallow brow at the temple side, and the authors wrote that it "may indicate a role for sunlight." 25
In a survey of 4,744 adults in the Australian state of Victoria, ultraviolet B exposure of the eye explained 10 percent of cortical cataract. The World Health Organization's 2022 fact sheet puts the number blind from cataract at 15 million, of whom "some 10% may be due to exposure to UVR." Its 2002 guide had said that "up to 20%" of that blindness may be "caused or enhanced by sun exposure," especially in the countries of a "cataract belt" near the equator. 26; 10; 2
In the one lens study in our library, a Manchester group in 2015 put purified lens proteins from cow eyes under a solar simulator, a lamp built to imitate sunlight, at up to 15.4 joules per square centimeter. Beta- and gamma-crystallin clumped and broke apart, and alpha-crystallin changed little. The study was funded in part by Alliance Boots, which sells sunscreen, and the authors disclosed that the company "was involved in the decision to submit this work for publication." 27
Pterygium rose with the estimated sun dose to the eye
A pterygium is a wedge of fleshy tissue that grows from the white of the eye onto the cornea. Among the 838 Chesapeake watermen, 140 had one, and 642 had a pinguecula, a small yellowish thickening of the membrane over the white of the eye. Pterygium rose with estimated exposure across ultraviolet B and A, and the link with pinguecula was weaker. The authors wrote that "simple measures such as wearing a hat or spectacles protect the eye." 28
In Perth in 1999, Threlfall and English compared people who had a pterygium removed with patients having ear, nose, or throat surgery. For each person they estimated a daily dose of sunlight to the eye from climate records, the hours spent outdoors out of shade, and the use of hats and spectacles, over a lifetime of homes. People in the top quarter of that dose had 6.8 times the odds of pterygium of the bottom quarter, and the true figure could be anywhere from about 2.6 to nearly 20 times. Exposure in no single period of life mattered more than in the others. 29
In Sydney in 2003, the surgeon Minas Coroneo cut seven freshly removed pterygia "symmetrically into halves" and sent them to the laboratory on ice. Nick Di Girolamo's team gave one half of each 20 millijoules per square centimeter of UVB from a lamp and left the other half in ordinary room light in a sterile cabinet for the same time. The lamp-exposed halves released 2.7 times as much MMP-1, an enzyme that cuts collagen. Cells grown from normal conjunctiva barely responded to the same lamp dose. 30
Ultraviolet damage on the eye's surface can be photographed as a glow under ultraviolet light, called conjunctival ultraviolet autofluorescence. In a group of young adults followed from about age 20 for eight years, the glowing area shrank on average, and it shrank faster in those who wore sunglasses at least half the time. Fourteen of them, 1.5 percent, developed a pterygium. 31
Studies of sunlight and macular degeneration disagree
Age-related macular degeneration destroys the center of the retina in older people. The adult retina behind a natural lens receives almost no ultraviolet: less than 1 percent of UVA reaches it, and UVB is absorbed completely except in newborns. Young eyes let through more light. A 2005 dermatology review describes a window near 320 nanometers in the newborn eye that lets through about 10 percent and closes around age 10, and blue light passing at 70 to 80 percent until about age 20 and at 40 percent by age 60. 6; 32
The Chesapeake watermen showed no link between macular degeneration and ultraviolet A or B, "even with high levels of sunlight exposure." When Taylor's team estimated the men's exposure to visible and blue light, those with advanced macular degeneration had received more blue light over the previous 20 years than men of the same age without it. The odds ratio was 1.36, and the true effect could be anywhere from no increase to 1.85 times the odds. 33; 34
At Beaver Dam, in men only, more summer leisure time outdoors went with 2.26 times the odds of the wet form of the disease, exudative macular degeneration, and 2.19 times the odds of late disease. Men who wore hats and sunglasses had fewer soft drusen, yellow deposits under the retina that are an early sign of the disease. Estimated ultraviolet B showed no link, and women showed none at all. Ten years on, people who had spent more than five hours a day in summer sun as teenagers, in their 30s, and at the start of the study were 2.2 times as likely to develop early disease as those who had spent less than two, and the authors' conclusion began, "Few significant relationships." 35; 36
In the Blue Mountains a 1998 survey found blue irises more often in people with late macular degeneration, an odds ratio of 1.69. When the same group was followed for ten years, iris color predicted nothing, and the authors "did not find a consistent pattern of association between sunlight-related factors" and the disease. Two meta-analyses, which pool earlier studies into one estimate, took in 14 studies each and reached different answers. One in 2013 found an odds ratio of 1.38 for more sunlight, and one in 2018, covering 43,934 people, found 1.10, where the true effect could be anywhere from slightly lower odds to 1.23 times. In the second, people who avoided the sun had no lower risk. 37; 38; 39; 40
In the one retina study in our library, a chemistry group at the University of Toledo in 2020 used no retina. It soaked HeLa cells, a line of human cancer cells, in retinal, the light-catching molecule of the eye, and took one dish into direct sunlight: "Al-foil of one plate was removed and exposed to sunlight for 30 minutes while keeping the lid closed." About 28 percent of the cells died at 1 micromolar retinal and 37 percent at 10, while cells without added retinal showed no disruption in the same sunlight. In a living eye, light reaches the retina through the cornea, the lens, and the macula's own pigment. 41
Children given more time outdoors became nearsighted less often
In myopia, or nearsightedness, the eyeball usually grows too long for its focus, and distant things blur. In 2000 about 1.4 billion people, 22.9 percent of the world, were nearsighted, and a 2016 projection put the share at 49.8 percent by 2050. In East Asian cities 80 to 90 percent of young people leaving school are nearsighted. 42; 43
In 2008 Rose's group in Sydney compared children of Chinese ethnicity aged 6 and 7 in two cities, 124 in Sydney and 628 in Singapore. In Sydney 3.3 percent were nearsighted and in Singapore 29.1 percent, although about the same share had a nearsighted parent, 68 and 71 percent. The Sydney children read more books and did more close work. They also spent 13.75 hours a week outdoors, against 3.05 in Singapore. 44
In October 2010 a team in Guangzhou began a cluster randomized trial, a design that assigns whole groups by chance, in 12 primary schools with 1,903 first-graders averaging 6.6 years old. Six added one 40-minute class of outdoor activity to every school day and encouraged parents to take their children out on weekends and holidays, and six carried on as usual. After three years, 30.4 percent of the 853 children at outdoor schools who had not been nearsighted at the start had become so, against 39.5 percent of 726 at the usual schools. The difference, 9.1 points, could plausibly be anywhere from about 4 to 14. Their eyeballs grew by about the same amount in both groups, 0.95 millimeters against 0.98. 8
In September 2010 a national program called Tian-Tian 120 began encouraging Taiwan's schools to take students outdoors for 120 minutes every day. Schools test pupils' unaided eyesight with a Tumbling E chart twice a year, 1.2 to 1.9 million primary school children a year. The share with reduced eyesight had risen from 34.8 percent in 2001 to 50.0 percent in 2011, and then it fell each year, to 46.1 percent in 2015. That measure was eyesight without glasses, not a direct test of nearsightedness. 45
Shanghai's trial of 6,295 children in 24 schools added 40 or 80 minutes of outdoor time a day and measured it in the second year with a device worn on each child's wrist. Both groups ended up outdoors about 127 minutes a day, against 106 at the usual schools, and new nearsightedness fell by 16 and 11 percent. The authors put the missing difference between the two doses down to poor compliance. A 2017 pooling of the trials found that outdoor time cut the risk of becoming nearsighted to about half, but did not slow its progress in eyes that were already nearsighted. 46; 47
Bright light, not vitamin D, went with the protection
Chicks can be made nearsighted in the laboratory by fitting them with translucent diffusers. In 2009 Ashby, Ohlendorf, and Schaeffel took the diffusers off for 15 minutes a day under one of three lights: ordinary laboratory lighting of 500 lux, bright laboratory lamps of 15,000 lux, or daylight at 30,000 lux. Lux measures brightness, and diopters measure focusing error. After five days, the chicks given daylight had −1.1 diopters of nearsightedness, against −5.3 under ordinary lab light and −3.4 under the bright lamps. Chicks that kept their diffusers on but spent six hours a day under 15,000 lux lamps developed essentially none. 48
The 2018 Taiwan trial, in 16 schools with 693 first-graders, put light meters on the children and encouraged those at the outdoor schools to spend up to 11 hours a week outside. After a year, their eyes had shifted less toward nearsightedness (0.35 diopters against 0.47) and their odds of rapid progression were 54 percent lower, in children already nearsighted as well as the rest. Children with 200 minutes or more of outdoor time at school, at 1,000 lux or above, showed less shift. The authors wrote that "outdoor activities with strong sunlight exposure may not be necessary," and that time in lower light, "such as in hallways or under trees, also can be considered." 49
In the Avon Longitudinal Study of Parents and Children, a birth cohort in England, 3,677 children whose mothers reported more time outdoors at age 8 had slightly more vitamin D in their blood at 10, 60.0 against 56.9 nanomoles per liter. With both in the analysis, vitamin D had no independent link to later myopia, while time outdoors kept its full effect, a hazard ratio of 0.69, or about 31 percent less risk of becoming nearsighted over time. A 2017 study of more than 45,000 adults used gene variants that set a person's vitamin D level and put its effect on focusing error at 0.02 diopters per 10 nanomoles per liter, which the authors called "indistinguishable from zero." 50; 51
A group at Keio University in Tokyo proposed a narrower active ingredient, violet light, the band from 360 to 400 nanometers at the edge of ultraviolet. In lamp-lit chicks with a goggle over one eye, adding violet light cut the covered eye's myopia from −15.18 to −4.59 diopters; goggles that fell off were put back, and those eyes were dropped from the analysis. The team then looked back at clinic records of Japanese students aged 10 to 15. Eyes of 211 who wore ordinary glasses, which block violet, grew 0.25 millimeters in a year, against 0.17 in 99 who wore contact lenses that pass it. 12
The two groups differed in age, starting focus, and starting eye length, and the authors noted that they had compared "different materials as well as different spectral transmissions." The paper reports no measure of the children's time outdoors. Four of the authors were applying for a patent "for potential products for myopia suppression." 12
Price and tint did not show which sunglasses blocked ultraviolet
The US federal rule for sunglasses, 21 CFR 801.410, requires lenses that resist impact. In its test, a 5/8-inch steel ball weighing about 0.56 ounces is dropped from 50 inches onto the lens. The rule says nothing about ultraviolet. Australia will not let sunglasses be sold unless they pass a mandatory safety standard, and it sorts lenses into five categories: categories 0 and 1 "are fashion spectacles, not sunglasses," and category 4 lenses "must not be used when driving at any time." 52; 53
Under the international sunglass standard, ISO 12312-1, the ultraviolet a lens may pass depends on its tint. A light category 1 lens, which lets through 43 to 80 percent of visible light, may let through as large a share of UVA and still comply. 54
In Tehran in 2015, researchers anonymously bought 348 pairs of sunglasses priced 20 to 80 US dollars, 189 from sellers licensed by the Ministry of Health and 159 from unlicensed sellers, and measured how much ultraviolet each lens let through. From licensed sellers, 92.6 percent met the Australian and New Zealand standard for ultraviolet and 95.2 percent the American one. From unlicensed sellers, none met the first and 8.2 percent the second. Among the licensed pairs, neither brand nor price predicted protection. The American Academy of Ophthalmology advises that a darker lens is not necessarily a more protective one and that "polarization itself does not provide UV protection." 55; 56
A common warning holds that dark lenses without ultraviolet filters are worse than none, because the pupil opens wider behind them. In 1987 H. L. Hoover called that supposition "wrong" after a survey of more than 400 retail lenses, all of which reduced the ultraviolet reaching the lens and retina. The 1990 rabbit eye-drop paper repeated the warning. In 2024 a group in Brazil modeled 214 unbranded lenses donated by the national association of sunglass makers and found that pupil size mattered little; what mattered was that people squint less behind dark glasses and so take in a wider field of view. Eight of the 214 lenses, all with less than 86 percent protection against UVA, let more ultraviolet into the eye than going without. 57; 11; 54
Johnson and Johnson Vision Care paid for a 2022 trial of a contact lens that filters about 60 percent of violet and blue light between 380 and 450 nanometers. Two of its four authors were company employees and the other two its consultants. Each of 61 volunteers wore the filtering lens in one eye and a nearly clear lens in the other and looked at a point of light from a 1,000-watt xenon lamp "emulating white sunlight." Halos around the light were 30 percent smaller and starbursts 23 percent smaller through the filtering lens. 58
Eye color settles in childhood; sunburns go with iris freckles
The Louisville Twin Study, published in 1997, followed 1,513 twins, matching each iris at every visit to one of 15 painted glass eyes, like those used in artificial eyes, mounted on a disk and running from light blue (1) to dark brown (15). Most people's eye color was stable by age six. Between six and adulthood, 3.8 to 8.6 percent of the white twins moved two or more steps lighter or darker over spans of three to nine years, and 9 percent of their mothers' irises lightened by two or more steps. Identical twins matched each other almost perfectly (a correlation of 0.98) and fraternal twins much less (0.49), so the authors suggested that the changes, or the tendency to them, may be inherited. 59
In a 2017 study at a skin cancer screening where ophthalmologists and dermatologists examined 632 volunteers, 76.1 percent had at least one iris freckle, most often in the lower part of the iris toward the temple. People with iris freckles were older, and more of them reported more than 10 sunburns in their lives (31 against 19 percent), had sun-damaged skin (26 against 11 percent), and had sun spots on the skin (72 against 45 percent). The authors concluded that sunlight "may trigger the formation of iris freckles." 60
The Grenoble clinic whose 303 volunteers appear on the sunburn page also sorted them by eye color. Under a xenon lamp, the 142 with blue, gray, or green eyes reddened at an average of 858 millijoules per square centimeter and the 161 with brown eyes at 1,101, although the authors found complexion a better guide than eye color. A 2004 Brazilian lamp study of 193 people found green eyes even in the darkest skin type, and its authors judged that eye color could not sort people into skin types. 61; 62
In the Blue Mountains, dark brown irises went with more nuclear cataract (an odds ratio of 1.59) and more posterior subcapsular cataract (2.50) than lighter irises. 63
Of 27 eye studies in our library, 13 study the eye and none measured its dose
Our evidence map screened 157,842 records, and 7,646 of them carry an eye or vision tag. Of those, 27 passed as eligible, 0.35 percent. The map admits a study when three AI models reading its record agree that it is eligible and agree on the kind of light exposure it measured. For eye studies the exposure is usually a job, a place of residence, or remembered hours outdoors, and the models often did not agree; for other records, eligibility could not be settled from the title and abstract. The Chesapeake, Beaver Dam, and Salisbury studies and the Guangzhou, Taiwan, and Shanghai trials are among the 5,125 eye records left unresolved.
Of the 27, 13 study the eye. Five are skin or bowel studies that recorded eye color as a sign of complexion, three used an eye dye but measured skin or cells, and six are keyword accidents, among them a sunscreen study tagged because redness "was assessed by eye." By exposure, 8 are real sun, 7 solar simulator, 7 lamp or device, and 5 proxies such as questionnaires. Four put real sun on a living human eye, and three of those are injury reports: the London eclipse series, the LSD report, and the Himachal Pradesh sunbathers.
The fourth is the only eligible study in which sunlight was the treatment. In a pilot in 1985, the authors treated 45 people with recurring inflammation at the front of the eye, anterior uveitis, with sun therapy at the Dead Sea, and 40 percent improved over the following year; the abstract describes no untreated comparison group. 64
Three people who went into the sun 3 to 18 hours after fluorescein angiography, a test in which a dye is injected into a vein to photograph the retina, described a tingling in the light and skin that felt "like a sunburn" afterward. Ronald Danis of Indiana University, one of the authors of the 1997 report, then tested the dye on himself. He put on a long-sleeved shirt with a 3-by-3-inch square cut out over his left shoulder blade, spread sunscreen with a sun protection factor (SPF) of 30 on the patch, and sat in the sun for an hour. Then he drank 6 milliliters of 25 percent fluorescein, sat in the shade for an hour, reversed the shirt so that the hole lay over his right shoulder blade, and sat in the sun again. 65
The first patch turned slightly red. The second reddened within about 20 minutes, burned and tingled, and took seven days to fade. The sunscreen did not stop it, the authors wrote, because "blue light at 487 nm will not be blocked by the lotion." In a follow-up challenge, 2 of 14 volunteers reacted. 65; 66
Misconceptions
What people believe
- Looking at an eclipse makes people go blind.
- The sun can change the color of your eyes.
- Darker or more expensive sunglasses protect more, and cheap dark lenses are worse than none.
- The midday sun under a clear sky gives the eyes their largest dose.
- Ultraviolet causes macular degeneration.
- Time outdoors protects children's eyes because of vitamin D.
- The sun causes 20 percent of cataracts.
What the studies show
- Of 70 people reported with eclipse-related vision loss across the UK in 1999, none reported continued loss at six months. In a London series from 1966, 15 of 29 burned eyes ended at 6/9 or worse. 18; 9
- Most of 1,513 Louisville twins had stable eye color after age six, and later changes tracked genes. Sunburns went with iris freckles. 59; 60
- In Tehran, brand and price did not predict the ultraviolet protection of sunglasses from licensed sellers. In a Brazilian model, 8 of 214 unbranded lenses let in more ultraviolet than none, through a wider field of view rather than a wider pupil. 55; 54
- The eye takes in light mainly from near the horizon, and its highest ultraviolet dose comes under light overcast with a visible horizon and reflective ground. 21
- The Chesapeake watermen showed no link between ultraviolet and macular degeneration and a weak link with blue light, and pooled analyses disagree. 33; 34; 40
- With time outdoors in the analysis, vitamin D had no independent link to myopia, and chicks given bright light developed less myopia. 50; 48
- The WHO's 2022 figure is "some 10%" of cataract blindness. Its 2002 "up to 20%" was a ceiling on blindness "caused or enhanced" by sun, especially near the equator. 10; 2
Questions people ask
Is sunlight good or bad for your eyes?
In children, more time outdoors went with fewer new cases of nearsightedness in trials in Guangzhou, Taiwan, and Shanghai. Staring at the sun can burn the retina, and higher estimated lifetime doses of ultraviolet went with more cortical cataract on Chesapeake Bay and more pterygium in Perth. 8; 49; 9; 7; 29
Can the sun change your eye color?
In the Louisville Twin Study most people's eye color was stable by age six, and the changes seen later tracked genes. Sun exposure has been tied to small iris freckles: people who had them more often reported more than 10 sunburns. 59; 60
Can bright sunlight cause floaters?
We found no study linking sunlight to new floaters. Floaters are clumps of gel or cells in the vitreous, the jelly that fills the eye, and people usually notice them against something plain, "like a blank wall or a blue sky." 67
Is sungazing good for your eyes?
Safety limits allow about 0.6 to 2.5 seconds of staring at the sun, and four people who gazed at the sun in religious rituals were left with irreversible loss of vision. A disk sitting on the horizon at sunset is the one safe exception, because its blue light has been scattered away. 14; 16
Do blue eyes make you more sensitive to sunlight?
No study in our library measured the eye's sensitivity to light by eye color. The eye-color studies we hold measured skin, and in Grenoble volunteers with blue, gray, or green eyes reddened under a lamp at lower doses than brown-eyed volunteers. 61
Does sunlight cause cataracts?
Among Chesapeake watermen and in Salisbury, more estimated ultraviolet B to the eye went with more cortical cataract. The WHO estimates that ultraviolet may account for some 10 percent of the 15 million people blind from cataract. 7; 23; 10
Does sunlight make glaucoma worse?
The one sun-related signal we found concerns exfoliation syndrome, a condition that can lead to glaucoma. Among 626,901 US eye patients, those living north of 42 degrees had about twice the hazard of those in the middle latitudes, and sunnier places a slightly higher one. 68
Is the sun bad for pink eye?
We found no study of sunlight and pink eye, the common name for infectious or allergic conjunctivitis. Ultraviolet causes its own inflammation of the conjunctiva, photoconjunctivitis, which the WHO describes as painful and reversible. 2
What we do not know
No study in our library measured the sunlight that reached an eye. The large cataract and pterygium studies estimated it from working lives, climate records, and questionnaires, and the myopia trials measured light at the wrist.
Whether sunlight plays a part in macular degeneration is unsettled, and the two pooled analyses of the studies disagree.
No trial has tested whether sunglasses or hats prevent cataract or pterygium. The evidence is association, plus one study in which sun damage on the eye's surface shrank faster in young adults who wore sunglasses.
How much brightness is enough to protect children's eyes is unsettled. Taiwan found an effect at 1,000 lux, Shanghai's analysis pointed to 120 to 150 minutes a day at around 5,000 lux, and a pooling of trials found no slowing in eyes already nearsighted while the 2018 Taiwan trial found it.
On a climbing expedition on the north side of Everest, reported in 2013, 15 climbers and 10 Sherpas wore ultraviolet dosimeters for the whole trip, one at the crown of the head and one in the armpit for comparison. Fourteen of the dosimeters were lost. The report lists cataract among the climbers' risks and recommends sunglasses. 69
Key studies
- MacFaul, 1969, London. Real sun, 1966 eclipse. Of 29 burned eyes, 15 ended at 6/9 or worse, 11 of them after looking through sunglasses, filters, or smoked or stained glass. 9
- Taylor and colleagues, 1988, Chesapeake Bay. Estimated ocular dose from working lives. Doubling ultraviolet B went with 1.6 times the risk of cortical cataract among 838 watermen. 7
- Taylor and colleagues, 1989, Chesapeake Bay. Pterygium in 140 of 838 watermen rose with estimated ultraviolet. 28
- Oldenburg and colleagues, 1990, Los Angeles. UVB lamp on rabbits. A drop of octinoxate gave a sun protection factor of at least 4 to the cornea. 11
- Taylor and colleagues, 1992, Chesapeake Bay. Advanced macular degeneration went with more estimated blue light, odds ratio 1.36. 34
- Cruickshanks, Klein, and Klein, 1992, Beaver Dam. Estimated ambient ultraviolet B went with more severe cortical clouding in men only. 24
- Bito and colleagues, 1997, Louisville. Eye color was stable after age six in most of 1,513 twins, and changes tracked genes. 59
- Danis and Stephens, 1997, Indianapolis. Real sun on skin after fluorescein. An author's shoulder reddened through SPF 30 sunscreen. 65
- West and colleagues, 1998, Salisbury. Estimated ocular dose. Cortical clouding rose with ultraviolet B in 2,520 older adults. 23
- Threlfall and English, 1999, Perth. Estimated ocular dose. The top quarter had 6.8 times the odds of pterygium. 29
- Sliney, 1999. Geometry of ocular exposure. The highest dose comes under light overcast with a visible horizon. 21
- Michaelides and colleagues, 2001, United Kingdom. Real sun, 1999 eclipse. None of 70 cases reported lasting loss at six months. 18
- Rochtchina and colleagues, 2001, Blue Mountains. Cortical clouding was four times greater in the lower nasal lens. 25
- Di Girolamo, Coroneo, and Wakefield, 2003, Sydney. UVB lamp on halved pterygia. Exposed halves released 2.7 times as much MMP-1. 30
- Rose and colleagues, 2008, Sydney and Singapore. Nearsightedness at 3.3 against 29.1 percent, with 13.75 against 3.05 hours a week outdoors. 44
- Ashby, Ohlendorf, and Schaeffel, 2009. Chicks in daylight and lamps. Fifteen minutes of daylight a day cut induced myopia to −1.1 diopters from −5.3. 48
- Guggenheim and colleagues, 2014, England. Birth cohort. Vitamin D had no link to myopia independent of time outdoors. 50
- He and colleagues, 2015, Guangzhou. Real sun, randomized by school. New nearsightedness at 30.4 against 39.5 percent. 8
- Torii and colleagues, 2017, Tokyo. Lamps on chicks, eyewear records in children. Violet light and myopia, with a patent pending. 12
- Schwab and colleagues, 2017. Iris freckles went with sunburns and sun-damaged skin in 632 people. 60
- Wu and colleagues, 2018, Taiwan. Real sun, randomized by school, light meters. Less myopic shift, including at 1,000 lux. 49
- Zhou and colleagues, 2018. Fourteen studies pooled. Sunlight and macular degeneration, odds ratio 1.10, with a range that includes no effect. 40
- Sharma and colleagues, 2021, Himachal Pradesh. Real sun. Three sunbathers with retinal damage. 17
- He and colleagues, 2022, Shanghai. Real sun, randomized by school, wrist monitors. New nearsightedness fell by 11 to 16 percent. 46
- Renzi-Hammond and colleagues, 2022, Georgia. Xenon lamp. A violet-filtering contact lens paid for by its maker cut halos by 30 percent. 58
- Masili, Duarte, and Ventura, 2024, Brazil. Model of 214 lenses. Field of view, not pupil size, drove the extra ultraviolet from poor lenses. 54
Definitions
- Cataract. A clouding of the lens. Cortical cataract forms in the outer layers, nuclear cataract in the center, and posterior subcapsular cataract under the back surface.
- Cluster randomized trial. A trial that assigns whole groups, such as schools, by chance, rather than single people.
- Confidence interval. The range of values that the data leave plausible for a result, usually at 95 percent confidence. A range that includes 1.0 for a ratio means no effect cannot be ruled out.
- Conjunctiva. The thin membrane over the white of the eye and the inside of the lids.
- Cornea. The clear window at the front of the eye. It absorbs most UVB.
- Diopter. The unit of a lens's focusing power. A negative number means nearsightedness.
- Dosimeter. A badge or sensor that records the dose of radiation reaching the place where it is worn.
- Drusen. Yellow deposits under the retina that come before macular degeneration.
- Fluorescein angiography. An eye test in which a dye is injected into a vein and the retina is photographed as the dye passes through its vessels.
- Hazard ratio. How much faster an outcome occurs in one group than in another over time. 1.0 means no difference.
- Lens. The clear disk behind the pupil that focuses light onto the retina. It absorbs nearly all UVA in adults.
- Limbus. The ring where the cornea meets the white of the eye. Its stem cells renew the corneal surface.
- Lux. A unit of brightness as the eye sees it. The chick study used 500 lux for ordinary lab light and 30,000 lux for daylight.
- Macula. The small central patch of the retina used for reading and recognizing faces.
- Macular degeneration. An age-related disease that destroys the macula and with it central vision. Exudative macular degeneration is its wet form.
- Meta-analysis. A study that pools the results of earlier studies into one estimate.
- Myopia. Nearsightedness, usually because the eyeball has grown too long for its focus.
- Odds ratio. The odds of an outcome in one group divided by the odds in another. 1.0 means no difference.
- Optical coherence tomography. A scan that images the layers of the retina with light.
- Photokeratitis. Inflammation of the cornea from ultraviolet, also called snow blindness or arc eye.
- Pinguecula. A small yellowish thickening of the conjunctiva over the white of the eye.
- Pterygium. A wedge of fleshy tissue that grows from the white of the eye onto the cornea.
- Retina. The layer of light-sensing cells at the back of the eye.
- Solar retinopathy. Damage to the retina, usually the macula, from sunlight, most often from looking at the sun.
- Solar simulator. A lamp built to give off a spectrum close to that of sunlight.
- UVA. Ultraviolet light from 315 to 400 nanometers. The lens absorbs most of it.
- UVB. Ultraviolet light from 280 to 315 nanometers. The cornea absorbs most of it.
- Uveitis. Inflammation of the uvea, the pigmented middle layer of the eye that includes the iris.
- Visual acuity. Sharpness of sight. In British notation 6/6 is normal and 6/9 means seeing at 6 meters what a normal eye sees at 9.
Statistics
A London eye hospital saw 19 people with solar retinal burns after real sun exposure (18 from watching the partial eclipse of May 20, 1966, one from sunbathing): 13 males and 6 females. 9
Of the 19 patients with solar retinal burns, 10 had both eyes burned, 6 the right eye only, and 3 the left eye only, 29 eyes in all. 9
Of the 18 people burned by watching the 1966 eclipse, 11 (61 percent) were schoolchildren aged 9 to 18. 9
All 19 patients with solar retinal burns were aged 9 to 24, except one woman aged 45. 9
Of the 19 people with solar retinal burns, 8 looked at the sun with nothing in front of their eyes; the other 11 used sunglasses, stained or smoked glass, filters, old photographic negatives, a pinhole, or half-closed lids, and one relied on passing thin cloud. 9
Among the 19 people with solar retinal burns, looking time was unknown in 9, up to 5 minutes in 6, and over 10 minutes in 2. 9
Counting all 29 eyes burned by the sun in the London series (eclipse and sunbathing), 15 had a final visual acuity of 6/9 or worse at the end of follow-up of 6 to 12 months. 9
Counting all 29 burned eyes (eclipse and sunbathing), 5 saw 6/6 and none saw 6/5 at the first examination; at the final examination 10 saw 6/6 and 4 saw 6/5. 9
Of the 15 burned eyes that never recovered better than 6/9 vision, 11 had looked at the sun through sunglasses, filters, smoked or stained glass, or old negatives. 9
A 13-year-old who watched the 1966 eclipse for 30 seconds through smoked glass still had only 6/9 vision in his left eye a month later, with a lasting blind spot beside the center of vision. 9
A 22-year-old psychiatric inpatient who watched the whole 1966 eclipse had 6/24 vision in both eyes 3 days later; neither eye improved, and he was registered partially sighted. 9
A 14-year-old who watched the 1966 eclipse through dark glasses and an old film negative for about 3 minutes had 6/36 vision in her right eye 4 days later, and 6 months later a small macular hole with 6/18 vision. 9
A 23-year-old man who gazed at the sky without sunglasses for about 6 minutes while sunbathing still had blind spots at the center of vision in both eyes a month later that hampered reading, though his acuity was back near 6/6. 9
An 18-year-old carpenter who watched the 1966 eclipse through stained glass for about one minute had already had one solar eye injury, from an eclipse 4 years before. 9
A 45-year-old amateur astronomer burned by watching the 1966 eclipse through smoked glass recovered 6/5 acuity in each eye after 6 weeks, 2 of them on steroids, but a small blind spot remained in each eye. 9
The May 20, 1966 partial eclipse that burned the eyes of 18 people in this London series lasted 2 hours, from 9:20 to 11:20 in the morning, in clear weather. 9
An 18-year-old who watched the 1966 eclipse 3 days before a national school exam fell to 6/9 vision with central blind spots that hampered his writing; he recovered fully in 4 weeks. 9
Outside the eclipse series, MacFaul describes a 27-year-old meteorological worker whose solar macular burn, sustained at work, left his right eye at 6/24. 9
In a case series of 3 people exposed to real sun, all had outer retinal defects on OCT scans typical of solar retinopathy, none gave a history of prolonged sun gazing, and the common factor was repeated sunbathing for a few hours. 17
The 3 sunbathers with solar retinopathy all lived in India's Solan district, at about 1,300 to 1,500 m above sea level, where snow often falls in winter. 17
A 42-year-old farmer who had sunbathed for long hours on the previous 3 days to fight the cold, and denied sun gazing, had best-corrected vision of 6/9 and a central blind spot in both eyes. 17
A 40-year-old woman who habitually sunbathed for long hours lying on her back with her eyes closed had best-corrected vision of 6/12 in the right eye and 6/6 in the left. 17
A 28-year-old man who sunbathed often, especially in November and winter, had 6/12 vision in his left eye, where the central retina measured 206 µm thick, against 236 µm in his right eye. 17
In a retrospective comparison of 310 Japanese students with myopia, eyeballs lengthened a mean 0.25 mm in a year in 211 wearing ordinary eyeglasses that block violet light, against 0.17 mm in 99 wearing violet-transmitting contact lenses. 12
The eyeglass group (211 children, mean age 12.2) and the violet-transmitting contact-lens group (99 children, mean age 13.9) differed significantly in age, starting refraction, and starting eye length. 12
Among contact-lens wearers aged 13 to 18 with myopia, eyes lengthened a mean 0.19 mm in a year in 31 wearing lenses that partly block violet light, against 0.14 mm in 116 wearing violet-transmitting lenses. 12
Spectral tests of 14 of the most common eyeglass lenses in Japan found no violet light (360 to 400 nm) passing through any of them. 12
Spectra measured on sunny early-summer days in Tokyo found no light below 400 nm coming through the UV-protected glass of an office, a taxi, or a hospital entrance hall. 12
Of 11 artificial lenses for cataract surgery tested for light transmission, most blocked violet light. 12
In lamp-lit chicks with one eye covered by a goggle for 7 days to induce myopia, the covered eye developed a mean −15.18 D of myopia without added violet light (4 eyes) and −4.59 D with it (5 eyes). 12
In the uncovered eyes of the same lamp-lit chicks, 15 without violet light drifted to −1.08 D of myopia while 15 with violet light became farsighted at +1.31 D. 12
In 17 goggled chick eyes, the more violet light the goggle let through, the less the eye lengthened (Spearman correlation −0.513). 12
In a chick experiment with a 305 nm UVB lamp, the birds kept both eyes shut all day because the UVB caused severe corneal inflammation, with 21 exposed chicks against 18 unexposed. 12
Adding violet light did not change chick blood vitamin D: both groups averaged around 20 ng/ml at 6 hours (8 per group) and at 1 week (14 and 16 chicks). 12
In lamp-lit chicks, the myopia-suppressing gene EGR1 in the back of the eye rose to almost twice its level under violet light compared with no violet light. 12
The chick cages were lit at a mean 1,262 lux without violet light and 1,349 lux with it, a difference that was not significant. 12
In a mouse cone photoreceptor cell line, 600 mJ/cm2 of lamp light at 370, 380, 390, or 400 nm raised the myopia-suppressing gene EGR1, but 360 nm did not (4 samples each); the authors equate the dose to about 100 minutes of north-sky sunlight at 10 a.m. 12
In the chick cages lit with added violet light, UV (290 to 390 nm) irradiance averaged 0.413 mW/cm2, against 0 in cages without it. 12
Four of the violet-light myopia study's authors were applying for a patent on products for myopia suppression when it was published. 12
In chicks given blue LED light (about 1,035 lux) instead, myopia progression was only weakly suppressed (14 per group), and violet light's effect on eye lengthening in the covered eye was significantly stronger than blue light's at matched total irradiance. 12
A randomized, double-masked trial fitted 61 people (77 percent women) with a blue-violet-filtering contact lens in one eye and a clear lens in the other, then tested them against a point of light from a lamp imitating sunlight; 58 completed the study. 58
The test contact lens filtered about 60 percent of visible light between 380 and 450 nm, against about 10 percent for the clear control lens. 58
Under a sun-imitating xenon lamp, halos around a bright point of light averaged 39.73 mm across with the blue-filtering contact lens and 55.14 mm with the clear lens, in 61 people. 58
Modeled across the 61 wearers, the blue-filtering contact lens shrank halo diameter by 30.3 percent (95% CI 20.1 to 39.2) compared with the clear lens in the other eye. 58
Starburst spokes around the lamp averaged 79.52 mm across with the blue-filtering contact lens and 106.29 mm with the clear lens, a modeled reduction of 23.4 percent (95% CI 18.4 to 28.1). 58
With the blue-filtering contact lens, the gap needed to see two points of violet (403 nm) light as separate was 30.1 percent smaller than with the clear lens (95% CI 22.6 to 36.9). 58
Under white, sun-imitating light, the two-point gap was 18.3 percent smaller with the blue-filtering contact lens than with the clear lens (95% CI 12.8 to 23.6). 58
The contact-lens glare trial was paid for by Johnson & Johnson Vision Care, which employed 2 of its 4 authors; the other 2 had worked as its consultants. 58
At 403 nm (violet), the filtering contact lens absorbed more than 90 percent of light at its peak, against about 5 percent for the clear lens. 58
A Sydney lab study of pterygium, the fleshy growth onto the cornea, used 10 surgically removed pterygia from 6 men and 4 women (mean age 47.6) and 10 samples of normal conjunctiva for tissue staining. 30
In pterygium cells grown in dishes and dosed with a UVB lamp, messenger RNA for MMP-1, a collagen-cutting enzyme, rose 2.3-, 3.7-, 4.6-, and 4.1-fold after 5, 20, 40, and 60 mJ/cm2. 30
MMP-1 protein from lamp-dosed pterygium cells peaked at 84.54 ng/mL after 20 mJ/cm2 of UVB; at 50 mJ/cm2 and above it fell below detection, perhaps because the cells died. 30
After one 20 mJ/cm2 UVB lamp dose, pterygium cells made 20 times as much MMP-1 as undosed cells at 48 hours and 80 times as much at 72 hours. 30
After 20 mJ/cm2 of lamp UVB, cells grown from pterygium released 49.6 ng/mL of MMP-1 and cells from the limbus 43.8 ng/mL, against 1.7 ng/mL from normal conjunctival cells. 30
Seven fresh pterygia were cut in half at surgery; the halves given 20 mJ/cm2 of lamp UVB released 2.7 times as much MMP-1 as the unexposed halves. 30
Only about 20 percent of cultured human corneal surface cells survived 20 hours after a 50 mJ/cm2 UVB lamp dose, a dose range the authors equate to about 20 minutes of sunlight on the cornea. 5
By 36 hours after the 50 mJ/cm2 UVB lamp dose, fewer than 10 percent of the cultured corneal cells were alive; cells pretreated with arsenite, a stress-protein inducer, survived at 40 percent. 5
In cultured human corneal cells, activation of caspase 3, an enzyme of programmed cell death, appeared after a UVB lamp dose of 20 mJ/cm2 and became marked at 35 to 50 mJ/cm2. 5
In colonies of human limbal stem cells, which renew the corneal surface, a non-lethal 20 J/cm2 UVA lamp dose raised the share of cells in autophagy, a self-cleaning stress response, from 35.7 to 59.4 percent at 6 hours. 70
After the 20 J/cm2 UVA lamp dose, the share of limbal stem cells with the master eye-surface gene regulator PAX6 in the nucleus fell from 22.1 to 8.0 percent. 70
In one experiment, the 20 J/cm2 UVA lamp dose cut the share of dividing limbal stem cells from 34.7 percent in unexposed colonies to 7.9 percent. 70
HeLa cells (a human cancer cell line) soaked in retinal, the light-catching molecule of the eye, and set in direct sunlight for 30 minutes began to die at 250 nM retinal; about 28 percent died at 1 µM and about 37 percent at 10 µM. 41
In retinal-loaded HeLa cells in direct sunlight (about 62 µW/cm2 at 445 nm), a membrane signaling lipid was knocked loose in 27.7 percent of cells after 1 minute and in 84.8, 85.5, and 99.3 percent after 5, 15, and 30 minutes. 41
In cultured human retinal pigment epithelium cells, about 77.6 percent showed DNA damage after retinal plus blue lamp light, against 16.9 percent with blue light alone and 0.8 percent with retinal alone. 41
Purified lens proteins from cow eyes given 15.4 J/cm2 from a solar simulator responded unevenly: alpha-crystallin only slightly paired up, while beta-crystallins paired up extensively and gamma-crystallins both paired up and broke into fragments. 27
The same study's solar-simulator doses reached up to 15.4 J/cm2 of light that was 95 percent UVA and 5 percent UVB, which the authors call physiologically relevant for skin and eye proteins. 27
Of 14 healthy Dutch adults aged 65 to 80 who wore light meters in real daylight, 9 never wore sunglasses, 3 always wore them outdoors, and 2 wore them occasionally. 71
Wrist-worn light meters on older Dutch adults recorded an average of 658 lux over 24 hours in summer against 178 lux in winter. 71
The older Dutch adults spent a mean 162 minutes a day above 1,000 lux in summer (13 people) against 65 minutes in winter (10 people), measured at the wrist. 71
The same older adults spent a mean 79 minutes a day above 3,000 lux in summer against 25 minutes in winter, measured at the wrist. 71
Among 9 older Dutch adults measured in both seasons, those who got the most bright light in summer also got the most in winter (r = 0.708 for time above 1,000 lux; r = 0.823 above 3,000 lux). 71
Excised corneas from 3 healthy rabbits passed an average 78.4 percent of light at 800 nm and 68.8 percent at 400 nm, but only 9 percent at 300 nm (UVB) and none below that. 11
In rabbits under a UVB fluorescent lamp (peak 308 nm) at 15 cm, 15 minutes of exposure caused mild keratitis (corneal inflammation) in every eye tested; longer exposures caused more inflammation. 11
One drop of the sunscreen filter octinoxate (Parsol MCX) in a rabbit's eye prevented keratitis through 30- and 60-minute UVB lamp exposures, 3 rabbits per exposure time, while 15 minutes reliably caused keratitis in castor-oil-treated control eyes. 11
The octinoxate eye drop let rabbit eyes take four times the UVB lamp dose without visible keratitis, which the authors expressed as a sun protection factor of at least 4, and the protection held when exposure started 60 minutes after the drop. 11
One drop of 10 percent avobenzone (Parsol 1789) cut UVA transmission through rabbit corneas by 20 percent; the effect lasted 120 minutes and was gone by 240 minutes. 11
The authors attribute the octinoxate drop's protection of rabbit corneas to its absorbing more than 75 percent of the lamp's UVB (peak 308 nm). 11
Eighteen female rabbits got eye drops six times a day in both eyes: saline, 50 µg/mL zinc oxide nanoparticles (a sunscreen ingredient), or 50 µg/mL vanadium pentoxide, 6 rabbits per group. 72
After a scraped wound to the cornea, rabbit eyes reached half-closure in 27.0 hours with saline drops, 35.5 hours with zinc oxide nanoparticle drops, and 27.8 hours with vanadium pentoxide drops. 72
Every wounded rabbit cornea in the saline and vanadium groups had healed by 96 hours, but half of the 6 zinc-oxide-treated rabbits still had a corneal ulcer at 96 hours and 2 of 6 corneas were unhealed at 105 hours. 72
Zinc oxide nanoparticle drops slowed rabbit corneal healing significantly from 33 to 81 hours after wounding compared with saline or vanadium drops. 72
Of 11 metal oxide nanomaterials tested on cultured human corneal surface cells for 24 hours, zinc oxide markedly cut viability at 5 µg/mL and vanadium pentoxide at 12.5 µg/mL. 72
Seven days of zinc oxide or vanadium nanoparticle drops caused no measurable change in rabbit tear production, tear film break-up time, or tear meniscus size. 72
The 50 µg/mL zinc oxide dose given to the rabbits was 100 times the highest concentration (about 0.5 µg/mL) that cultured corneal cells tolerated, chosen to allow for dilution by tears on the eye. 72
In the hen's-egg membrane test, a lab stand-in for eye irritation, a semi-solid sunscreen carrying oxybenzone in nanocapsules caused no bleeding, breakdown, or clotting of the membrane over 300 seconds and was classed non-irritating. 73
Three patients who went into real sun after an intravenous fluorescein dye test of the retina all developed red, swollen, painful skin on sun-exposed areas within 1 hour; each noticed redness and tenderness within about 30 minutes. 65
In a one-person sun challenge, after swallowing 6 ml of 25 percent fluorescein, a patch of skin put in direct sun began to redden within about 20 minutes, with burning and tingling. 65
In the fluorescein sun-challenge volunteer, the redness and tenderness of the post-dye patch resolved over 7 days, like a severe sunburn, even though SPF 30 sunscreen had been applied. 65
A literature, CDC, and ocular drug side-effect database search by the fluorescein authors found only one published case of probable fluorescein phototoxicity and two of photoallergy. 65
One of the 3 patients, referred with a diagnosis of fluorescein allergy, had a repeat fluorescein angiogram with no adverse effects. 65
Among 303 volunteers tested under a xenon solar simulator, the mean skin burn threshold (minimal erythema dose) was 846 mJ/cm2 for 71 with blue eyes, 800 for 27 with gray, 910 for 44 with green, and 1,100 for 161 with brown eyes. 61
Grouping blue, gray, and green eyes together, the 142 light-eyed volunteers burned at a mean 858 mJ/cm2 of solar-simulator light against 1,101 for the 161 brown-eyed volunteers (P < 0.001). 61
In 193 healthy Brazilians sorted into six skin types before UVB lamp testing, blue eyes were found in 50 percent of skin type I, 31 percent of type II, 14 percent of type III, and none of types IV to VI. 62
In the same Brazilian study, green eyes appeared in 3 percent of the darkest skin type (VI, 32 people), so the authors judged that hair and eye color could not be used to sort people into the skin-type groups. 62