A 65-year-old man in San Diego began taking amiodarone, a heart-rhythm drug then still being tested in the United States. Within a week, five to ten minutes of summer noon sun left his skin stinging and burning, worst on his bald scalp, and on a trip to Hawaii it took two minutes. In his car he could sit for an hour in the same sun and feel nothing. His doctors at the University of California, San Diego, took a piece of tinted glass from the car's windshield, which had since been broken, laid it on the inside of his forearm, and shone a solar simulator, a lamp built to imitate sunlight, through it. 1
Skin under the windshield glass stayed pale. Skin under a filter that removed only wavelengths shorter than 350 nanometers, or billionths of a meter, reddened. The damage came from ultraviolet A, or UVA, the longer ultraviolet that makes up most of what reaches the ground, in a band from 350 to 380 nanometers that the windshield stopped and the sunscreens of the day mostly let through. Normal skin reddens after a UVA dose of about 30 joules per square centimeter, and his reddened after 4 to 8. When he stopped the drug the sensitivity faded, and by ten weeks it was gone. 1
Light that most people tolerate brings on a rash, hives, blisters, or burning pain in a minority. The commonest of these reactions, polymorphic light eruption, an itchy rash that comes hours or days after sun, turned up in about 10 percent of 19,287 people in population surveys pooled in 2023. At the other end, erythropoietic protoporphyria, an inherited disease in which visible light burns the skin, affects about 1 person in 150,000. Some of these reactions start in the body, in the immune system or a gene, and others need a second ingredient in or on the skin, such as a drug, a plant juice, a soap antiseptic, a painkilling gel, or a sunscreen chemical. 2; 3
This page covers polymorphic light eruption, hives from light, lupus, the porphyrias, the drugs and plants that make skin sensitive to light, and allergy to sunscreen itself. Each study is named as real sun, a lamp or solar simulator, a product applied to people, animals, or cells.
Key facts
- In population surveys pooled in 2023, polymorphic light eruption affected about 10 percent of 19,287 people, from 0.65 percent in China to 21.4 percent in Ireland. 2
- In Vienna, lamp tests brought on polymorphic light eruption in 69 of 142 patients, and UVA alone triggered it in 56 percent of those 69. 4
- Before treatment, 39 Swiss patients with erythropoietic protoporphyria could stay in direct sun for a median of 10 minutes before a painful reaction. 3
- The US Food and Drug Administration (FDA) puts the added skin cancer risk from hydrochlorothiazide, a blood pressure drug, at about 1 squamous cell carcinoma per 16,000 patients a year. 5; 6
- Of 40 Toronto patients who described a sunscreen allergy, 26 tested positive; in Portland, 8 of 11 who believed they were allergic tested negative. 7; 8
Everard Home showed in 1820 that sunlight, not heat, scorched skin
In 1817 the London physician Thomas Bateman published an engraving of a rash on the backs of the hands and fingers, which he called eczema solare. It came in summer, he wrote, from "the direct irritation of the sun or heated air". 9
Three years later Sir Everard Home held both hands in the sun with a thermometer on each, one hand under black cloth. After ten minutes the covered hand was the hotter, 106 degrees Fahrenheit against 98 in his third trial. "In every one of these trials the skin that was uncovered was scorched, the other hand had not suffered in the slightest degree," he told the Royal Society in November 1820. Part of what set him going was a silver fish in a pond at Spring Grove whose back was scorched after the trees shading the pond were cut down. 10
"I have had much difficulty in finding a name which should be even tolerably appropriate," the surgeon Jonathan Hutchinson told his students in 1879, and he settled on summer prurigo, "a form of prurigo hitherto undescribed". It usually began around puberty, flared in summer on the face and arms, and left tiny scars. He promised that at the end of the lecture he would "produce the lad" whose portrait he had shown, painted at 13 in 1867, to prove that the rash had gone. 11
In 1900 the Danish dermatologist Carl Rasch named an itchy summer rash polymorphic light eruption. Lecturing in London in 1926, he credited Home with the finding that the sun's harm to skin "is due to the light and not to the heat". He judged it "impossible to keep Hutchinson's prurigo distinct" from his own rash. In Copenhagen, 230 such cases had turned up among 31,000 skin patients from 1911 to 1925, which he put down to Danish skin and hair being "so frequently blonde". 10
About one person in ten has polymorphic light eruption, most of them women, and few see a doctor for it
The rash of polymorphic light eruption is small raised spots, blisters, or red blotches that itch or burn. It appears within hours or days of sun, most often in spring and summer, and it favors skin that spends the winter covered, such as the arms, legs, and chest. According to Britain's National Health Service, it usually clears in about a week out of the sun. 12
In 1982 Morison and Stern asked 271 apparently healthy people about rashes after sun, and 10 percent described this one. Those people needed more than three hours of sun to bring it on, against 30 minutes for patients seen in a clinic. 13 In 1986 researchers sent a questionnaire to the 412 employees of a Swedish drug company, and 83 of the 397 who answered had the rash, 89 percent of them women. Three percent of the 83 had seen a doctor about it, and 84 percent said their skin grew used to the sun as summer went on, which doctors call hardening. 14
A 1994 survey found the rash in 5.2 percent of 172 people in Perth, 3.6 percent of 196 in Ballarat, and 14.8 percent of 182 in London. A 2010 survey of 6,895 Europeans from the Mediterranean to Scandinavia put it at 18 percent, with no rise toward the north. 15; 16
When a 2023 meta-analysis, a study that pools the results of earlier studies, combined nine population surveys covering 19,287 people, it put the rash at 10 percent. The true figure could be anywhere from about 6 to 15 percent. Countries ranged from 0.65 percent in China to 21.4 percent in Ireland, and the farther a country lay from the equator, the more of it there was. In the 1994 survey women outnumbered men three to one, and among 167 patients in Vienna 4.7 to one. 2; 15; 4
Lamps bring on polymorphic light eruption in about half of patients, most often with UVA
In a study published in 1986, Ortel, Hönigsmann, and their colleagues in Vienna tried to raise the rash on purpose in 142 patients. They chose two 4-by-4-centimeter squares of skin, where possible on places the rash had appeared before, and every day for up to eight days gave one square UVA from a lamp filtered through window glass and the other UVB, the shorter ultraviolet that causes most sunburn. In 69 patients the rash came, after two to eight exposures. It came from UVA alone in 56 percent of them, from UVB alone in 17 percent, and from both in 26 percent. 4
Before testing, the team had sorted patients by what they said kept the rash away. Those who said window glass or sunscreen protected them were assumed to react to UVB, which glass stops. The comparison "unexpectedly did not confirm this assumption", and that group had more positive results with UVA than with UVB. The authors wondered whether the rash "may perhaps represent a 'normal' cutaneous reaction that anyone can develop under adequate irradiation and certain environmental conditions." 4
Before the summer, the same clinic gave 122 patients three to four weeks of PUVA, in which a light-sensitizing plant chemical called psoralen is taken by mouth before a dose of UVA from a lamp. The treatment itself brought out the rash in 27 of them. Of the 51 who came back in the autumn, 64 percent said they had been fully protected that summer, 26 percent partly, and 10 percent not at all. 4
At St Thomas' Hospital in London in 2004, Chantalle van de Pas, Antony Young, and their colleagues looked at what goes wrong in the skin. In most people a dose of ultraviolet damps the skin's immune reactions for a time. The team gave 22 patients and 23 matched volunteers one dose from a solar simulator, put a chemical called dinitrochlorobenzene on the same spot to set up a contact allergy, and later measured the allergic swelling by ultrasound. At one minimal erythema dose, the smallest dose that leaves skin just visibly red a day later, ultraviolet cut the allergic response by 78 percent in the volunteers and by 44 percent in the patients, although both groups reddened alike. The authors read this as "a flaw in their immunoregulatory response to ultraviolet-radiation", seen only in a narrow band of doses. 17
Sunscreens with strong UVA protection prevented provoked polymorphic light eruption in trials paid for by their makers
In Tübingen in 2008, dermatologists spread 2 milligrams per square centimeter of Daylong, a sunscreen sold as giving very high protection against UVB and UVA, on one forearm of each patient, and the same cream without its sunscreen chemicals on the other. Both arms then took three days of UVA and UVB from lamps. Of 12 patients, 2 were dropped because the rash never came on the unprotected arm, and in the other 10 it came on the unprotected arm and on none of the protected ones. The team had planned to test 30 and stopped at 12 after checking the results partway through. Spirig Pharma, which makes Daylong, paid for the study. 18
The largest trial ran in real sun, at one site in Puerto Rico, with 144 people with the rash recruited from across the United States, who took rising doses of sun on up to six days. Each wore a cream with four sunscreen chemicals and a sun protection factor, or SPF, of 40 on one side of the body and, on the other, the same cream without one of its two UVA-absorbing chemicals, ecamsule or avobenzone. The full cream did better on 41 of 73 people and the cream without ecamsule on 8, and against the cream without avobenzone the count was 26 to 11 of 71. L'Oréal USA paid for the trial, and three of its authors worked for the research arm of Galderma. 19
The trial counted a later or milder rash as better, and most people still broke out on both sides, 75.4 percent in one comparison and 83.1 percent in the other, and 22 of the 144 never broke out at all. Both sides carried an SPF 40 cream, so the trial measured what one UVA-absorbing chemical added. 19
In Graz, 14 patients had four small patches of skin given low solar-simulator doses on four days in a row. An SPF 30 sunscreen put on beforehand prevented the rash in all of them. An after-sun lotion containing DNA-repair enzymes from two microbes, switched on with blue light an hour after each dose, lowered the rash score by 61 percent against untreated skin, and the true effect could be anywhere from about a third to five sixths lower. 20
Solar urticaria raises hives within minutes of light, sometimes from visible light alone
In 1923 W. W. Duke reported "urticaria caused by light" in the Journal of the American Medical Association. In solar urticaria, itchy wheals, the raised welts of hives, rise within minutes of light on the skin. A 2022 review called it "a rare, immunologically mediated photodermatosis" in which light sets off mast cells, the immune cells that release histamine in ordinary hives. 21; 22
In London in the 1970s, Charles Ramsay at the Institute of Dermatology treated nine patients with it. Two of the men had undressed on the beach and walked toward the sea, and as they entered the water hives spread over their bodies and their blood pressure fell. "Both patients had been rescued from the sea semiconscious after such a reaction," Ramsay wrote. 23
Ramsay exposed a small area of skin, usually one arm, to fluorescent lamps and repeated the dose every hour until the hives stopped coming, which took two to five exposures in all but one patient. The tolerance lasted 48 to 72 hours unless it was topped up every day. Six patients went on to build tolerance with sunlight itself, some spending "the first day or two of the vacation inducing tolerance with the sun". Two were fully helped, six partly, and one, who worked irregular hours, could not keep to the schedule. 23
A 28-year-old veterinary research scientist in Boston got itchy welts three to five minutes after sun, including when she sat "within 3 feet of a window through which the sun was shining," and chemical sunscreens did not help. Standard tests with UVA, UVB, and a slide projector raised nothing. In 1998 her doctors tried lasers and raised welts within three minutes at red wavelengths of 610, 650, and 690 nanometers, and none with blue or green light at five times the dose. 24 In a 1982 case, longer light from a lamp, given straight after the blue light that raised a woman's hives, switched them off, and heat from an electric hair dryer did not. 25
For a 24-year-old dental hygienist in Vancouver, even the lamps over the dental chair made her hands itchy and red. Sunscreens, antihistamines, two antimalarial drugs, four months of UVA hardening, and 53 sessions of PUVA had all failed. In 1999 her doctors at the University of British Columbia replaced her plasma, the liquid part of the blood, five times in ten days, 11.9 liters in all. Afterward their lamps could not raise the hives, and 21 months later she was almost free of them. 26
In twentieth-century experiments, patients' blood serum passed the reaction to other people, and the leading explanation holds that light turns some molecule in the skin into one the patient's antibodies recognize. The 2022 review said this model "does not account for several observations" and called the molecule "putative". Omalizumab, a drug that blocks IgE, the antibody behind ordinary allergies, helped 38 of 48 patients described in 22 published reports. 22; 27
Lamps provoke lupus rashes days after the dose, and sunscreens with strong UVA protection prevented them
In 1907 Rasch saw a man who had begun to wear "a very low-necked sailor's jersey (in the Argentine)" and then developed lupus erythematosus, an autoimmune disease that can attack the skin, on the parts of his chest and back that the jersey left bare. On the fingers, he noted, the disease "entirely spares those parts encircled by the patient's rings", and on the hand it stopped where the sleeve began. His assistant sorted 94 of the clinic's cases by the month they started, and 54 began between March and June, "when the light is most irritating". 10
In 1990 Lehmann and colleagues tested 128 patients with UVB and UVA lamps. They raised lesions like each patient's own lupus in 64 percent of those with subacute cutaneous lupus and 42 percent with discoid lupus, two forms that mainly affect the skin, and in 25 percent with systemic lupus, which can reach the joints, kidneys, and blood. A third of the provoked lesions came from UVB alone, 14 percent from UVA alone, and about half from both. They appeared within a few days, and some discoid lesions lasted for months. 28
With rising lamp doses of UVA, UVB, and visible light and two months of follow-up, Sanders and colleagues found abnormal reactions in 93 of 100 patients in 2003, and patients' own histories did not predict who would react. A 2006 study that filtered its lamps to pure UVB or pure UVA1, the longer part of UVA, raised no lesions in nine patients with subacute cutaneous lupus. 29; 30
In Düsseldorf in 2000, Stege, Krutmann, and colleagues raised lupus lesions with UVA and UVB lamps in 11 patients and put three commercial sunscreens on neighboring patches, with neither patients nor readers told which was which. One, with three UVA-absorbing chemicals and titanium dioxide, prevented lesions in all 11. The other two, whose only UVA-absorbing chemical was avobenzone, with titanium dioxide, prevented them in 5 and in 3. The three had similar protection against UVB, as their SPFs showed, and one author was supported in part by La Roche-Posay, a sunscreen maker. 31
In a 2011 trial in 25 patients, a broad-spectrum sunscreen and the same cream without its sunscreen chemicals went on different patches of back skin, in random order, before three days of UVA and UVB lamps. Lupus lesions appeared in 16 untreated patches, 14 patches under the plain cream, and none under the sunscreen. The authors cautioned that results from standardized phototests "might not be transferable to a clinical setting". 32
Sunlight can bring on lupus flares in people who have it, and it has not been shown to start the disease
In 2007 doctors in Würzburg reported a 34-year-old woman whose systemic lupus had been in remission without treatment for more than eight years, during which she had twins. On holiday in Turkey she sunbathed and burned, and a lupus rash spread over the sun-exposed skin. Six weeks later she had lupus nephritis, an inflammation of the kidneys, and treatment led to blood poisoning, more than two weeks on a ventilator, and dialysis. "Interestingly, in previous summer holidays, sunbathing did not lead to" lupus flares, the authors wrote. 33
In 1992 Rihner and McGrath polled 30 people with systemic lupus who were sensitive to light, and 13 said unshielded cool-white fluorescent lamps made their disease worse. The lamps gave off measurable UVB, the standard acrylic covers absorbed it, and where the covers were used patients reported almost no problems. 34 A European panel recommended sun protection for people with systemic lupus in 2024 at its lowest level of evidence. Data were scarce, it wrote, because the known role of sun in the disease imposes "ethical limitations for the conduct of RCTs", or randomized controlled trials. 35
The American College of Rheumatology's 1982 criteria, the checklist used to classify patients for research, included "skin rash as a result of unusual reaction to sunlight, by patient history or physician observation". The 2019 criteria from the European and American rheumatology societies dropped it. Photosensitivity "has otherwise disappeared," one of their authors wrote, surviving only in the sun-exposed pattern of skin lupus, because polymorphic light eruption "is much more common than SLE", systemic lupus, and redness soon after sun is easier to remember than lupus lesions, "which develop several days to a week after UV exposure". 36; 37
Whether sun can start lupus in someone who does not have it was examined in the Nurses' Health Studies, which enrolled about 238,000 American nurses in 1976 and 1989. Researchers estimated each woman's ultraviolet from her home address, and 297 women developed systemic lupus over 6.05 million years of follow-up. Women in the highest third of ultraviolet had 1.28 times the risk of those in the lowest third; the true effect could be anywhere from 4 percent lower to 70 percent higher, and the authors reported no association. The same women were 1.62 times as likely to have the butterfly-shaped malar rash across the cheeks at diagnosis, with a range from 4 percent higher to about two and a half times. 38
Soap antiseptics sensitized thousands of Americans to light before the FDA banned them in 1975
In 1939 Epstein described photoallergy to sulfanilamide, one of the first antibacterial drugs. In photoallergy, light changes a chemical on or in the skin into a form the immune system has learned to attack, so only people who have become sensitized react. In phototoxicity, the light-activated chemical injures the skin of anyone who has enough of it, like an exaggerated sunburn. 39
In 1961 Wilkinson reported photodermatitis in England from tetrachlorosalicylanilide, or TCSA, a germ-killing agent added to soap, and in 1963 Jillson and Baughman reported the same in the United States from a related antiseptic, bithionol. By 1975 the commissioner of the Food and Drug Administration wrote that "there are literally thousands of people in this country who are sensitized to TCSA, in most cases, by products such as soap and diaper rinses". In Denmark, which never sold TCSA, a soap with 2 percent of a related chemical, tribromsalan, had produced "approximately 80 verified cases". The FDA barred this family of chemicals, the halogenated salicylanilides, from drugs and cosmetics from 1 December 1975, by which time the main manufacturer said American sales were "virtually nil". 40; 41; 42
The FDA rule also recorded that in some people "severe reactions can reoccur in the presence of sunlight even though the individual is not exposed again", and dermatologists called such patients persistent light reactors. In 1984 Kaidbey and Messenger described eight whose light-triggered dermatitis had lasted 5 to 19 years; three reacted on testing to the soap antiseptics and five to musk ambrette, a fragrance since taken off the market. 42; 43; 44
In a 1992 report, a 64-year-old man with nine years of dermatitis on sun-exposed skin was first thought to have a lymphoma of the skin and was found to have actinic reticuloid, a severe chronic sensitivity to light. His rash flared "from sunlight through a car window on journeys of more than 30 min", and under a lamp he and two other patients were up to 27 times more sensitive to UVB than healthy people. At St John's Institute of Dermatology in London, 34 of the 62 patients found photoallergic from 1983 to 1998 already had a light-sensitivity disease, 19 of them chronic actinic dermatitis, a lasting eczema of sun-exposed skin. 45; 44
Ketoprofen gel caused more photoallergy than any sunscreen chemical in a 2012 European study
In Belgium from May 2001 to June 2002, dermatologists tested 20 people with rashes after using gels of ketoprofen, an anti-inflammatory painkiller, and found photoallergy to it in 17. Nine needed corticosteroids by mouth or injection, five were admitted to hospital, and one stayed sensitive to light for four months. 46
In southern Sweden, where the gel had been sold without prescription for more than three years, 35 patients with photoallergy to it reached the Malmö clinic in two years. Two who had rubbed it on their knees were first checked for a blood clot and were sent to the dermatologists when the scan came back clear. The Malmö authors cited relapses from "ketoprofen-contaminated objects as bandages and slippers". Of 26 tested, 19 were also photoallergic to fenofibrate, a cholesterol drug with a similar structure, and 40 percent of the 35 reacted to TCSA, the soap antiseptic of the 1960s. 47
In 2012 a European study photopatch tested 1,031 patients at 30 centers in 12 countries, taping chemicals to the back in two sets and shining UVA on one set. Of the 1,031, 200 had a photoallergic reaction. Ketoprofen caused it in 128 and another painkiller gel, etofenamate, in 59. Among sunscreen chemicals octocrylene caused the most, in 41, and 15 people reacted to ketoprofen, octocrylene, and oxybenzone together. Chemotechnique Diagnostics, which donated the test chemicals, and L'Oréal Research and Innovation funded the study. 48
In 2009 French regulators had suspended every marketing authorization for ketoprofen gels over their interactions with octocrylene. 48 French cases that tied octocrylene allergy to earlier ketoprofen use are on the sunscreen page.
In Coimbra, Portugal, another painkiller, benzydamine, caused photoallergy in eight patients from 2003 to 2007. Seven had rashes on the chin and lower lip from gargling an anti-inflammatory mouthwash. The eighth got his on the face, forearms, and hands from rubbing a benzydamine gel on his wife. 49
Lime, fig, celery, and hogweed juices burn skin in sunlight without any allergy
In 1942 Klaber gave the reaction its name, phyto-photo-dermatitis, in the British Journal of Dermatology. Plants such as limes, figs, wild parsnip, and giant hogweed carry furocoumarins, chemicals of the psoralen family, and UVA turns them into agents that damage skin cells whether or not a person has any allergy. 50; 51
In 1987 epidemiologists investigated 19 cases among grocery workers in Oregon. Two stores of the same chain carried the same brands of celery, and all 19 cases came from one of them, which had held a celery sale that quadrupled the usual volume. Workers who used tanning salons had the worst reactions. Two of three celery samples from that store had raised psoralen levels. 52
A 26-year-old woman in Texas sliced and squeezed two dozen limes for margaritas before a pool party. She felt burning as she got into the pool, and about seven hours after the squeezing a sharply bordered rash appeared on her lower abdomen and thighs. By the third day it had blistered, despite steroid tablets and cream. "She was the only attendee at the pool party who developed this rash," her doctors wrote. 53
In London a 46-year-old tree surgeon pruned a fig tree on a hot, sunny day in a T-shirt. Over two days blisters spread over both forearms and the back of his neck, about 8 percent of his body surface, while he treated them as sunburn with aloe vera, wet tea towels, and paracetamol. His face, also in the sun, was spared. His supervisor knew about fig sap and told him to go to hospital, and at the emergency department, his doctors wrote, "No member of the ED team on duty, up to consultant level, was aware of this skin reaction". 54
In Busan, South Korea, five people aged 57 to 87 soaked their feet in a folk remedy of boiled fig leaves, three of them for toenail fungus, and then walked out in the sun. Within a day their feet burned, swelled, and blistered everywhere except under the straps of their sandals, and one woman took two months to heal. 55 On holiday in Greece, an Italian couple ground fresh fig leaves into a homemade tanning lotion, covered themselves in it, and lay in the midday sun. Both blistered over much of the body, and their 9-year-old son, who touched the mixture by accident, was burned less. 56
A German dermatologist warned in 1984 that power lawn mowers spray the juice of giant hogweed over bare skin. The plants, he added, "should by no means be exterminated". 57
In erythropoietic protoporphyria visible light burns within minutes, and sunscreens made for ultraviolet block little of it
In 1961 Magnus and colleagues described "a new porphyria syndrome with solar urticaria due to protoporphyrinaemia". The porphyrias are disorders in making heme, the iron-holding part of hemoglobin. In erythropoietic protoporphyria, or EPP, protoporphyrin IX, a step on the way to heme, builds up in red blood cells and skin. It absorbs light most strongly around 408 nanometers, violet light at the short end of the visible range, and light absorbed by it sets off the painful reaction. 58; 59
In May 1985 the chemist David Dolphin of the University of British Columbia told a meeting of the American Association for the Advancement of Science that vampires and werewolves may have been people with porphyria, whose "legendary thirst for blood was a primitive effort to obtain heme," in the words of a United Press International report. He added that garlic "might well act to keep the vampires away". Stanley Schrier, head of hematology at Stanford, called it a "cute idea". 60
In Cambria, California, Michael McMillan, 28, had an acute porphyria, a form that had brought him abdominal pain and bouts of partial paralysis since he was 10 and had not marked his skin. After the story ran, a local resident asked whether the McMillans "were the kind of people who consumed blood", and he stopped his twilight walks along the shore. "If people believe I'm a werewolf or a vampire, I don't want to be out walking at the same time a child disappears," he told the Los Angeles Times. 61
In 2001 Moseley, Ferguson, and colleagues at the photobiology unit in Dundee tested seven patients whose skin reacted to blue light at 430 nanometers, four of them with EPP. Their skin reddened at 4.7 to 15 joules per square centimeter, against more than 82 for normal skin, and an ordinary SPF 25 sunscreen protected it by a median factor of 2. Protecting in the visible range, the authors noted, "would impact on the colour of the product", and the unit's own creams of zinc oxide and pigment-grade titanium dioxide came in Coral Pink, Beige, and Coffee. Beige protected by a median factor of 8. 62
A 17-year-old girl came to the university clinic in Mannheim with cheeks that burned and swelled for days after sun, and on the day of her visit her skin showed nothing. The diagnosis rested on her own photographs of past attacks, a history that as a small child she had often screamed hard after sun, and her blood. Free protoporphyrin in her red cells was 915.5 micrograms per deciliter, against a normal range of 0.5 to 5.0, and under a fluorescence microscope the cells glowed orange-red. 63
People with EPP "develop painful burns in everyday situations, when they are unable to avoid sunlight, for instance, when waiting at the bus stop, and even when crossing a street," doctors in Zurich wrote in 2020, and light through windows affects them too. 3 A 2021 laboratory test spread six sunscreens on plastic plates and found that five let 75 to 95 percent of light through at 500 nanometers. The best, which contained iron oxide pigments, had a median "porphyrin protection factor" of 4.22. 59
Afamelanotide added hours of pain-free sun over six to nine months in trials funded by its maker
Afamelanotide is a synthetic analogue of alpha-melanocyte-stimulating hormone, which drives pigment cells to make melanin, and it is given as a 16-milligram implant under the skin every two months. In two trials published in 2015 and funded by its maker, Clinuvel Pharmaceuticals, among others, 94 American and 74 European patients with EPP were randomly given afamelanotide or a dummy implant. Over six months the American patients on the drug spent a median of 69.4 hours in direct sunlight without pain, against 40.8 hours on the dummy. Over nine months in Europe the medians were 6.0 hours against 0.8, and the drug group had 77 painful reactions against 146. 64; 65
The European Medicines Agency authorized it in December 2014 under "exceptional circumstances", because the disease is too rare for complete information on its benefits, and noted that "the additional time spent in sunlight was small". The FDA approved it on 8 October 2019 as the first treatment in the United States to increase pain-free light exposure in EPP. Its US label tells patients to "maintain sun and light protection measures during treatment". 66; 67; 65
In Zurich, where 39 patients had used it outside a trial, the median longest time in sun without a painful reaction rose from 10 minutes to 180, by the patients' own report and with no comparison group. The pain of their worst reaction fell from a median of 10 out of 10 to 6. The study was paid for by a research foundation of the Zurich city hospitals. 3
Drugs taken by mouth lowered the burning dose to a fraction of normal, and two carry added skin cancer risk
In a case reported in 1975, a 34-year-old man being treated for Hodgkin disease with the cancer drug vinblastine went to the Daytona auto races in a wide-brimmed hat and sleeves to the wrist. Both hands blistered, the left one worse, because his seat put it in more direct sun. At the Miami Veterans Hospital a lamp reddened his skin in 5 to 10 seconds, where healthy people took 20 to 30. His doctors then injected a tiny amount of vinblastine into the skin of five healthy volunteers, and every injected spot reacted to a lamp dose too small to redden normal skin. Window glass blocked the reaction. 68
The US label for amiodarone says it "induces photosensitization in about 10% of patients" and that long use can turn exposed skin blue-gray, and all 10 of the San Diego group's amiodarone patients had reacted. The label for doxycycline, a common antibiotic, warns of "an exaggerated sunburn reaction" and tells patients to "avoid excessive sunlight or artificial ultraviolet light" and to consider sunscreen. It gives no number of hours to stay out of the sun after a dose, and the drug's half-life in the blood is 18 to 22 hours. 69; 1; 70 In 1990 a woman who had taken quinine for muscle cramps was still sensitive to light three years after she stopped it. 71
The label for voriconazole, an antifungal drug sold as VFEND, reports squamous cell carcinoma and melanoma "during long-term VFEND therapy in patients with photosensitivity skin reactions". It recommends sun avoidance and skin checks for children with sun damage "even after treatment discontinuation". About 21 percent of patients in its trials had vision changes that included photophobia, pain from light in the eyes. 72
Hydrochlorothiazide, a common blood pressure drug, makes skin sensitive to light. A 2018 Danish study compared the prescriptions of everyone in the national registry with a basal or squamous cell carcinoma from 2004 to 2012 with those of 20 people of the same age and sex for each. People who had taken 50,000 milligrams or more, about 11 years at 12.5 milligrams a day, had 3.98 times the odds of squamous cell carcinoma, and those above 200,000 milligrams 7.38 times. The authors noted that "no data on sun exposure were available". 73; 74
In 5.2 million matched American users, the FDA's Sentinel system found a smaller rise. Against users of angiotensin-converting enzyme (ACE) inhibitors, another class of blood pressure drug, the risk of squamous cell carcinoma was 4 percent higher, with a range of 2 to 6 percent, and 15 percent higher among white patients. The FDA put the added risk at about 1 squamous cell carcinoma per 16,000 patients a year, and 1 per 6,700 for white patients who had taken 50,000 milligrams or more. It told patients to keep taking the drug unless their doctor directed otherwise. 75; 5; 6
Photopatch tests confirmed a suspected sunscreen allergy in 15 to 65 percent of people, and clinics find different culprits by country
In 1978 Kaidbey and Kligman in Philadelphia were checking a new photosensitivity test on products they "felt certain were nonphotosensitizing". "To our astonishment, we obtained phototoxic responses to a popular proprietary sunscreen." The culprit was padimate A, a form of para-aminobenzoic acid, or PABA. Taped under a patch on students' backs and then given 14 minutes of UVA from a solar simulator, it raised a sunburn-like redness in 12 of 17. "Millions of units of sunscreens containing it have been sold annually for years," they wrote, and users who burned had concluded that "the product was ineffective". 76
In the summer of 1989 every Swedish pharmacy had to report customers' complaints about sunscreens. Of 58 complainants, 27 were photopatch tested, and 4 reacted to a sunscreen chemical. Complaints per unit sold rose with the product's SPF. 77
In Portland, Oregon, 27 people who said they were allergic to sunscreen came forward for a study published in 2010. "Photopatch testing is difficult for patients; hence, only 11 agreed to proceed with the testing," the authors wrote, and 8 of the 11 tested negative. At Toronto Western Hospital from 2001 to 2010, 26 of 40 patients who came in describing a sunscreen allergy reacted to their own sunscreen or its ingredients. The North American Contact Dermatitis Group noted in 2013 that "at least 1 study has documented that most individuals who suspect that they are allergic to sunscreen are not". 8; 7; 78
The chemicals in a photopatch test stay taped to the back for a day or two before the UVA. A 2026 European study compared a 24-hour wait with a 48-hour wait on the same people, and among 132 tested with sunscreen chemicals the shorter wait missed 53 percent of the sunscreen photoallergies. In Bogotá, 8 of 15 patients with photoallergy showed it only at a day-6 reading. The rates of sunscreen allergy in the UK and North American series are on the sunscreen page. 79; 80
In Shanghai from 2015 to 2019, the antipsychotic drug chlorpromazine gave a positive photopatch test in 562 of 2,365 patients, and doctors judged it relevant to the rash in 73. In New Delhi, a standard commercial test series found photoallergy in 2 of 70 patients with rashes on sun-exposed skin, and the commonest pattern, in 32, was a reaction to parthenium, a weed. 81; 49; 82
Misconceptions
What people believe
- Sun allergy is rare, and it works like hay fever.
- Light rashes are worst near the equator.
- Window glass protects people with light-sensitive skin.
- A high SPF protects against every reaction to light.
- A drug's effect on the skin ends when the dose wears off.
- Sunlight causes lupus.
What the studies show
- Polymorphic light eruption was found in about 10 percent of people in pooled surveys, and in lamp tests it goes with a failure of ultraviolet to damp the immune response. 2; 17
- A 1994 survey found the rash in 14.8 percent of people in London and 5.2 percent in Perth. 15
- Vienna patients who said glass protected them reacted more often to UVA, which glass lets through, and people with EPP react to light through windows. 4; 3
- Three sunscreens with similar SPFs protected 11, 5, and 3 of 11 lupus patients. 31
- A quinine sensitivity lasted three years after the drug was stopped, and persistent light reactions from soap antiseptics lasted up to 19 years. 71; 43
- Ultraviolet at women's homes was not associated with developing systemic lupus. 38
Questions people ask
Can you be allergic to the sun?
What people call sun allergy is usually polymorphic light eruption, an itchy rash hours or days after sun that about 10 percent of people in population surveys report. Lamp studies tie it to an immune response that ultraviolet fails to damp, which differs from the antibody allergy of hay fever. Solar urticaria, hives within minutes of light, is rare. 2; 17; 22
Can the sun cause lupus or a lupus flare?
Lamps raised lupus lesions in 25 to 64 percent of 128 patients with lupus, and case reports describe flares after sunbathing. In the Nurses' Health Studies, ultraviolet at women's homes was not associated with developing lupus. 28; 33; 38
Which medicines make you sensitive to the sun?
A 2019 review named amiodarone, chlorpromazine, doxycycline, hydrochlorothiazide, nalidixic acid, naproxen, piroxicam, tetracycline, thioridazine, vemurafenib, and voriconazole among the most consistently implicated. 83
How long should you stay out of the sun after taking doxycycline?
The US label gives no number of hours. It tells patients to avoid excessive sunlight while taking the drug, whose half-life is 18 to 22 hours. 70
Does hydrochlorothiazide cause skin cancer?
Its use is linked with more squamous cell carcinoma, and the FDA puts the added risk at about 1 case per 16,000 patients a year. The Danish study that raised the question had no data on sun exposure. 5; 73
Why did lime juice give me a rash in the sun?
Limes, like figs, celery, and giant hogweed, carry furocoumarins that UVA turns into chemicals that damage skin, with no allergy needed. 51; 53
Does the sun make poison ivy worse?
We found no study of sunlight and poison ivy rash.
Can you be allergic to sunscreen?
Yes. Sunscreen chemicals can cause contact allergy, which needs no light, and photoallergy, which does. Among people who suspected it, photopatch tests confirmed it in 26 of 40 in Toronto, and 8 of 11 in Portland tested negative. 7; 8
Does vitamin D deficiency cause sun allergy?
In Graz, 23 patients with polymorphic light eruption had lower vitamin D than 23 matched people in every season, and UVB lamp hardening that prevented the rash raised it. We found no trial of vitamin D pills to prevent the rash. 84
Can bright sunlight cause migraines?
The Sunlight Report's library holds no study of sunlight and migraine or of light sensitivity in the eyes.
What we do not know
Nobody has identified what light changes in the skin to set off polymorphic light eruption or solar urticaria. The 2004 London study found that patients' skin fails to damp its response to ultraviolet, and the 2022 review of solar urticaria still calls the molecule at the center of its leading model "putative". 17; 22
The library holds no primary study of how often doxycycline, hydrochlorothiazide, or voriconazole makes skin sensitive to light. The drug evidence here is single cases, clinic series, and labels, and the plant evidence is case reports, with no count of how often the reactions happen.
The sunscreen trials in polymorphic light eruption and lupus used lamps or deliberate heavy sun, applied the cream at 2 milligrams per square centimeter, and were mostly paid for by the companies whose products were tested. None followed people through ordinary summers.
Whether sunlight can start lupus rests on estimates of the ultraviolet at women's homes, since no one measured the sun that reached them. The surveys of polymorphic light eruption disagree on whether it rises with distance from the equator.
Key studies
- Rasch, 1926, Copenhagen and London. Lecture and clinic records. Home's 1820 hand experiment, 230 light-eruption cases among 31,000 skin patients, and 54 of 94 lupus cases starting from March to June. 10
- Ramsay, 1977, London. Fluorescent lamps and real sun. Hourly exposures built tolerance to solar urticaria in eight of nine patients. 23
- Kaidbey and Kligman, 1978, Philadelphia. Solar simulator on students' backs. A padimate A sunscreen caused sunburn-like reactions in 12 of 17. 76
- Walter and colleagues, 1984, San Diego. Solar simulator and a windshield filter. Amiodarone sensitivity to UVA from 350 to 380 nanometers. 1
- Ortel and colleagues, 1986, Vienna. Lamps. Polymorphic light eruption reproduced in 69 of 142 patients, by UVA alone in 56 percent. 4
- Seligman and colleagues, 1987, Oregon. Real sun and celery. Nineteen grocery workers burned after a celery sale. 52
- Lehmann and colleagues, 1990. Lamps. Lupus lesions provoked in 25 to 64 percent of 128 patients, depending on the form of lupus. 28
- Fischer and Bergström, 1991, Sweden. Photopatch tests of people who complained about sunscreens. Four of 27 reacted to a sunscreen filter. 77
- Pao and colleagues, 1994, Perth, Ballarat, and London. Questionnaire. Polymorphic light eruption in 5.2, 3.6, and 14.8 percent. 15
- Alora and Taylor, 1998, Boston. Lasers. Solar urticaria from red light only. 24
- Stege and colleagues, 2000, Düsseldorf. Lamps and sunscreens on 11 lupus patients. Three sunscreens protected 11, 5, and 3. 31
- Darvay and colleagues, 2001, London. Photopatch tests. Photoallergy in 62 of 2,715 patients. 44
- Moseley and colleagues, 2001, Dundee. Blue light. An SPF 25 sunscreen protected by a median factor of 2, a tinted cream by 8. 62
- van de Pas and colleagues, 2004, London. Solar simulator. Ultraviolet damped the immune response less in polymorphic light eruption. 17
- Matthieu and colleagues, 2004, Belgium, and Hindsén and colleagues, 2006, Malmö. Ketoprofen gel on skin, photopatch tests. Severe photoallergy to ketoprofen. 46; 47
- Schleyer and colleagues, 2008, Tübingen. Lamps and sunscreen. The rash on 10 of 10 unprotected arms and none of the protected ones. 18
- DeLeo and colleagues, 2009, Puerto Rico. Real sun and sunscreens. Ecamsule added protection against the rash. 19
- European photopatch taskforce, 2012. Photopatch tests at 30 centers. Ketoprofen the commonest photoallergen. 48
- Langendonk and colleagues, 2015. Randomized trials. Afamelanotide added pain-free hours in sunlight in EPP. 64
- Pedersen and colleagues, 2018, Denmark. Registry records. Hydrochlorothiazide and squamous cell carcinoma. 73
- Barman-Aksözen and colleagues, 2020, Zurich. Patients' reports. Time in sun before pain in EPP rose from 10 to 180 minutes on afamelanotide. 3
- Redgrave and Solomon, 2021, London. Real sun and fig sap. Blistering over 8 percent of a tree surgeon's body. 54
- Burfield and colleagues, 2023. Meta-analysis. Polymorphic light eruption in about 10 percent of 19,287 people. 2
- Barbhaiya and colleagues, 2023. Nurses' Health Studies. Ultraviolet at home was not associated with developing lupus. 38
- European photopatch study, 2026. Photopatch tests. A 24-hour wait missed 53 percent of sunscreen photoallergies. 79
Definitions
- Action spectrum. The wavelengths of light that bring on a given reaction.
- Afamelanotide. A synthetic analogue of alpha-melanocyte-stimulating hormone, given as an implant, that darkens the skin. It is approved for erythropoietic protoporphyria.
- Chronic actinic dermatitis. A long-lasting eczema of sun-exposed skin in people who react to very small doses of ultraviolet and sometimes visible light, usually in older people.
- Confidence interval. The range in which the true effect probably lies, given the size of the study.
- Contact allergy. An allergic skin reaction to a chemical touching the skin, with or without light.
- Erythropoietic protoporphyria (EPP). An inherited porphyria in which protoporphyrin IX builds up in red blood cells and skin, and visible light causes burning pain within minutes.
- Furocoumarins. Plant chemicals of the psoralen family, found in limes, figs, celery, parsnip, and hogweed, that UVA turns into agents that damage skin.
- Hardening. The skin's growing tolerance to light after repeated exposure, from the sun or from lamps given on purpose.
- Hazard ratio and odds ratio. How often one group gets a disease compared with another. 1.0 means no difference.
- Lupus erythematosus. An autoimmune disease. Discoid and subacute cutaneous lupus mainly affect the skin; systemic lupus can reach the joints, kidneys, and blood.
- Malar rash. A butterfly-shaped rash across the cheeks and nose, seen in systemic lupus.
- Mast cell. An immune cell in the skin that releases histamine and causes the wheals of hives.
- Meta-analysis. A study that pools the results of earlier studies into one estimate.
- Minimal erythema dose. The smallest dose of ultraviolet that leaves a patch of a person's skin just visibly red a day later.
- Nanometer. A billionth of a meter, the unit used for wavelengths of light. Ultraviolet runs from about 280 to 400 nanometers at the ground, and visible light from about 400 to 700.
- Persistent light reaction. A light sensitivity that continues for months or years after the chemical that caused it is gone.
- Photoallergy. An allergic reaction to a chemical that happens only where light falls on the skin carrying it.
- Photodermatosis. A skin disease caused or made worse by light.
- Photopatch test. A test in which suspect chemicals are taped to the back in two sets and one set is given UVA from a lamp.
- Phototoxicity. Damage to skin from a chemical activated by light, which happens in anyone with enough of the chemical and light, with no allergy.
- Phytophotodermatitis. A phototoxic burn from plant juice on the skin followed by sunlight.
- Polymorphic light eruption. An itchy rash that appears hours to days after sun, mostly in spring and summer, and clears in about a week out of the sun.
- Porphyria. Any of a group of disorders in the making of heme. Some affect the skin, and others, the acute porphyrias, mainly affect the nerves and gut.
- Protoporphyrin IX. A step on the way to heme. It absorbs light most strongly around 408 nanometers.
- Psoralen. A light-sensitizing plant chemical, used as a medicine with UVA in PUVA.
- PUVA. Psoralen taken by mouth or put on the skin, followed by UVA from a lamp.
- Randomized trial. A study that assigns people to groups by chance, so that the groups differ only in what the researchers gave them.
- Solar simulator. A lamp built to give off a spectrum close to that of sunlight.
- Solar urticaria. Hives raised by light, usually within minutes.
- UVA. Ultraviolet light from 315 to 400 nanometers. It passes through ordinary window glass.
- UVB. Ultraviolet light from 280 to 315 nanometers. It causes most sunburn, and ordinary window glass stops it.
- Visible light. Light the eye can see, from about 400 to 700 nanometers.
- Wheal. A raised, itchy welt of hives.
Statistics
Of 370 patients with suspected light-triggered dermatitis patch and photopatch tested in Toulouse from 1990 to 1994, 57 (15.4 percent) had contact allergy or photoallergy to sunscreen chemicals. 85
In the Toulouse series, 41 of the 57 positive sunscreen reactions (72 percent) were judged relevant to the patient's rash. 85
Testing 370 Toulouse patients with suspected photodermatitis produced 107 positive patch test reactions and 126 positive photopatch test reactions. 85
Of the Toulouse patients allergic or photoallergic to sunscreens, 13 also had polymorphic light eruption. 85
In the Toulouse series, 8 patients reacted to more than one sunscreen chemical. 85
Of 4,957 patients photopatch tested for suspected light-triggered rashes in Shanghai from 2006 to 2012, 2,454 (49.5 percent) had at least one photoallergic reaction. 86
In the Shanghai series, the antipsychotic drug chlorpromazine caused 1,768 of the 3,472 photoallergic reactions (50.9 percent). 86
PABA, the old sunscreen filter, gave a photoallergic reaction in 546 of 3,718 Shanghai patients tested with it (14.7 percent). 86
In the same Shanghai series, benzophenone gave a photoallergic reaction in 0.6 percent of patients tested. 86
Thimerosal, a mercury preservative, gave a photoallergic reaction in 387 of 4,331 Shanghai patients tested (8.9 percent). 86
Ordinary contact allergy without light was found in 399 of 4,957 Shanghai patients (8 percent), with 409 reactions. 86
In the Shanghai series, 33.5 percent of patients aged 11 to 20 had a photoallergic reaction, against 57.8 percent of those aged 51 to 60. 86
Of 6,153 patients photopatch tested at a Shanghai hospital from 2005 to 2014, those tested in summer had the highest photoallergic reaction rate, 63.74 percent, against 56.50 percent in winter. 87
In the Shanghai 10-year series, photoallergic reactions to PABA fell from 14.17 percent of patients in 2005 to 2009 to 11.09 percent in 2010 to 2014. 87
In the Shanghai 10-year series, benzophenones gave a photoallergic reaction in 35 of 5,654 patients tested (0.62 percent), against 601 of 5,032 for PABA (11.94 percent). 87
Of 6,153 Shanghai patients photopatch tested over 10 years, 3,668 (59.61 percent) had at least one photoallergic reaction. 87
Of 2,372 patients photopatch tested at a Shanghai referral centre from 2015 to 2019, 897 (37.8 percent) had a photoallergic reaction. 81
Of the 1,208 photoallergic reactions in the 2015 to 2019 Shanghai series, 348 (28.9 percent) were judged relevant to the patient's current rash and 795 (65.8 percent) were of unknown relevance. 81
The antipsychotic chlorpromazine gave a positive photopatch test in 23.8 percent of 2,365 Shanghai patients, and the reaction was judged relevant to the rash in 3.1 percent. 81
In the 2015 to 2019 Shanghai series, 1,479 of 2,372 patients (62.4 percent) had a phototoxic test reaction, and chlorpromazine caused 92.9 percent of those reactions. 81
The sunscreen filters benzophenone and PABA each gave a photoallergic reaction in under 1 percent of 2,365 Shanghai patients tested from 2015 to 2019 (10 and 15 patients). 81
In the Shanghai 2015 to 2019 series, photoallergy was found in 41.7 percent of men tested against 35.8 percent of women. 81
Of the 897 Shanghai patients with a photoallergic reaction, 42.3 percent already had a diagnosis of a light-sensitivity disease such as chronic actinic dermatitis, polymorphic light eruption, or solar urticaria, or of atopic dermatitis. 81
Of 220 Shanghai patients with chronic actinic dermatitis, 101 (45.9 percent) had a photoallergic reaction on testing, against 172 of 446 with polymorphic light eruption (38.6 percent) and 13 of 33 with solar urticaria (39.4 percent). 81
Of 100 patients photopatch tested for suspected light-triggered allergy in Bogota from 2011 to 2013, 15 had a photoallergic reaction. 80
In the Bogota series, 8 of the 15 photoallergic patients (53.3 percent) showed their reaction only at the day-6 reading, missed at day 4. 80
In the Bogota series, UV filters also caused 53 percent of the ordinary contact allergy reactions, found in 32 of the 100 patients. 80
Of 70 patients in New Delhi with rashes on sun-exposed skin, 2 had a photoallergy to a standard commercial photopatch series. 82
In the New Delhi series, the weed parthenium was the commonest pattern, in 32 of 70 patients, with photoallergy to it in 4 and light-worsened allergy in 6. 82
Of 83 patients photopatch tested in Coimbra, Portugal from 2003 to 2007, 36 (43.3 percent) had at least one positive photopatch test. 49
In the Coimbra series, drugs (phenothiazines and anti-inflammatory painkillers) caused photosensitivity in 29 of the 36 reactive patients (80.5 percent). 49
In the Coimbra series, 21 of 30 patients tested with anti-inflammatory painkillers (70 percent) had a relevant photoallergy to one. 49
In the Coimbra series, UV filters caused photoallergy in 10 of 83 patients tested (12.0 percent). 49
In Coimbra, 7 patients had photoallergy to benzydamine from gargling an anti-inflammatory mouth rinse, with rashes mainly on the chin and lower lip. 49
Of 116 patients photopatch tested with the European baseline series in Valencia from 2014 to 2016, 25 (21.5 percent) had 55 photoallergic reactions. 88
In the Valencia series, anti-inflammatory painkiller gels caused 41 of the 55 photoallergic reactions, ketoprofen and dexketoprofen 16 each. 88
In the Valencia series, 14 of 116 patients (12.1 percent) had ordinary contact allergy without light, with 21 reactions. 88
Of the positive reactions in the Valencia series, 51.3 percent were relevant to the patient's current rash. 88
Of 160 patients photopatch tested at Toronto Western Hospital from 2001 to 2010, 54 (33.8 percent) had at least one photoallergic reaction. 7
In the Toronto series, 59 of 160 patients (36.9 percent) reacted to at least one sunscreen ingredient or sunscreen product. 7
Of 40 Toronto patients who came in describing a sunscreen allergy, 26 (65 percent) tested positive to their own sunscreens or sunscreen components. 7
In the Toronto series, 17 of 37 men (45.9 percent) had a photoallergic reaction, against 37 of 123 women (30.1 percent). 7
Of the 160 Toronto patients, 72 were referred for rash, swelling, or itch after sun exposure and 30 for a light-distributed rash after lupus and porphyria had been ruled out. 7
Of 27 people in Portland, Oregon who said they were allergic to sunscreen, 11 agreed to full patch and photopatch testing, and 8 of those 11 tested negative. 8
Of the 11 Portland patients who believed they were sunscreen-allergic, 9 had a history of atopy such as hay fever, asthma, or eczema. 8
None of the 11 Portland patients who believed they were sunscreen-allergic reacted to the newer filters Tinosorb S and Tinosorb M or to Anthelios SX. 8
Of 207 patients tested for suspected light-triggered rashes in Athens from 1992 to 2006, 28 (13.52 percent) had a photoallergy and 86 (41.54 percent) an ordinary contact allergy. 89
In the Athens series, the antihistamine promethazine caused 25 percent of the 40 photoallergic reactions, and chlorpromazine 12.5 percent. 89
Of the 207 Athens patients tested for suspected light-triggered rashes, 93 (44.92 percent) reacted to nothing on either patch or photopatch testing. 89
Of the 62 London patients found photoallergic among 2,715 tested at St John's from 1983 to 1998, 34 (55 percent) already had a light-sensitivity disease: 19 chronic actinic dermatitis and 15 polymorphic light eruption. 44
Of 80 photoallergic reactions in the St John's series, drugs caused 16 (20 percent) and the old perfume musk ambrette 11 (14 percent). 44
In 12 years of photopatch testing at St John's, the UVB filter octyl methoxycinnamate produced two photoallergic reactions. 44
At St John's, 49 of 2,715 patients (1.8 percent) had ordinary contact allergy on the photopatch series, with 51 of their 75 reactions to UV filters. 44
Of 58 Swedish pharmacy customers who complained of skin problems from a sunscreen in 1989, 35 came in for testing and 27 were photopatch tested. 77
Among the complaining Swedish customers, 20 had symptoms immediately after applying the sunscreen and 11 only minutes after sun exposure. 77
Of the complaining Swedish customers, 20 had short-lived stinging or burning and 26 had longer-lasting itchy, swollen rashes followed by scaling. 77
Among the complaining Swedish customers, complaints per unit of sunscreen sold rose as the product's SPF rose. 77
Of 219 North American patients whose allergy was traced to a sunscreen from 2001 to 2010, 61.2 percent reacted to at least one allergen missing from the group's standard screening series. 78
Among the 219 North American patients with a sunscreen allergy, the final diagnosis was allergic contact dermatitis in 59.4 percent and photodermatitis in 20.1 percent. 78
After oxybenzone, the commonest sunscreen allergens in the North American series were inactive ingredients: vitamin E (DL-alpha-tocopherol) at 4.8 percent and fragrance mix I at 4.0 percent. 78
Of 23,908 North American patients patch tested from 2001 to 2010, 32 (0.1 percent) had a sunscreen coded as the source of an irritant, not allergic, reaction. 78
Fewer than 40 percent of positive patch test reactions in North American sunscreen-allergic patients were caught by the group's standard series of 65 to 70 allergens. 78
Of 1,527 patients patch tested at a Vancouver clinic, 23 were tested with a sunscreen series, and 4 of them reacted to a sunscreen chemical or their own sunscreen. 90
Across all 1,527 patients patch tested in Vancouver, 12 (0.8 percent) had a positive reaction to a sunscreen chemical or sunscreen product. 90
Of 11 patients with actinic prurigo, a sun-triggered itchy rash, seen at Singapore's National Skin Centre from 1990 to 1998, 10 were men, and all fell ill as adults, unlike the young American Indian girls in earlier reports. 91
Of the 11 Singapore actinic prurigo patients, 4 had noticed that sun made their rash worse. 91
Phototesting of the 11 Singapore actinic prurigo patients found a lowered burn threshold to both UVA and UVB in 4 and to UVA alone in 2. 91
In the Singapore actinic prurigo series, the rash sat on the backs of the hands in 10 of 11 patients and on the forearms in 9. 91
Among 167 Vienna patients with polymorphic light eruption, women outnumbered men 4.7 to 1. 4
Repeated lamp exposures reproduced the rash in 69 of 142 Vienna patients with polymorphic light eruption, after two to eight exposures. 4
In the Vienna patients whose rash was reproduced under a lamp, UVA alone triggered it in 56 percent, UVB alone in 17 percent, and both in 26 percent. 4
Of 51 Vienna patients followed up after a preventive course of psoralen plus UVA lamp treatment, 64 percent reported full protection from polymorphic light eruption that summer, 26 percent partial, and 10 percent none. 4
Of 122 Vienna patients given psoralen plus UVA to prevent polymorphic light eruption, 27 (22 percent) broke out in the rash after the first treatments. 4
In a trial of 144 people with polymorphic light eruption given up to 6 days of deliberate outdoor sun, a four-filter SPF 40 sunscreen beat the same cream without ecamsule in 41 of 73 participants (56 percent), against 8 of 73 (11 percent) the other way. 19
In the same outdoor trial, 75.4 percent and 83.1 percent of participants in the two groups still developed polymorphic light eruption on both sunscreen-treated sides of the body. 19
Of the 144 people with polymorphic light eruption in the outdoor sunscreen trial, 22 never developed the rash on either side despite up to 6 days of increasing sun doses. 19
Of the 144 participants in the polymorphic light eruption sunscreen trial, 81.9 percent were women. 19
In 14 patients with polymorphic light eruption given four days of low solar-simulator doses, an after-sun lotion with DNA-repair enzymes cut the rash score by 61 percent against untreated skin (1.73 against 4.39). 20
In the same 14 patients, an SPF 30 sunscreen applied before the solar-simulator doses prevented polymorphic light eruption completely in every patient. 20
In 9 London patients with solar urticaria (hives from light), repeated hourly lamp exposures made a patch of skin stop reacting after two to five exposures in all but one patient. 23
Lamp-induced tolerance in the London solar urticaria patients wore off 48 to 72 hours after the last exposure unless exposures were repeated every 24 hours. 23
Of the 9 London solar urticaria patients treated by building tolerance with lamps or sunlight, 2 had complete success, 6 partial success, and 1 failed. 23
Two of the 9 London solar urticaria patients had collapsed with low blood pressure after sun on a beach and had been rescued from the sea semiconscious. 23
None of the 9 London solar urticaria patients went into spontaneous remission during up to two years of follow-up. 23
A 24-year-old Vancouver dental hygienist with solar urticaria broke out in hives under a UVA lamp at 4.4 joules per square centimetre and under visible light at 0.75. 26
After five plasma exchanges over 10 days, the Vancouver solar urticaria patient was almost symptom-free more than 21 months later. 26
A 28-year-old Boston woman with solar urticaria got itchy welts 3 to 5 minutes after sun exposure, and they cleared within 20 minutes of taking cover. 24
The Boston solar urticaria patient did not react to UVA, UVB, or a slide projector, but welted within 3 minutes under red laser light at 610 to 690 nanometres. 24
Of 10 London patients with dermatomyositis, a muscle and skin autoimmune disease, 5 reported light sensitivity, and monochromator testing found a lowered burn threshold in 2. 92
Solar-simulator light on unaffected skin reproduced a dermatomyositis lesion in 1 of 6 London patients tested. 92
In the London dermatomyositis group, a rash came first in 7 of 10 patients and preceded muscle weakness by a mean of 5.9 months. 92
In Sofia, 9 of 19 patients with dermatomyositis (47.4 percent) burned at a lower UVB dose than normal, against 5 of 26 healthy controls (19.2 percent). 93
In the Sofia study, 14 of 30 lupus patients (46.7 percent) showed increased UVB sensitivity on lamp testing, including 6 of 11 with discoid lupus and 2 of 6 with subacute cutaneous lupus. 93
Of 19 Sofia dermatomyositis patients, 7 (36.8 percent) reported increased sun sensitivity and 4 (21.1 percent) said sun worsened their rash. 93
In a 38-year-old Yokohama man with discoid lupus, UVB lamp test doses above his burn threshold left scaly lupus-like red patches that lasted at least 2 months, but the same doses 7 and 12 months later produced none. 94
In 11 Düsseldorf lupus patients whose skin lesions were provoked by UVA plus UVB lamps, one sunscreen with several UVA filters prevented lesions in all 11, and two others protected 5 and 3. 31
In 2 of 5 infants with neonatal lupus described in 1983, the rash did not appear until the babies had been out in direct sunlight. 95
A 34-year-old Würzburg woman whose systemic lupus had been in remission for more than 8 years without treatment developed lupus kidney disease 6 weeks after a sunburn from sunbathing. 33
At diagnosis after the sunbathing holiday, the Würzburg patient's anti-double-stranded DNA antibody level was 197 IU/mL against a normal of under 10, and she lost 5,390 mg of protein a day in her urine. 33
Of 19 Tokyo patients with atopic dermatitis, all had a normal UVB burn threshold, but 3 (15.7 percent) developed abnormal rashes 24 to 72 hours after repeated UVB lamp doses. 96
None of the 19 Tokyo atopic dermatitis patients reacted to pure UVA lamp doses of up to 9 joules per square centimetre. 96
In 7 Dundee patients sensitive to blue light (4 with erythropoietic protoporphyria, 2 with chronic actinic dermatitis, 1 with solar urticaria), skin reddened at 4.7 to 15 joules per square centimetre of 430-nanometre light, against more than 82 for normal skin. 62
On the 7 blue-light-sensitive Dundee patients, an ordinary SPF 25 sunscreen gave a median protection factor of 2 against blue light, and the new Dundee reflectant sunscreens a median of 8. 62
On 10 Zurich volunteers made light-sensitive with a topical photodynamic drug, a dark tinted cover cream protected the skin 3.4 times better than a standard sunscreen. 97
All 10 patients given the heart-rhythm drug amiodarone (400 to 800 mg daily) at the University of California, San Diego had photosensitivity reactions, and 4 of them had considerable redness, swelling, and burning. 1
One week after starting amiodarone, a 65-year-old San Diego man felt stinging and burning after 5 to 10 minutes of summer noon sun; in Hawaii three months later, after 2 minutes. 1
The amiodarone patient reddened after 4 to 8 joules per square centimetre of UVA, where normal skin needs 30 or more. 1
The amiodarone patient's UVA sensitivity faded after he stopped the drug and was gone by ten weeks. 1
With a sunscreen, the amiodarone patient could stay in sun about 30 minutes before severe stinging, instead of 5 to 10. 1
A Miami man receiving the cancer drug vinblastine burned under a lamp in 5 to 10 seconds, where normal controls took 20 to 30 seconds. 68
All 5 healthy controls injected with a small dose of vinblastine in the skin reacted to a lamp dose too low to redden normal skin. 68
Five Busan patients who soaked their feet in a fig-leaf folk remedy and then went out in sun all developed burning, red, swollen feet within 24 hours, sparing only skin under the sandal straps. 55
In the first Busan fig-leaf case, a 57-year-old woman felt burning in her hands and feet 5 hours after walking in the sun. 55
One Busan fig-leaf patient, who soaked her feet three times a day for two days before going out in the sun, took two months to heal. 55
A 26-year-old Texas woman developed a sharply bordered rash on her abdomen and thighs about 7 hours after squeezing 2 dozen limes for margaritas at a pool party. 53
A 46-year-old London tree surgeon who pruned a fig tree in hot sun had red, blistering skin covering about 8 percent of his body surface two days later. 54
In 39 Swiss patients with erythropoietic protoporphyria, an inherited disease in which visible light burns the skin, the longest time in direct sun without a painful reaction was a median of 10 minutes before treatment. 3
On the drug afamelanotide, the Swiss protoporphyria patients' median sun tolerance rose from 10 to 180 minutes, an individual gain of 1.8-fold to 180-fold. 3
The Swiss protoporphyria patients rated the pain of their worst light reaction a median 10 out of 10 before treatment and 6 under afamelanotide. 3
In 23 Lausanne patients given the photodynamic drug mTHPC, a typical dose kept the skin measurably light-sensitive under a solar simulator for up to about 6 weeks. 98
For mTHPC patients in Lausanne, skin light sensitivity peaked 2 to 6 days after injection and direct sun had to be strictly avoided for about 2 weeks. 98
In one Lausanne patient given both drugs three years apart, Photofrin kept the skin light-sensitive for more than 13 weeks, while mTHPC left no effect at 6 weeks. 98
Of 16 healthy Beijing volunteers given the photodynamic drug hemoporfin, 8 had a forearm patch exposed to 20 minutes of noon sun 52 hours later, and none showed any skin reaction. 99
Of 20 Belgian patients tested for suspected dermatitis from ketoprofen painkiller gel in 2001 and 2002, 17 had a photoallergy to it, and 47 percent had symptoms severe enough to need oral or injected steroids. 46
Of the 20 Belgian ketoprofen patients, 5 were hospitalized and 1 stayed light-sensitive long after the reaction. 46
Of 35 patients in southern Sweden with photoallergy to ketoprofen gel, 73 percent were also photoallergic to the cholesterol drug fenofibrate. 47
In a Besançon case-control study, 12 HIV-infected patients, most taking potentially photosensitizing drugs, burned at the same solar-simulator dose as 12 healthy volunteers. 100
In 4 patients with severe solar urticaria, eosinophil white cells appeared in skin vessels 5 minutes after solar-simulator light, with marked eosinophil infiltration of the skin 2 hours later. 101
In a Procter and Gamble guinea pig test, PABA applied to the skin with UVA made 9 of 19 animals photoallergic, against 1 of 19 challenged with PABA without light. 102
In the same guinea pig test, the old perfume ingredient musk ambrette with UVA made 12 of 20 animals photoallergic, and 0 of 20 reacted to it without UVA. 102