Melanin

In brief

What melanin is, and why it has no formula

Most entries in this Atlas can open with a formula. Melanin cannot, and the reason is the whole subject. It is made by taking one amino acid, tyrosine, oxidizing it with an enzyme, and letting the reactive products attack one another until they have joined into an irregular, insoluble mass. The mass is not one compound. It is a polymer, a chain of repeating units, but the units are not all the same, they are joined at different positions and in different oxidation states, and no two preparations have quite the same size or composition. The standard chemistry review from 2009 says plainly that the molecular structure of melanins, the mechanism by which the pigment is produced, and how it is organized within the melanosome remain to be fully understood, and that most natural melanins are mixtures of two different pigments ². Treating melanin as a process, a thing cells do, is therefore more honest than treating it as a thing cells contain.

The two pigments are eumelanin and pheomelanin. Eumelanin is the brown-to-black one, insoluble in nearly everything, built from indole units, which are a benzene ring fused to a five-membered nitrogen ring. Pheomelanin is yellow to reddish brown, dissolves in alkali, and contains sulfur because its units, benzothiazines, are built with the amino acid cysteine bolted in. Both come from the same first step: tyrosinase, a copper enzyme, oxidizes tyrosine to dopaquinone, a highly reactive ortho-quinone. What happens next depends on the company dopaquinone keeps. With no thiol nearby it folds on itself to cyclodopa, is oxidized to dopachrome, and rearranges to 5,6-dihydroxyindole and, less often, 5,6-dihydroxyindole-2-carboxylic acid, in a ratio set by a second enzyme, dopachrome tautomerase; the indoles then oxidize and polymerize into eumelanin. If cysteine is present, dopaquinone grabs it first, making cysteinyldopa isomers that go on to pheomelanin. In pulse radiolysis experiments the mixed process ran in three stages, cysteinyldopa first, pheomelanin second, eumelanin last, and the authors proposed a casing model in which a pheomelanin core is buried under a eumelanin surface ¹.

That order matters more than it looks. If the casing model is right, even a dark granule would carry a pheomelanin core, and the amount of pheomelanin in a person's hair or skin would not simply be the inverse of their eumelanin; the model is inferred from reaction kinetics in a test tube rather than seen in granules ¹. The quantity involved is in any case small. Measuring melanin taken from the skin of African, Indian, Mexican, Chinese and European volunteers in 2002, Alaluf and colleagues found that spheroidal, pheomelanin-type melanosomes made up less than 10 percent in every group, and concluded that pheomelanin is a very minor component of human epidermal melanin even in the lightest skin; what distinguished the lighter skin types was about half as much epidermal melanin, up to three times as much of the pale, alkali-soluble fraction, and progressively smaller melanosomes ⁴¹. A 2021 method that feeds melanocytes tyrosine labeled with heavy carbon and traces where the atoms go confirmed the known differences between melanocytes from people of different skin colors and different MC1R genotypes, and found that different ways of changing the acidity inside the melanosome drive new eumelanin and new pheomelanin differently; it also noted that the traditional way of measuring melanin, breaking it down chemically and quantifying the fragments, cannot tell synthesis from degradation ⁴².

There is a third pigment that shares the name and little else. Neuromelanin is the dark material that accumulates with age in the dopamine neurons of the substantia nigra, the midbrain region whose loss causes Parkinson's disease, and in the norepinephrine neurons of the locus coeruleus in the brainstem. It is not made by tyrosinase in a melanosome. It forms when dopamine or norepinephrine oxidizes inside the neuron and the products, with proteins and lipids the cell could not degrade, are swept into an autophagic organelle that then never empties; it increases linearly with age and is lost when the neurons die ^25,43. Elemental mapping of human locus coeruleus in 2026 identified the pigment compartment of these organelles by its sulfur-rich, pheomelanin-like portion, so even the brain pigment is a mixture ⁴⁴. It has its own section below.

The whole history, including the reversals

Melanin's history runs on two tracks that only recently met: a chemistry track, which worked out how the pigment is built, and a medical track, which kept making claims about what the pigment does and then having to check them. The timeline below carries both. The reversals are marked, because they are the sentences this page exists for: that melanin only protects (2015), that melanoma in red-haired skin needs sunlight (2012), that the low vitamin D of dark-skinned Americans was not true deficiency because their binding protein was low (2013, undone 2015-2016), that dark skin needs several times more sun to make vitamin D (1982, revised 2020), that a pulse oximeter reads skin as neutral (2005, 2020), and that the Fitzpatrick scale measures skin color (1988, challenged 2013-2026).

From tyrosine to pigment: enzymes, organelle, handover

Three related enzymes do the work, and they are related the way cousins are: alike in shape, different in what they hold. Tyrosinase carries two copper ions and performs both the first oxidation of tyrosine and the oxidation of dopa. Tyrosinase-related protein 1 looked like a second tyrosinase for decades, until its crystal structure in 2017 showed two zinc ions in the active site instead of copper, which explains why it has no tyrosinase activity; the structures also suggest that most albinism mutations in it affect the protein's stability or activity ⁴⁵. The third, dopachrome tautomerase, sets the ratio of the two indoles that eumelanin is built from ¹. All three are glycoproteins made in the endoplasmic reticulum and delivered to the melanosome, and the delivery is most of the story.

The melanosome is a lysosome-related organelle: a compartment that shares machinery with the cell's digestive lysosomes but is built, from early endosomes, to make and store pigment rather than to break things down. A 2007 review describes its formation as a novel pathway from endosomal intermediates in which a protein scaffold is laid down first and the enzymes arrive afterward, so that pigment is deposited on fibrils in stages; the inherited diseases that disrupt it, such as Hermansky-Pudlak syndrome, have shed light on the machinery that controls specialized endosomal sorting ⁵. Because the pigment is made inside a sealed compartment, the acidity of that compartment is a control: in the 2021 tracing study, different ways of shifting melanosomal pH drove new eumelanin and new pheomelanin differently ⁴².

The finished melanosome does not stay where it was made. Each melanocyte sits in the basal layer of the epidermis and sends dendrites up between the keratinocytes around it, and the pigment ends up in those keratinocytes, which do not make any. This unit, one melanocyte and the keratinocytes it supplies, was named the epidermal melanin unit, and a 2007 review argued it should include the Langerhans immune cells too, since inflammatory signals act on all three ⁷⁸. How the handover happens has been argued over since the 1960s. The candidate routes are exocytosis of the pigment core followed by phagocytosis by the keratinocyte, shedding of pigment-filled vesicles from the dendrite tip, direct fusion of the two cells' membranes, and the keratinocyte biting off the dendrite tip; a 2021 review finds the recent evidence favors the first two, proposes that different routes may run under different conditions, and calls what the keratinocyte then does with the pigment ill-characterized ⁶.

What the keratinocyte does with it is visible under the electron microscope and is half of what we call skin color. In 1969 Szabó and Fitzpatrick's group reported that melanosomes meet different fates in the epidermis of different races ⁵⁷. The measured version came in 2003 from photoprotected forearm skin: melanosomes in keratinocytes of dark skin were distributed singly in 88.9 percent of cases, in light skin in membrane-bound clusters in 84.5 percent, and in Asian skin 62.6 percent single to 37.4 percent clustered; mean melanosome area was 1.44 in dark, 1.36 in Asian and 0.94 in Caucasian skin, in hundredths of a square micrometer, and the single ones were the larger ones ⁷. The 1979 transmission study had already concluded that the superior protection of Black epidermis came from packaging and distribution as well as from melanin content ¹⁰. The number of melanocytes is not where the difference lies, and that has been measured directly: in 2005 Tadokoro and colleagues found melanocyte density at the junction of epidermis and dermis remarkably similar across racial groups, and unchanged in the week after a single minimal erythema dose of ultraviolet light; what changed was the expression of the pigment enzymes and, most of all, the movement of melanin from the lower layer up into the middle layer of the epidermis, more dramatically in darker skin ⁷⁹. The amount, the type and the arrangement of what the melanocytes make is where the difference lies.

What sets the color: MC1R, red hair, and the genes of light skin

If one gene had to stand for pigmentation it would be MC1R, the melanocortin 1 receptor, a seven-pass receptor on the melanocyte surface. Its hormone, alpha-melanocyte-stimulating hormone, raises cyclic AMP inside the cell and drives the machinery toward eumelanin; a receptor that signals weakly leaves dopaquinone to take the cysteine fork, and the result is pheomelanin. In 1995 Valverde and colleagues in Newcastle sequenced the gene in British volunteers and found variants in over 80 percent of those with red hair or fair skin that tans poorly, in fewer than 20 percent of those with brown or black hair, and in less than 4 percent of good tanners ⁸. A 2008 meta-analysis of 11 melanoma studies and 9 phenotype studies found seven variants associated with melanoma, with odds ratios from 1.42 (95 percent confidence interval 1.09 to 1.85) for R163Q to 2.45 (1.32 to 4.55) for I155T, and found R151C carried an odds ratio of 8.10 (5.82 to 11.28) for red hair; some variants were tied to melanoma but not to red hair, which the authors took as evidence for a non-pigmentary route to cancer as well as the obvious one ⁹. That hint is taken up in the next section.

The receptor is old and has been lost more than once. Lalueza-Fox and colleagues amplified a fragment of MC1R from two Neanderthal specimens and found a mutation not present in about 3,700 modern humans; in cell assays it reduced the receptor's activity to a level that alters hair or skin color, so some Neanderthals were probably light-skinned or red-haired, and inactive variants evolved independently in the two lineages ⁶⁷. The hormone that works the receptor is made, surprisingly, by the keratinocytes, not the melanocytes. In 2007 Cui and colleagues showed that after ultraviolet light the DNA-damage protein p53 directly switches on the pro-opiomelanocortin gene in keratinocytes, whose protein is cut into alpha-melanocyte-stimulating hormone and also beta-endorphin; mice lacking p53 do not tan, and several conditions of darkening without sunlight turned out to involve p53 mimicking the tan ⁴⁶. A tan is therefore a wound response: the keratinocyte reports DNA damage and orders more pigment from its neighbor. Skin can be made to darken without the damage: small molecules that inhibit salt-inducible kinase, designed to cross human epidermis, raised the pigment master regulator MITF and produced eumelanin in red-hair mice and in normal human skin explants, though no clinical trial has followed ⁸¹.

MC1R is not the whole of skin color; it is the whole of red hair. The genes of light skin were found by a fish. The golden zebrafish mutant has melanosomes that are fewer, smaller and less dense, exactly the change seen in light human skin, and the gene, SLC24A5, encodes a cation exchanger on an internal membrane, probably the melanosome itself; the human version restores color to the fish, its ancestral allele predominates in African and East Asian populations, and the variant allele is nearly fixed in Europeans and tracks lighter skin in people of mixed ancestry ⁶⁴. That was 2005, and it explained part of European lightening. African populations, whose pigment genetics had been little studied, were examined in 2017: genome-wide association in ethnically diverse African populations found variants in and near several genes, showed that a light variant at one locus entered East Africa by gene flow from non-Africans, and found that at the other loci the dark-pigmentation variants are identical by descent in South Asians and Australo-Melanesians, meaning they are ancient and shared, not separately evolved ⁶⁵.

Why the cline exists at all is a different question, and the standard answer is Jablonski and Chaplin's. Matching skin reflectance of indigenous peoples to satellite ultraviolet data, they found reflectance strongly correlated with latitude and ultraviolet level, found females lighter than males in every population for which both were measured, and argued that the gradient is a compromise: dark skin protects folate, which ultraviolet destroys and which the developing neural tube and sperm production need, while light skin at high latitude permits vitamin D synthesis ⁵⁶. It is a correlational argument, and the folate half of it rests on older laboratory work the paper cites rather than on anything measured in populations; it is the best-supported account, not a proven one. One last consequence of the receptor sits far from the skin. Redheaded women, nine of ten of them carrying two MC1R mutations, needed 6.2 volume percent of the inhaled anesthetic desflurane to stop moving to a painful stimulus, against 5.2 percent in dark-haired women, which the authors present as a genotype linked to inhaled anesthetic requirement ⁴⁷.

The 2012 mouse: what pheomelanin does, and does not, prove for people

The headline experiment is clean. Mitra and colleagues in David Fisher's laboratory took mice carrying an inactivating Mc1r mutation, whose yellow coats and pheomelanin-rich skin stand in for red-haired, fair-skinned people, and gave their melanocytes the commonest human melanoma oncogene, BRAF V600E. Kept away from ultraviolet light, the mice developed invasive melanomas at high incidence. The control was the point: adding an albino allele, which abolishes all pigment synthesis on the same red-hair background, was protective, and ordinary red-hair skin carried significantly more oxidative damage to DNA and lipids than albino red-hair skin. The pheomelanin pathway itself, the authors concluded, makes an ultraviolet-independent contribution to melanoma through oxidative damage ³⁹. The chemistry that would do it had been proposed already and was reviewed in 2014: pheomelanin can set up a pro-oxidant state without light, depleting glutathione, the cell's main antioxidant, in an oxygen-dependent way, and has been proposed as a 'living' polymer and biocatalyst that may keep growing and may trigger its own oxidation ⁴⁹.

What the mouse does not prove is that people with red hair get melanoma without sunlight. The animals had the oncogene switched on in their melanocytes by tamoxifen painted on the shaved skin of the back at 6 to 10 weeks of age, a melanocyte-targeted oncogenic hit that no person has, and the tumors arose mostly in the treated skin of the back; mice also have no human skin thickness, no human sun history and no human immune surveillance of pigment cells. The human evidence is of a different kind and points in the same direction. In sequenced human melanomas, carrying even one disruptive MC1R allele was associated with an estimated 42 percent more C-to-T mutations, the signature of ultraviolet damage, with a wide confidence interval of 15 to 76 percent, which the authors compared to the burden added by 21 extra years of age; carriers also had more mutations of classes that are not the ultraviolet signature, which supports additional mutagenic processes in red-haired skin ⁴⁰. That is an association in tumors that had, mostly, seen sun; it says red-haired skin is damaged more per unit of light and perhaps also without light, not that light is unnecessary. The 2008 meta-analysis points the same way from the other end: some MC1R variants raised melanoma risk without being tied to red hair at all, which is hard to explain by sunburn alone ⁹.

Then eumelanin lost its innocence too. In 2015 Premi, Brash and colleagues at Yale reported that in melanocytes the cyclobutane pyrimidine dimers that sunlight stamps into DNA within picoseconds of a photon's arrival went on forming for more than 3 hours after a dose of ultraviolet A ended. These dark dimers were the majority of all dimers and included the cytosine-containing ones that start the C-to-T mutations of melanoma. They arise when ultraviolet-triggered reactive oxygen and nitrogen species combine to excite an electron in a fragment of melanin into a triplet state carrying the energy of an ultraviolet photon, which is then handed to DNA by energy transfer, with no light involved. Pheomelanin did it more, but eumelanin did it. Melanin, the paper concluded, may be carcinogenic as well as protective against cancer ¹². The work was done in cultured cells and in mice, and no one has measured dark dimers in a living person's skin after a day in the sun.

The idea has grown rather than faded. A 2025 review of ultraviolet DNA photochemistry describes the chain in detail: ultraviolet light switches on two enzymes in the melanocyte, nitric oxide synthase, which makes the small signaling gas nitric oxide, and NADPH oxidase, which deliberately makes superoxide, a reactive form of oxygen (NADPH is the cell's electron-carrying cofactor); the two products combine into peroxynitrite, a strong oxidant, and that oxidant drives the chemiexcitation ⁸². The original group argued in 2026 that the action spectrum for dimer formation, the curve of which wavelengths do the damage, is no longer the absorption spectrum of DNA alone, because redox chemistry extends it into the ultraviolet A and into the hours after exposure, so that sunburn is an incomplete surrogate for mutagenic risk and sunscreens should be judged on indirect damage as well ⁸³, and proposed inhibiting nitric oxide synthase or quenching triplet states as interventions while conceding that how pigment oxidation state and enzyme flux coordinate in living tissue is unknown ⁸⁴. Those are the hypothesis's authors; the independent chemistry is more modest, and a 2026 infrared study of synthetic pigments bleached under light found pheomelanin shifting from benzothiazine to benzothiazole rings and losing its internal interactions while eumelanin became more conjugated, which is consistent with the two polymers degrading differently but is model chemistry, not skin ⁸⁵.

The honest summary is this. In mice, pheomelanin causes melanoma without light; in people, MC1R variants raise melanoma risk and mutation load by routes that are at least partly independent of pigmentation, and the chemistry for a light-independent route exists. No study has shown a human melanoma arising from pheomelanin in the dark, and sun protection remains the only intervention with evidence. What the mouse changed is the frame: a red-haired patient's risk is not entirely in their hands, and a dark tan is not a neutral shield.

Two 2026 papers argue the other way, and both are animal or test-tube work. Blocking pheomelanin synthesis pharmacologically in male zebra finches, by inhibiting the enzyme that strips a fatty acid off MC1R and so leaving the receptor signaling, raised systemic oxidative damage in the birds whose follicles make pheomelanin and not in those whose follicles make eumelanin, which suggests the pigment exists partly to dispose of cysteine that the cell cannot otherwise hold safely ⁸⁶. And melanin bound to nucleic acids under physiological conditions protected them rather than damaging them: in complexes made from chicken follicular melanocytes, eumelanin shielded DNA from strand breaks and from oxidative modification, and pheomelanin shielded RNA ⁸⁷. In people the natural experiment runs the other way from the mouse: among 123,179 patients with vitiligo matched to as many comparators in a United States network of health records, losing melanocytes went with less skin cancer rather than more (hazard ratio 0.83, 95 percent confidence interval 0.78 to 0.88), with the largest reduction for melanoma (0.69, 0.58 to 0.81), although an exploratory subgroup suggested more squamous cell carcinoma in Black patients with vitiligo (2.16, 1.09 to 4.28) ⁸⁸. That is an association in a cohort drawn from records, and people with vitiligo are watched, treated and counseled about the sun more than anyone; it does not overturn the mouse, and it is the most direct human test of the pigment-and-melanoma question yet published.

Melanin as sunscreen: the measured equivalent and the cancer rates

The number everyone wants is a sun protection factor for dark skin, and the truthful answer is that the best measurement gave a ratio rather than a factor. Kaidbey and colleagues in 1979 measured ultraviolet transmission through skin samples from Black and White donors two ways, biologically, by the erythema and the phototoxic response the transmitted light produced, and spectroscopically. On average, five times as much ultraviolet, both A and B, reached the upper dermis of Caucasian skin as of Black skin. The difference at the stratum corneum, the dead outer layer, was far smaller: in White skin the stratum corneum was the main filter, in Black skin the living layers beneath it. Melanin behaved as a neutral density filter, attenuating all wavelengths about equally, and the authors credited the extra protection to melanosome packaging and distribution as well as to melanin content ¹⁰. Fivefold attenuation of the dose reaching the dermis is the measured equivalent; it is not the same as a factor on a sunscreen bottle, which is defined by erythema on the surface, and the abstract gives no such factor. A sun protection factor has nonetheless been estimated for dark skin, and it is small: the 2012 Nature paper that put pheomelanin in the dock gives the range as SPF 2.0 to 4.0, against the SPF 20 to 40 of a sunscreen, and notes in the same breath that the evidence for sunscreen against melanoma, unlike against squamous cell carcinoma, is weak ³⁹. A fivefold cut in the dose reaching the dermis and a sun protection factor of two to four are the same fact measured two ways, and neither is permission to sit in the sun.

What the filter does to DNA was measured directly in 2003. Volunteers of different ethnic groups had their minimal erythema dose determined, the smallest dose that reddens the skin, and were biopsied before and 7 minutes, 1 day and 1 week after one such dose. Every group showed significant DNA damage; the amount fell as constitutive melanin rose; the variation between individuals was large; and the erythema dose predicted damage better than ethnic origin or skin phototype did, with the rate of repair looking like the real determinant of sensitivity ¹¹. Darker skin burns at a higher dose, and at that higher dose it is still damaged. The National Cancer Institute review of 2008 lists melanin's broadband absorption and radical scavenging and the lower skin-cancer incidence in darker skin across many epidemiological studies, and still calls the role of melanin in photoprotection controversial ⁴⁸; the dark-dimer finding seven years later gave a further reason for that caution ¹².

The cancer rates are the strongest human evidence that melanin protects, and the strongest evidence that protection is not the whole story. In the California registry for 1988 to 1993, age-adjusted melanoma incidence per 100,000 was 17.2 in men and 11.3 in women among non-Hispanic Whites, 2.8 and 3.0 among Hispanics, 0.9 and 0.8 among Asians and 1.0 and 0.7 among non-Hispanic Blacks, a difference of roughly 17- to 19-fold in men and 14- to 16-fold in women between non-Hispanic Whites and Asians or non-Hispanic Blacks; but the tumors in Black men sat on the lower limb in 50 percent of cases against 9 percent in White men, and late-stage diagnosis was commoner in every non-White group ¹³. The tumors that arise in dark skin are disproportionately acral, on palms, soles and nail beds, where there is little pigment and little sun exposure, and where ultraviolet light is not thought to be the main cause: in the National Cancer Database, 251,864 patients, Black patients had 27.6-fold higher odds (95 percent confidence interval 24.4 to 31.2) of an acral melanoma than non-Hispanic Whites ⁸⁹, and in SEER for 2006 to 2015 acral melanoma, with an incidence of 2.0 per million person-years overall, made up 32.6 percent of melanomas in Black patients ⁹⁰.

Survival runs the other way from incidence. Across 96,953 SEER patients from 1992 to 2009, White patients survived longest, then Hispanic, then Asian American, Native American and Pacific Islander, then Black; within stage I disease the hazard ratio for death in Black patients was 3.037 (2.335 to 3.951), and within stage III 1.864 (1.211 to 2.87) ¹⁴. For acral melanoma specifically, five-year melanoma-specific survival was 84.3 percent in non-Hispanic Whites and 66.9 percent in Blacks, who presented older with thicker, more ulcerated tumors ⁹⁰, and in 4,188 acral cases from 2000 to 2020 the adjusted hazard of acral-melanoma death was 2.19 (1.74 to 2.76) in Black patients, 1.58 in Asian and Pacific Islander patients and 1.38 in Hispanic patients against Whites, while incidence rose 2.48 percent a year, driven by Hispanics and Whites and flat in Black patients ⁹¹. The usual explanation is late diagnosis in skin no one examines and amputations rather than excisions, and the amputation part is documented: Black patients had higher odds of major amputation at every stage even after adjusting for acral histology ⁸⁹. But a 2025 SEER analysis found no significant difference in disease severity at diagnosis between Black and White acral melanoma patients and still a doubled risk of melanoma death (hazard ratio 2.04, 1.31 to 3.11) after adjusting for clinical and socioeconomic covariates, and concluded that the delayed-diagnosis explanation does not hold on its own, while noting the lack of large genetic studies that might ⁷⁶.

The cleanest natural experiment on melanin and skin cancer is albinism. A 2026 meta-analysis of 12 studies and 1,747 people with albinism found a pooled prevalence of actinic keratosis, the precancerous sun lesion, of 38 percent (17.7 to 60.3) and of skin cancer of 19 percent (13.5 to 24.4); among the cancers, squamous cell carcinoma made up 55.4 percent and basal cell carcinoma 33.7 percent, and melanoma 0 percent, with an upper confidence limit of 0.56 percent. The studies disagreed with one another almost completely: the I-squared statistic, which estimates how much of the spread between studies is real rather than chance, ran from 89 to 99 percent, and anything above about 75 percent is usually called extreme. Most of the studies were cross-sectional ¹⁷. A single South African center saw 221 Black patients with albinism contribute 618 confirmed skin cancers over twenty years, 59.7 percent presenting before age 40, with squamous and basal cell carcinoma nearly equal at 50.2 and 49.7 percent and one melanoma ⁹². Removing melanin from dark skin therefore produces an epidemic of the keratinocyte cancers that ultraviolet causes directly, and almost no melanoma. That absence is suggestive, and it is not evidence of a mechanism: the studies are cross-sectional, mostly African, extremely heterogeneous, and they count people who are dying of squamous and basal cell carcinoma young, so melanoma has neither the follow-up nor the surviving population in which to appear. The melanoma that dark skin does get is disproportionately acral, about a third of melanomas in Black patients in SEER ⁹⁰, arising where pigment and sun are both scarce and where whole-genome sequencing finds structural changes and mutation signatures of unknown origin rather than the signature of ultraviolet light ⁹³. The albinism data are consistent with a contribution of pigment chemistry to melanoma ^12,39, and consistent with much else.

• Fitzpatrick sun-reactive skin typeSix categories from always burns, never tans (I) to never burns (VI), built by Thomas Fitzpatrick to dose ultraviolet treatment for psoriasis and later used for everything ⁶⁰.Sunscreen and phototherapy dosing, trial enrollment, and, wrongly, as a proxy for skin color and race ¹³⁴In 270 ethnically diverse people, 42% gave burning and tanning answers that could not be classified at all; a dermatologist's type separated types IV and V on the melanin index and self-report did not (P = 0.90) ⁶⁸.It grades sunburn, not pigment. In 789 adults, perceived Fitzpatrick called 26% dark where reflectance put 14%, and it captured differences in a melanoma classifier's scores less well than the Monk scale did ^118,135.
• Reflectance spectrophotometry: melanin index and individual typology angleA handheld instrument shines light on the skin and reads what returns; the melanin index and the individual typology angle (ITA, a number that falls as skin darkens) are computed from it ^68,118.Characterizing pigment in oximeter and device trials, and as the objective reference in studies of the verbal scales ^19,135A spectrophotometer and a colorimeter agreed strongly on the individual typology angle and on lightness (correlation above 0.7) in a 789-person study with 33,856 assessments in all ¹¹⁸; colorimetry is described as the gold standard for measuring color ¹³⁵.Site matters and thresholds have been set too light: about half the people below -30 degrees were below -50, where only 7% of the cohort sat, so cohorts counted as diverse may hold few genuinely dark participants ¹¹⁸.
• Chemical degradation and isotope fate tracingEumelanin and pheomelanin are measured by breaking the polymer down and quantifying marker fragments by liquid chromatography, the assay Ito and Fujita published in 1985 ⁵⁹; since 2021 labeled tyrosine can be traced into new pigment by mass spectrometry ^1,42.Research: hair and skin pigment composition, melanocyte cultures, the nitisinone pilot's hair measurements ⁹⁴The degradation method is described as robust; the tracing method reproduced known differences between melanocytes of different skin colors and MC1R genotypes ⁴².Degradation cannot tell synthesis from breakdown and is suited to changes over long periods; neither method is clinical ⁴².
• Neuromelanin-sensitive MRIA magnetic resonance sequence in which iron-bound neuromelanin in the substantia nigra and locus coeruleus shows bright, first demonstrated at 3 tesla in 2006 ^25,66.Research imaging of Parkinson's disease and, as a proxy for dopamine function, of psychosis ⁷⁰Pooled sensitivity 89% (95% CI 86-92%) and specificity 83% (76-88%) for Parkinson's disease against healthy controls in 12 studies of 403 patients and 298 controls ²⁶.The studies are case-control against healthy people, not against the mimics a clinician faces; accuracy was better with longer disease and with volume-based measures, and no trial has shown that the scan changes a diagnosis or an outcome ^26,125.
• Pulse oximetry (a measurement melanin disturbs)Red and infrared light are passed through a fingertip and the ratio of their pulsatile absorption is converted to oxygen saturation by a calibration curve built in volunteers; melanin absorbs both wavelengths ^63,114.Every hospital bed; oxygen titration and treatment eligibility, including during COVID-19 ⁷³Overestimation of 3.56 ± 2.45% at 60-70% saturation in dark skin against 0.37 ± 3.20% in light skin in 2005; under low perfusion a missed hypoxemia rate of 21.1% in dark against 1.1% in light skin in 2024 ^33,63.The clinical cost: occult hypoxemia in 11.7% against 3.6% of readings by self-identified race, a one-hour longer median delay for Black than for White patients in recognizing eligibility for COVID-19 treatment, and in 2026 only 1 of 34 devices passing the criteria regulators are moving toward ^18,19,73.

When it runs short: albinism, vitiligo, gray hair

Three different failures produce a shortage of pigment, and they fail at three different points: the enzyme, the cell and the stem cell. Oculocutaneous albinism is the enzyme and its organelle. It is a group of recessive disorders of melanin synthesis with a worldwide prevalence of about 1 in 17,000, which means about 1 person in 70 carries an allele. OCA1A, from complete loss of tyrosinase, makes no melanin for life; OCA1B, OCA2, OCA3 and OCA4, from partial tyrosinase loss or from loss of the OCA2, TYRP1 and MATP proteins of the melanosome, accumulate some pigment over the years. The eye is the organ that defines it: nystagmus, a translucent iris, a pale retina, an underdeveloped fovea, visual acuity usually between 20/60 and 20/400, and optic nerves that are misrouted at the chiasm so that stereoscopic vision is poor. Lifespan, development, intelligence and fertility are normal ¹⁵. The 2017 structure of TYRP1 suggested that most of its albinism mutations act by affecting the stability or the activity of a zinc-containing protein rather than by blocking a chemical step ⁴⁵. The skin consequence, in the sunlit regions of Africa where the condition is commonest, is the keratinocyte cancer epidemic described in the previous section ^17,92.

There is no treatment that restores the pigment, and the one serious attempt is instructive. Nitisinone, a drug used in a rare liver disease that raises blood tyrosine, had increased fur and eye pigment in a mouse model of OCA1B, the partial tyrosinase form, on the logic that more substrate might push a weak enzyme. In an open-label pilot, five adults with OCA1B took 2 milligrams a day for 12 months. Iris transillumination, the marker of eye melanin, did not change (a mean change of 1.0 plus or minus 1.54 points); the number of letters read on an eye chart improved by 4.2 in the right eye and 5 in the left; skin pigment on the inner arm rose by a melanin index of 1.72 (0.03 to 3.41) and hair pigment by 17.3 (4.4 to 30.2), confirmed chemically ⁹⁴. Hair and skin darkened a little, the eye did not, and five people without a control group cannot show whether the acuity gain was real. In a mouse model of OCA3, the TYRP1 form, the same drug raised plasma tyrosine about sevenfold without changing the color of the fur or of the pigmented structures of the eye, and the only difference it made was in the number of pigmented melanosomes in the iris, so the authors judged it unlikely to help that type ⁸⁰. The other idea tried in people was levodopa. Dopamine is an intermediate on the way to melanin, and animal work had suggested that supplying the precursor might improve vision by strengthening the retina's wiring. In a randomized, double-masked, placebo-controlled trial, 45 people with albinism, mean age 14.5 years and every one of them carrying a mutation in a known albinism gene, took levodopa 0.76 mg/kg, levodopa 0.51 mg/kg or placebo for 20 weeks, with complete follow-up; there was no statistically significant improvement in best-corrected visual acuity at either dose, and no serious adverse events ⁹⁵. That is a clean negative result in a disease with almost no trials at all.

Vitiligo is the cell. The melanocytes worked, and then the immune system killed them, leaving white patches; it is the commonest cause of depigmentation in the world with an estimated prevalence of 1 percent, it divides into segmental and the commoner non-segmental form with different prognosis and treatment, and it is psychologically devastating especially in darker skin, where the contrast is greatest ¹⁶. A 2026 meta-analysis of 11 East African studies, 8,444 people reported between 2000 and 2025, pooled a prevalence of 6 percent (3 to 8): 3 percent in children, 9 percent in adults, 7 percent in general-population samples, 1 percent in Tanzania and 8 percent in Ethiopia. Those figures sit far above the 1 percent global estimate and come from a small number of mostly Ethiopian studies, so they are not a settled regional prevalence ⁹⁶. The mechanism was pinned down in 2014: lesional skin from patients showed an interferon-gamma signature led by the chemokine CXCL10, raised in skin and serum, with its receptor on the killer T cells; in a mouse model, mice lacking CXCL10 or its receptor, or given a neutralizing antibody, developed minimal depigmentation, and neutralizing CXCL10 in mice with established, widespread disease produced repigmentation ⁹⁷. Interferon-gamma signals through Janus kinases, which is why the first drugs are Janus kinase inhibitors.

The first approved treatment was a cream. In two phase 3 trials, TRuE-V1 and TRuE-V2, 674 patients aged 12 and over with non-segmental vitiligo covering 10 percent or less of the body applied 1.5 percent ruxolitinib cream or vehicle twice daily for 24 weeks. The primary end point, a 75 percent improvement in the facial Vitiligo Area Scoring Index, was reached by 29.8 percent on the cream against 7.4 percent on vehicle in the first trial (relative risk 4.0, 1.9 to 8.4) and 30.9 against 11.4 percent in the second (2.7, 1.5 to 4.9); application-site acne occurred in 6.3 and 6.6 percent, and the authors, funded by the manufacturer, called for larger and longer trials ³⁰. The FDA approved the cream for non-segmental vitiligo in July 2022 ³¹. The mechanism trial published in 2026 closed the loop: in 41 patients on cream against 19 on vehicle, serum interferon-gamma-related proteins including CXCL10 fell at every time point, the fall in CXCL10 in serum and skin correlated with improvement, and facial scores fell 49.1 percent by week 24 ⁹⁸. For the two thirds who respond poorly in the first 12 weeks, adding narrow-band ultraviolet B, the phototherapy that stimulates the melanocyte reservoir in hair follicles, took the mean total-score change from minus 7.9 percent at week 12, in the 35 of 55 enrolled patients who added it, to minus 50.4 percent at week 48 in the 26 of them still assessed, in an uncontrolled exploratory phase 2 study ⁹⁹.

Oral drugs arrived in 2026. The Viti-Up program randomized 308 and 306 patients aged 12 and over to upadacitinib 15 milligrams daily or placebo; at week 48, a 50 percent total-body improvement was reached by 19 and 21 percent on the drug against 6 and 6 percent on placebo, and a 75 percent facial improvement by 25 and 23 percent against 6 and 7 percent. There were no adjudicated major cardiovascular events, venous thromboses, perforations, tuberculosis, lymphoma or non-melanoma skin cancer, and herpes zoster was the only opportunistic infection; the trials were funded by AbbVie, and at publication no systemic drug had yet been approved for the disease ³². A phase 2 of oral povorcitinib reported, post hoc and without a control arm, that 34.0 percent of 103 patients reached the total-body milestone and 45.6 percent the facial one at 52 weeks, with the feet responding worst at 25.8 percent and the head and neck best at 57.3 percent ¹⁰⁰, and three phase 3 ritlecitinib studies, planned at about 600, 400 (an extension drawing on the first) and 1,450 participants, have published protocols and no results ¹⁰¹. The pattern across all of them is the same: in the controlled trials a quarter to a third of patients reach a 75 percent improvement in the facial score within 24 to 48 weeks, and about 45 percent did so at 52 weeks in the uncontrolled povorcitinib extension; the body lags, the feet lag most, and the disease is suppressed rather than cured.

Gray hair is the stem cell. Each hair follicle keeps a small reserve of melanocyte stem cells in its bulge, which must supply new pigment cells to the bulb at every growth cycle. In 2005 Nishimura, Granter and Fisher showed in melanocyte-tagged mice and aging human follicles that graying is caused by defective self-maintenance of that reserve, that loss of the survival protein Bcl2 dramatically accelerates it by killing the stem cells as they enter dormancy, and that aging stem cells differentiate and pigment in the wrong place, inside the niche ³⁵. The folk belief that stress grays hair was tested in 2020: in mice, acute stress depleted the stem cells within days, not through immune attack or adrenal stress hormones but through the sympathetic nerves that innervate the niche, whose burst of norepinephrine drove the quiescent stem cells to proliferate, differentiate, migrate and disappear for good, and blocking that proliferation prevented the graying ³⁴. In people, profiling pigment along single hairs in 2021 showed gray hairs that spontaneously regained color across sexes, ethnicities, ages and body sites, with graying and reversal running in parallel with psychological stressors, and a model suggesting a threshold that can be crossed in either direction while the stem cells last ¹⁰².

The 2023 finding changed the picture of how the reserve works. Live imaging and single-cell sequencing in mice showed that most melanocyte stem cells do not sit in a protected state waiting to be called; they toggle between a stem state and a transit-amplifying state, moving between compartments of the follicle and changing their differentiation reversibly under local cues such as WNT signaling, a developmental signal with which neighboring cells tell a stem cell whether to remain one or to specialize, and the system is maintained by cells that revert rather than by a reserved pool exempt from change. With age, stranded stem cells accumulate that are stuck in a compartment where they no longer contribute pigment cells, and the authors suggest that restoring their mobility is a route to preventing graying ³⁶. A 2026 review from the same laboratory frames the system as a model of cellular aging generally ¹⁰³. All of this is mouse work; the human hair data are observational, and nothing has been shown to re-gray a human head on purpose.

In 2025 the tracking of individual stem-cell fates in mice turned graying into one half of a choice. Melanocyte stem cells that take a DNA double-strand break, a break through both strands of the helix, respond by differentiating and becoming senescent at once, which uses them up and grays the hair, and which protects against melanoma by taking damaged cells out of the pool. Carcinogens block that response: even in stem cells carrying double-strand breaks they switch on arachidonic acid metabolism and the niche's KIT ligand, so the damaged cell renews itself instead. Graying and melanoma are, in this mouse model, antagonistic outcomes of one injury, and the exposure decides which happens ⁷⁷. It is mouse work, and nothing of the kind has been shown in a person; but it is the first account in which going gray is the good outcome.

When it runs away: melasma and post-inflammatory hyperpigmentation

Too much melanin is rarely a medical danger and is a very common medical complaint. Melasma is the symmetrical brown patching of the cheeks, forehead and upper lip that appears in midlife, mostly in women and mostly in people with brown skin, worsens with sun and pregnancy and hormonal contraception, and relapses after almost every treatment. For decades it was treated as an overactive melanocyte and attacked with bleaching agents such as hydroquinone. Two 2026 reviews describe the shift in thinking. One proposes melasma as persistent melanocyte activation sustained by a photoaged dermis around it, a field of senescent fibroblasts, ultraviolet-activated sebaceous cells and dilated vessels that keeps sending pigment-promoting signals, arising in a window when the dermis has already aged but the melanocytes are still able to respond, which would explain why it starts in midlife, relapses for years and fades in old age ³⁷. The other organizes the triggers as an exposome: hormones, oxidative stress, thyroid disease, vitamin D metabolism and psychological stress from inside; ultraviolet, visible light, air pollution, drugs and skincare from outside; all converging on MITF, on oxidative cycles and on the same pro-opiomelanocortin pathway that orders a tan, and argues that treating formed pigment while leaving the drivers in place is why relapse is the rule ¹⁰⁴.

The treatment evidence is wide and thin. A 2025 network meta-analysis of 14 randomized trials published between 2022 and 2025, 738 women and 15 modalities, ranked intradermal platelet-rich plasma, alone or with oral tranexamic acid, above hydroquinone 4 percent, intradermal tranexamic acid, growth-factor plasma plus hydroquinone and tranexamic acid plus laser on the severity index, with mostly mild adverse events and, by the authors' grading, high-quality evidence; the trials were small and the first author works at an aesthetic clinic ¹⁰⁵. Tranexamic acid, an antifibrinolytic drug borrowed from the treatment of bleeding, runs through the modern melasma and hyperpigmentation literature because it blocks plasmin, which otherwise releases pigment-promoting signals from keratinocytes ¹⁰⁶; the evidence that it works is a collection of small trials and reviews rather than one decisive study. Strict photoprotection is the foundation, and since 2010 that has meant protection against visible light as well as ultraviolet. In skin types IV to VI, visible light produced pigmentation that was darker and longer-lasting than that produced by long-wave ultraviolet A, and produced none at all in type II skin ¹⁰⁷; and in 68 melasma patients randomized to a sunscreen containing iron oxide, which absorbs visible light, against a conventional ultraviolet-only sunscreen of the same protection factor, both groups also using 4 percent hydroquinone, the visible-light group did better at 8 weeks by 15 percent on the severity index, 28 percent on colorimetry and 4 percent on measured melanin ¹⁰⁸; the same comparison has been used to prevent relapse ¹⁰⁹. Full-spectrum photoprotection is placed at the center of management in the 2026 exposome review ¹⁰⁴.

Post-inflammatory hyperpigmentation is simpler to explain and harder to prevent. Any inflammation of the skin, acne, eczema, a burn, a laser, a chemical peel, can leave a brown stain that outlasts the insult by months, because inflammatory mediators drive melanocytes and because damage to the basement membrane lets pigment drop into the dermis where it clears slowly. It is commoner and more distressing in darker skin, and it is often the reason a person with brown skin sought treatment in the first place ³⁸. The oldest good evidence for treating the stain is for a retinoid. In a 40-week randomized, double-blind, vehicle-controlled trial, 54 Black patients applied 0.1 percent tretinoin cream or vehicle to post-inflammatory hyperpigmented lesions of the face or arms: the treated lesions were significantly lighter clinically (P less than 0.001), colorimetry showed 40 percent lightening toward normal skin color against 18 percent on vehicle (P equals 0.05), epidermal melanin fell 23 percent against 3 percent, which was not significant, normal skin lightened only minimally, and retinoid dermatitis occurred in half the treated patients and faded as the study went on ¹¹⁰. The other agent aimed at the handover step described earlier on this page is niacinamide, which does not touch tyrosinase or melanin synthesis at all: it inhibited melanosome transfer by 35 to 68 percent in a keratinocyte-melanocyte coculture and reduced facial hyperpigmentation against vehicle in 18 women over 4 weeks; the work was done at Procter and Gamble, and the clinical arm is small and short ¹¹¹. Prevention is the question that matters clinically, and a 2026 systematic review of 14 randomized trials, 11 of them in a network meta-analysis, of interventions before or after laser and energy-based treatments answered part of it: intradermal tranexamic acid (risk ratio 0.02, 0.00 to 0.53), topical corticosteroids, topical vasoconstrictors, oral tranexamic acid and cooling of the epidermis all reduced the incidence of hyperpigmentation compared with sunscreen alone, whitening agents and growth-factor preparations did not, and sunscreen alone was no better than placebo, which is the finding worth remembering; the trials were small and the authors ask for caution ¹¹². A 2024 narrative review of oral tranexamic acid reported that every study it found showed improvement without serious adverse effects, which is what narrative reviews of small positive studies tend to report ¹¹³.

Both conditions share a lesson with the rest of this page: the melanocyte is answering orders. In melasma the orders come from an aged dermis and from hormones; in post-inflammatory hyperpigmentation from the inflammatory cascade; in a tan from a keratinocyte's damaged DNA. Treatments aimed at the pigment, the bleaching agents and lasers, work on what has been made and not on who is ordering it, and that is why the conditions relapse and why the laser itself can cause the very stain it was meant to remove.

Pulse oximetry and skin tone: from a 2005 warning to the 2025 draft guidance

A pulse oximeter clips to a fingertip, shines red and infrared light through it, and measures how much of each the pulsing arterial blood absorbs; oxygenated and deoxygenated hemoglobin absorb the two colors in different proportions, and the ratio is converted to a saturation by a calibration curve built in volunteers. Melanin absorbs both colors. The finding is older than the attention paid to it. In 1990 Jubran and Tobin, titrating oxygen in 54 patients on ventilators, reported that an oximeter target of 92 percent reliably indicated adequate oxygenation in White patients but in Black patients was commonly associated with an arterial oxygen tension as low as 49 mm Hg, so that a target of 95 percent was needed; readings differed from the measured arterial saturation by more than 4 percent in 27 percent of Black and 11 percent of White patients ⁶¹. Fifteen years later, in 2005, Bickler, Feiner and Severinghaus in San Francisco desaturated 11 darkly and 10 lightly pigmented volunteers under controlled conditions and compared three finger oximeters against arterial blood: at arterial saturations of 60 to 70 percent the oximeters read too high by 3.56 plus or minus 2.45 percent in dark skin against 0.37 plus or minus 3.20 percent in light skin (P less than 0.0001), the bias rising roughly in proportion to desaturation and differing between devices ⁶³. That paper, like the 1990 one, then sat unacted on for fifteen years.

The clinical cost was counted in December 2020. Sjoding and colleagues at Michigan paired 10,789 oximeter readings with arterial blood gases from 1,333 White and 276 Black inpatients, and 37,308 pairs from 7,342 White and 1,050 Black intensive-care patients at 178 hospitals. Among readings of 92 to 96 percent, the range in which a clinician would not act, arterial saturation was below 88 percent in 88 of 749 measurements from Black patients (11.7 percent, 8.5 to 16.0) and 99 of 2,778 from White patients (3.6 percent, 2.7 to 4.7) in the Michigan cohort, and in 17.0 percent (12.2 to 23.3) against 6.2 percent (5.4 to 7.1) in the multicenter cohort; adjusting for age, sex and circulatory failure changed nothing, and the oximeter's ability to detect hypoxemia, measured as area under the curve, was 0.84 in Black and 0.89 in White patients. Black patients, the letter concluded, had nearly three times the frequency of occult hypoxemia ¹⁸. The study used self-identified race and measured no skin, which is both its weakness and the point: that is how the device fails in a hospital.

Larger data followed. Across 87,971 encounters in 215 hospitals, hidden hypoxemia, an oximeter reading of 88 percent or more with an arterial value below 88, occurred in 5.5 percent overall, 6.8 percent of Black, 6.0 percent of Hispanic, 4.8 percent of Asian and 4.9 percent of White patients, and the patients in whom it occurred went on to higher organ-failure scores and higher in-hospital mortality, 21.1 against 15.0 percent among Black patients ⁷². In 1,216 COVID-19 patients with 32,282 concurrent pairs, occult hypoxemia occurred in 28.5 percent of Black, 29.8 percent of non-Black Hispanic and 30.2 percent of Asian patients against 17.2 percent of White patients; the oximeter overestimated arterial saturation by 1.2 percent (0.6 to 1.9) in Black patients; and because treatment eligibility was defined by an oximeter threshold, Black patients had a 29 percent lower hazard of having their eligibility recognized (0.71, 0.63 to 0.80), a median delay of 1.0 hour among those eventually recognized, and made up 54.8 percent of the 451 patients whose eligibility was never recognized at all ⁷³.

The mechanism was then taken back to the laboratory, because retrospective studies cannot separate skin from the timing of blood draws, from how race was recorded or from the state of the circulation. In 2024 the San Francisco group studied 146 healthy volunteers, 25 with light, 78 with medium and 43 with dark skin by Fitzpatrick class, with 9,763 matched readings on two widely used oximeters during controlled hypoxemia. Pigment, the perfusion index, which is the strength of the pulse signal, and the depth of hypoxemia each added error; in light skin only poor perfusion did, while in medium and dark skin bias rose with both poor perfusion and hypoxemia; sex and age did not matter; and under low perfusion the combined rate of missed hypoxemia was 1.1 percent for light, 8.2 percent for medium and 21.1 percent for dark skin ³³. Dark skin and a weak pulse together, which is exactly the state of a sick patient, are the likely cause of the retrospective findings. A 2026 physics model adds a source-side explanation: with the broad-spectrum light-emitting diodes most oximeters use, the calibration curve itself depends on melanin, and it would not with monochromatic laser diodes ¹¹⁴; that is a simulation, not a device trial.

The 2026 data sharpen the picture and complicate it. In English intensive care units, the EXAKT study paired 11,018 oximeter readings with arterial co-oximetry in 903 critically ill adults across the five devices the National Health Service had distributed for home COVID-19 monitoring, grading skin tone with a handheld spectrophotometer: readings averaged 0.6 to 1.5 percentage points higher at the darkest skin tone than at the lightest, and the proportion of readings above 94 percent despite a paired arterial saturation of 92 percent or less was 5.3 to 35.3 percentage points higher in darker skin, a rate ratio of 2.3 to 7.1 ¹¹⁵. Against that, two prospective studies found nothing: in 70 newborns, 40 percent of them Black, with 136 tightly paired measurements, the oximeter slightly underestimated arterial saturation and no categorical difference by pigment appeared, with occult hypoxemia in a single lightly pigmented infant ¹¹⁶; and in 100 intensive-care adults whose pigment was measured by colorimeter, the estimated pigment effect was indistinguishable from random error, while a lower saturation and an older age predicted the error reliably ¹¹⁷. The bias is real in adults and it depends on the device; it is not yet demonstrated in newborns, and in a sick adult it competes with perfusion, age and the depth of hypoxemia.

Regulation moved in steps, all of them primary documents outside PubMed: a safety communication on oximeter limitations on 19 February 2021, an advisory panel meeting on 1 November 2022, and draft guidance on performance testing in January 2025 ⁷¹. The 2026 test of what the new rules would mean is sobering. Hughes and colleagues ran 34 oximeters through controlled desaturation in cohorts of at least 24 volunteers of whom at least 25 percent were darkly pigmented, grading pigment with the Monk Skin Tone scale and with the individual typology angle from a spectrophotometer. Twenty-eight of 34 passed the 2017 international standard, 22 passed the 2013 FDA guidance, 21 passed the anticipated international criteria and 1 of 34 passed the anticipated FDA criteria; 11 devices read higher in dark than in light skin, by a median 1.42 percent; and 18 devices could pass or fail depending on which participants were analyzed. Only some devices, the authors stress, show the bias, and whether new guidelines will improve real-world performance or reduce disparities is undetermined ¹⁹. The companion paper on how to measure pigment found that defining dark skin as an individual typology angle below minus 30 degrees made 14 percent of a cohort of 789 dark, where perceived Fitzpatrick type made 26 percent and self-identification as Black 23 percent, and that about half the dark group was below minus 50 degrees, where only 7 percent of the cohort fell; published thresholds for dark skin, they conclude, have been too light, and cohorts counted as diverse have held few genuinely dark participants ¹¹⁸.

Melanin and vitamin D: what is settled, what was an assay, what is not known

Skin makes vitamin D when ultraviolet B light splits a cholesterol precursor in the epidermis, and melanin absorbs ultraviolet B, so darker skin should make less. The correction was available before the claim. In 1981 Holick, MacLaughlin and Doppelt exposed human skin to simulated solar ultraviolet light and found that the formation of previtamin D3 reaches a plateau at about 10 to 15 percent of the skin's supply of its precursor, 7-dehydrocholesterol, after which further light destroys what it has made; more melanin, like a higher latitude, lengthened the time needed to reach that maximum but did not change the total reached, and they ranked the limiting factors as photochemical regulation first, pigmentation second, latitude third. Their subtitle was that skin pigment is not an essential regulator ⁵⁸. The classic demonstration of the opposite emphasis is from 1982: Clemens, Adams, Henderson and Holick gave one minimal erythema dose to two lightly pigmented and three heavily pigmented volunteers; serum vitamin D rose up to 60-fold within 24 to 48 hours in the first group and did not change significantly in the second, and when one dark-skinned subject was re-exposed to a dose six times larger, that subject's level matched the light-skinned subjects' after the smaller dose ²⁰. Five people, one dose each, and from them came the teaching that dark skin needs several times more sun. The teaching was wrong in degree. In 2020 Young and colleagues in London gave 102 healthy volunteers of Fitzpatrick types II to VI the same sub-erythemal dose, well below sunburn, to 85 percent of the body, five times at intervals of 3 to 4 days, with either solar-simulated light or narrowband ultraviolet B. The marker 25-hydroxyvitamin D rose linearly with dose in every skin type; only type II had a significantly steeper slope than the others; and the melanin inhibition factor between the extremes, type II and type VI, was about 1.3 to 1.4, which the authors call small compared with melanin's effect on sunburn, while allowing that it may be enough to explain the population data ²¹. The 1982 study used a sunburn dose, at which light skin was damaged and dark skin was not; at doses people actually receive, the pigment's effect on vitamin D is modest.

Two 2025-2026 studies fill in the edges. Thirty-one adults were exposed to 0.8 standard erythema doses three times a week for 9 weeks; a standard erythema dose is a fixed physical quantity of ultraviolet, the same for everyone, unlike the minimal erythema dose, which is the dose that reddens a particular person's skin. They raised their minimal erythema dose by 50 percent, their melanin index by 12 percent and their stratum corneum thickness by 34 percent, and their 25-hydroxyvitamin D by 21 percent after correcting for the seasonal fall, with local immunosuppression in the skin but no systemic change: a tan is a small vitamin D gain bought with a thicker, more suppressed epidermis ¹¹⁹. And in northern Britain, screening 299 people between December 2024 and August 2025 by mass spectrometry on dried blood spots found 25-hydroxyvitamin D below 50 nanomoles per liter in 54.8 percent of adults aged 65 and over and 72.1 percent of adults of Fitzpatrick types IV to VI, with no decline in summer, which the authors read as sunlight being an inadequate corrective at that latitude for either group ¹²⁰. The population association between dark skin and low measured vitamin D is real; it is the size of the pigment's contribution, against latitude, clothing, time indoors and diet, that the 1982 study overstated.

The second argument is subtler and was undone more thoroughly. In 2013 Powe and colleagues measured 2,085 Black and White adults in Baltimore: total 25-hydroxyvitamin D was 15.6 against 25.8 nanograms per milliliter (39 against 64 nanomoles per liter, since 1 nanogram per milliliter is 2.5 nanomoles per liter) and vitamin D-binding protein, the carrier that holds most of the vitamin in blood, was 168 against 337 micrograms per milliliter in Black against White participants; two common polymorphisms in the binding protein's gene explained 79.4 percent of the variation in its level; bone mineral density was higher in Black participants, 1.05 against 0.94 grams per square centimeter; and estimated bioavailable vitamin D, the fraction not bound to the carrier, was the same, 2.9 against 3.1 (P equals 0.71). The conclusion was that Black Americans were not deficient at all, only differently distributed, and it explained the paradox of lower vitamin D with stronger bones ²². It was elegant and it rested on a monoclonal immunoassay. In 2015 Hoofnagle, Eckfeldt and Lutsey built a mass spectrometry assay that measures peptides shared by all isoforms of the binding protein and ran it on 187 genotyped participants of the Atherosclerosis Risk in Communities study, a long-running United States population cohort: by genotype, the mass spectrometry results matched a polyclonal immunoassay, while the monoclonal assay read far lower whenever the Gc1f haplotype was present, and Gc1f is commoner in Black Americans; whether binding protein differs by race at all, they wrote, is unclear ²³. In 2016 Nielson and colleagues settled it in 1,057 men from the United States, the United Kingdom and The Gambia: by the monoclonal assay, binding protein in men of African ancestry was about 50 percent lower than in Whites; by polyclonal antibodies or proteomics it was not lower at all; directly measured free 25-hydroxyvitamin D correlated with total (r equals 0.84) and was lower in African Americans; the racial difference in binding protein was an assay artifact, and total 25-hydroxyvitamin D stands as the useful measure ²⁴. That is a reversal, and a clean one.

What is not settled is what the low measured level means for health in people with dark skin, because the bone paradox survived the assay correction: Black Americans in the 2013 cohort still had the higher bone density ²², and the explanation for that is not the binding protein. The 2024 Endocrine Society guideline, built on randomized trials rather than associations, found no clinical-trial evidence to support routine screening for 25-hydroxyvitamin D in the general population, in people with obesity or in people with dark complexion, and suggested against routine testing in all the populations it considered, while suggesting empiric supplementation without testing for children, adults over 75, pregnancy and high-risk prediabetes and admitting that the optimal dose is unclear ⁷⁵. The evolutionary argument that skin lightened at high latitude to permit vitamin D synthesis ⁵⁶ is plausible and consistent with the measurements, and it does not tell a clinician in Baltimore or Newcastle whether a particular dark-skinned patient with a level of 40 nanomoles per liter needs anything. What the two reversals leave standing is modest: melanin reduces vitamin D synthesis by a factor of about 1.3 to 1.4 at ordinary doses; the measured level is low in many dark-skinned people in northern cities, for reasons that include but are not dominated by pigment; the binding-protein rescue of that number was an artifact; and no trial has shown that correcting the number by pigment group changes an outcome.

Neuromelanin: the pigment of the substantia nigra, Parkinson's disease, and MRI

The substantia nigra is named for a pigment. Its dopamine neurons, and the norepinephrine neurons of the locus coeruleus, fill with age with dark granules that common laboratory animals do not make ⁵⁰, which long limited experimental study of the pigment. Neuromelanin is an autophagic product: when dopamine or norepinephrine oxidizes in the cytoplasm, the quinones react with proteins and lipids, and the cell sweeps the mass into an organelle that is an autolysosome, carrying lysosomal membrane and matrix proteins at lower numbers than true lysosomes, lipid bodies rich in dolichols, double membranes and the proteins of autophagy, but which never empties; it appears to be a compartment of final destination for molecules that other systems could not degrade, with a very slow turnover across the life of the neuron ⁴³. The pigment rises linearly with age in both nuclei ²⁵. Analytical microscopy of human substantia nigra finds single pigment granules and lipid bodies of about 200 to 600 nanometers, the granules holding sulfur, iron and copper, with iron at about 0.15 atomic percent, and the lipid bodies holding phosphorus consistent with phospholipids ¹²¹. Elemental mapping of seven human locus coeruleus samples at 5 to 10 nanometer resolution in 2026 placed the metals, aluminum, calcium and iron, mainly in the pigment compartment identified by its sulfur-rich, pheomelanin-like portion, found sulfur and iron rising with age, and found more of all three metals in the right hemisphere, an asymmetry the authors flag without explaining ⁴⁴.

Whether the pigment protects or kills is the question, and the answer in the literature is both, in sequence. The neuromelanin-iron complex is the main iron compound of these neurons, because the pigment is an effective chelator; trapping iron and converting excess dopamine into a stable polymer protect the cell from oxidative stress. But an organelle that has spent decades accumulating iron and toxins can trigger degeneration, and when a neuron dies and spills its neuromelanin, the pigment activates microglia, the brain's resident immune cells, which kill more neurons and release more pigment in a self-propelling loop; that, the review argues, is the link between aging and Parkinson's disease ¹²². The experiment that made the hypothesis testable came in 2019. Rats given human tyrosinase in the substantia nigra, where it is normally present only at low levels, built human-like neuromelanin with age up to the amounts seen in elderly people, and once the intracellular load crossed a threshold they developed slowness of movement, Lewy-body-like inclusions and loss of the nigrostriatal pathway; enhancing lysosomal function reduced the pigment and prevented the degeneration, and the authors propose that intracellular neuromelanin sets the threshold at which Parkinson's disease begins ⁵⁰. It is a rat with a human enzyme, not a person, but it is the first rodent model that makes the pigment, which ordinary laboratory animals lack ⁵⁰. In 2026 the same strategy in non-human primates produced extracellular pigment release, iron accumulation, inflammation and inclusions without substantial dopamine cell loss, and repeated low-intensity focused ultrasound with microbubbles, which transiently opens the blood-brain barrier, reduced both intracellular and extracellular neuromelanin and tended to lower free ferric iron; it is offered as a route toward a therapy and is not one ¹²³.

The pigment can be seen in living people because it binds iron, which makes it paramagnetic. Sasaki's group in Morioka showed in 2006 that a 3-tesla sequence makes the locus coeruleus and substantia nigra pars compacta glow in healthy volunteers, matching their positions in gross specimens, and that in Parkinson's disease the signal in both is greatly reduced ⁶⁶. Validation followed: in 18 patients and 18 controls, nigral volume and contrast on neuromelanin MRI were reduced and correlated with striatal dopamine transporter density on single-photon emission tomography, volume predicting best, while iron-sensitive measures did not differ ¹²⁴; and in 2019 the signal was shown to track neuromelanin concentration in postmortem midbrain even without degeneration, to relate to dopamine release in the dorsal striatum on positron emission tomography in living people, and to correlate with the severity of psychosis in schizophrenia, consistent with the nigrostriatal pathway's role there ⁷⁰. Two meta-analyses of its diagnostic performance for Parkinson's disease against healthy controls gave pooled sensitivity and specificity of 0.82 and 0.82 ¹²⁵ and 89 percent (86 to 92) and 83 percent (76 to 88) in 12 studies of 403 patients and 298 controls, better with disease of more than 5 or 10 years, with volume rather than intensity and with automated segmentation ²⁶. A 2026 systematic review of 17 studies and 2,228 participants found that combining the neuromelanin sequence with quantitative susceptibility mapping of iron gave areas under the curve of 0.86 to 0.99 and separated Parkinson's disease from some of its mimics, with simultaneous acquisition and automated segmentation shortening scans ¹²⁶. None of these is a trial showing the scan changes a diagnosis a neurologist would otherwise have missed, and all but the last compare patients with healthy people rather than with the diseases that look like Parkinson's. Two newer papers narrow that gap. A 2025 review of clinical magnetic resonance biomarkers sets neuromelanin imaging and the nigrosome-1 swallow-tail sign against the differential a neurologist actually faces: both are abnormal in Parkinson's disease and in most atypical parkinsonian syndromes, and both are normal in essential tremor, in drug-induced parkinsonism and in vascular parkinsonism, so they rule those mimics out but do not separate Parkinson's disease from the atypical syndromes, which dopamine transporter imaging, still the reference standard, cannot do either; the authors' proposal is to add the sequence to a standard scan and use it to choose who needs the nuclear scan ¹²⁷. And the sequence has now been used before diagnosis: among 66 people with idiopathic rapid eye movement sleep behavior disorder, the dream-enactment disorder that often precedes Parkinson's disease, 37 had abnormal dopamine transporter imaging, and neuromelanin volume and signal-to-noise ratio in the associative and limbic parts of the substantia nigra separated them from those with normal scans with areas under the curve of 0.73 and 0.75, which is fair rather than good ¹²⁸.

The skin pigment's receptor reappears here, unexpectedly. In 2026 Schumacher and colleagues asked whether the common loss-of-function variants of MC1R, carried by more than 60 percent of people of European descent, change the course of Parkinson's disease. Among 809 patients in the Parkinson's Progression Markers Initiative, 505 carriers and 304 non-carriers, the 383 with sporadic disease showed carriers declining 30 percent faster on the motor scale, 0.57 points a year (0.16 to 0.98), and in a replication cohort of 587 patients from three trials, 50 percent faster, 1.37 points a year (0.28 to 2.46); in a small prodromal group of 53, carriers had a 4.75-fold (1.48 to 15.27) risk of converting to Parkinson's disease ⁵¹. It is observational and confined to European ancestry, and whether the receptor acts through neuromelanin, through oxidative stress or through something else in the brain is not known; what it does is tie the red-hair receptor of the 1995 paper to a disease of the pigmented neurons.

Tanning drugs: afamelanotide, dersimelagon, and the illicit melanotan II

Because MC1R orders eumelanin, an agonist of the receptor orders a tan without sun. The first to be approved was afamelanotide, a synthetic analog of alpha-melanocyte-stimulating hormone, and the disease that earned it was not cosmetic. Erythropoietic protoporphyria is an inherited defect of heme synthesis in which a light-activated precursor accumulates in red cells and skin, so that minutes of sunlight cause hours of burning pain without visible change, and patients live indoors. In two randomized, double-blind, placebo-controlled trials, 74 patients in the European Union and 94 in the United States received a 16-milligram implant under the skin every 60 days, five implants in Europe and three in the United States. In the American trial, pain-free direct sun exposure at 6 months was a median 69.4 hours on the drug against 40.8 on placebo (P equals 0.04); in the European trial at 9 months, 6.0 against 0.8 hours (P equals 0.005), with 77 phototoxic reactions against 146; quality of life improved in both and adverse events were mostly mild. The trials were funded by Clinuvel, the manufacturer ²⁷. The FDA approved the implant in October 2019 as the first treatment to increase pain-free light exposure in the disease; the approval is a primary document outside PubMed ²⁸. The European Medicines Agency had authorized the implant for the prevention of phototoxicity in adults with the disease five years earlier, in 2014 ²⁹. The European post-authorization cohort the regulator required, 117 patients at one Dutch center over a median 2.0 years, found 98 percent continued treatment, time outdoors rose by 6.1 hours a week (3.62 to 8.67), the disease-specific quality-of-life score rose 14.01 percent, and phototoxic reactions were less painful but neither fewer nor shorter, with minor adverse events ⁶⁹. The drug works by darkening skin, which is to say it is a tanning drug prescribed for a disease in which a tan is medicine.

An oral version followed. Dersimelagon, a selective MC1R agonist taken by mouth, was tested in a phase 2 trial of 102 patients, 93 with erythropoietic and 9 with X-linked protoporphyria, at 100 or 300 milligrams daily or placebo for 16 weeks; the daily time to the first warning symptom in sunlight lengthened by 53.8 minutes at the lower dose (P equals 0.008) and 62.5 minutes at the higher (P equals 0.003) relative to placebo, and the commonest adverse events were nausea, freckles, headache and darkening of the skin, which is the mechanism showing. It was funded by Mitsubishi Tanabe and is a phase 2 result ⁷⁴. A different route to the same end has been shown in the laboratory only: small molecules that inhibit salt-inducible kinase raised MITF and produced eumelanin in red-hair mice and in normal human skin explants, a chemical tan aimed at reducing cancer risk in people who cannot tan ⁸¹; no trial has been published.

The illicit version precedes both, and it has a published parent. In 1996 a phase 1 study at the University of Arizona gave three healthy men melanotan II by subcutaneous injection on alternate days, escalating by 0.005 milligram-per-kilogram steps from 0.01 to 0.03 milligrams per kilogram in two of them and to 0.025 in the third: two developed measurable pigmentation of the face, upper body and buttock by reflectance a week after the dosing ended, mild nausea occurred at most doses, one of the two men at the top dose had grade II somnolence and fatigue, and a stretching-and-yawning complex accompanied spontaneous erections lasting one to five hours ¹²⁹. Five doses in three people is the whole of the controlled human evidence behind a drug now sold online. Melanotan II is an unlicensed synthetic melanocortin analog, sold online and in gyms as an injection and lately as a nasal spray, that activates MC1R and other melanocortin receptors and produces a tan without sun; its sale is illegal in the United Kingdom and many other countries ¹³⁰. The published harms are case reports, which is a statement about surveillance rather than about safety. A 22-year-old woman who used a melanotan II nasal spray for tanning developed a malignant melanoma of the mucosa of the upper jaw, resected and followed by immunotherapy; the authors raise the spray as a possible risk factor and cannot show causation from one case ¹³⁰. A 2026 report describes five primary melanomas in situ in a person who combined tanning beds, melanotan and anabolic hormones, three exposures in one patient, so that no one of them can be blamed ¹³¹. A dental case followed a patient who injected melanotan II for 64 days: symmetrical brown pigmentation of the gums and the lining of the cheeks appeared, the cheek pigment nearly cleared a month after stopping, and the gum pigment was still there, fainter, at three months ¹³². The drug pigments mucous membranes as well as skin. Beyond the skin, a 2026 systematic review of 68 studies of tanning agents records rhabdomyolysis, renal infarction and priapism among the serious adverse effects reported with unregulated melanotan I and II ¹³³. Whether it causes melanoma is not known; that it is taken without a prescription, in unknown doses, by people who are then not followed is.

The Fitzpatrick scale and its limits

The scale every dermatology form uses was built for a dosing problem. Thomas Fitzpatrick's six sun-reactive skin types, from type I, always burns and never tans, to type VI, never burns, were devised to choose the starting dose of ultraviolet A for psoralen phototherapy of psoriasis, and his 1988 paper on their validity and practicality is indexed under psoralen-ultraviolet therapy, psoriasis, erythema and sunburn; the record carries no abstract ⁶⁰. The questions that assign a type, how much you burn and how much you tan after the first exposure of summer, assume a person for whom that is a meaningful event; the argument that the scale was built around, and for, lighter skin is the basis of the 2021 call to stop using it as a measure of skin color ¹³⁴.

That assumption was tested in 2013 in 270 ethnically diverse people against reflectance spectrophotometry, which reads melanin from the skin directly. Forty-two percent, 114 of 270, gave answers to the burning and tanning questions that could not be classified at all under the standard definitions. A dermatologist's assigned type separated types III from IV and IV from V on the melanin index; the participants' self-assigned type separated III from IV but not IV from V (P equals 0.90); and the tanning answers did not correlate with the dermatologist's type at all, though the burning answers did ⁶⁸. A scale on which nearly half the people cannot place themselves, and on which the self-report cannot tell brown skin from dark brown, is not a measure of skin color, and in 2021 Okoji, Taylor and Lipoff argued in the British Journal of Dermatology that it is time to replace it as one ¹³⁴.

The replacement candidates are a verbal scale and an instrument. The Monk Skin Tone scale, ten swatches developed by the sociologist Ellis Monk, has been adopted in device studies; in a 2025 prospective comparison against a colorimeter, the Monk ratings showed slightly tighter clustering in color space and high repeatability, colors extracted from dermoscopic photographs correlated poorly with the colorimeter, and the Monk ratings captured differences in an artificial-intelligence melanoma classifier's scores better than Fitzpatrick types did; the authors state flatly that Fitzpatrick type is not a proxy for skin tone, and that labeling image datasets by tone matters because the algorithms perform differently across it ¹³⁵. The 2026 cohort study of 789 adults and 33,856 assessments compared perceived Fitzpatrick, Monk and Von Luschan scales with two instruments, a spectrophotometer and a colorimeter, found that the two instruments agreed strongly, found that each self-reported race category spanned a wide range of pigment, and found that the choice of method changed how diverse a cohort looked: 14 percent dark by instrument, 16 percent by Monk, 26 percent by perceived Fitzpatrick, 23 percent by self-identification as Black, with the published instrument threshold itself too light ¹¹⁸. A 2026 analysis overlaying the scales on 15,256 spectral reflectance measurements from a global archive concluded that neither Fitzpatrick nor Monk captures the continuous, multidimensional nature of pigmentation and that printed color targets fail under changing light; all four authors work for a spectral-imaging company, which is a conflict to note and does not make the measurement wrong ¹³⁶.

The practical rule that falls out of thirty years of this work is short. The Fitzpatrick type is a fair question about sunburn and a poor description of skin, race is not a measurement of pigment at all, and wherever the number matters, in an oximeter trial, in a cancer-screening cohort, in a dataset that trains a classifier, pigment should be measured by an instrument at a stated site, with a stated threshold, and reported as such.

What was overturned

Melanin only protects. True as far as it goes, and in 2015 it went further: fragments of melanin excited by the oxidants that ultraviolet light leaves behind go on making the DNA dimers of sunlight for hours in the dark, and these dark dimers are the majority; melanin may be carcinogenic as well as protective ¹². The 2026 extension of the argument to the action spectrum itself, that sunburn is an incomplete surrogate for mutagenic risk, comes from the same authors and awaits independent measurement in living skin ⁸³.

Melanoma in fair skin is a sunlight disease. Partly: in red-hair mice carrying a melanoma oncogene, pheomelanin drove invasive melanoma with no ultraviolet exposure, and removing pigment was protective ³⁹; in human melanomas, MC1R variants were associated with more mutations of both the ultraviolet signature and other classes ⁴⁰; and some MC1R variants raise melanoma risk without being tied to red hair at all ⁹. No human melanoma has been shown to arise from pheomelanin in the dark. Two 2026 papers push back from the other side: blocking pheomelanin synthesis raised systemic oxidative damage in the birds that make it, which suggests the pigment serves cysteine disposal ⁸⁶; melanin bound to DNA or RNA protected it in a test tube ⁸⁷; and in 123,179 matched patients, vitiligo, which destroys melanocytes, went with less skin cancer and less melanoma rather than more ⁸⁸.

Dark skin needs several times more sun to make vitamin D. From five volunteers given a sunburn dose in 1982 ²⁰ to 102 volunteers given ordinary doses in 2020: a melanin inhibition factor of about 1.3 to 1.4 between skin types II and VI ²¹.

Black Americans are not vitamin D deficient because their binding protein is low and their bioavailable vitamin D is normal. The binding protein was not low; the monoclonal assay misread the Gc1f haplotype that is commoner in Black Americans, and by mass spectrometry, polyclonal antibody or proteomics there is no racial difference; directly measured free vitamin D is lower, like the total ^22,23,24. The bone paradox the 2013 paper tried to explain still stands, unexplained.

A pulse oximeter reads skin as neutral. Warned against in 2005 in 21 volunteers ⁶³, shown false at scale in 2020 ¹⁸, mechanism supplied in 2024 ³³, and in 2026 one of 34 devices met the criteria regulators are moving toward ¹⁹.

The racial disparity in acral melanoma survival is delayed diagnosis. A 2025 SEER analysis found no difference in severity at diagnosis between Black and White acral melanoma patients and still a doubled death rate after adjustment ⁷⁶; the amputation disparity, though, is real and independent of histology ⁸⁹.

The Fitzpatrick scale measures skin color. It measures sunburn, in people for whom sunburn is an event; 42 percent of a diverse sample could not be classified by its questions ⁶⁸, it is not a proxy for skin tone ¹³⁵, and it overcounted dark skin nearly twofold against an instrument ¹¹⁸.

Melanocyte stem cells rest in a protected niche until called. In mice they shuttle between states and compartments and are maintained by reversion, and the ones that gray a hair are the ones that get stranded ³⁶; and the folk claim that stress grays hair turned out to be true in mice, through sympathetic nerves rather than stress hormones ³⁴, with reversal documented in human hairs ¹⁰².

Vitiligo is untreatable. The first approved drug, a Janus kinase cream, produced at least a 75 percent improvement in the facial vitiligo score in about 30 percent of patients at 24 weeks ^30,31, and an oral Janus kinase inhibitor did the same in about a quarter at 48 weeks in two phase 3 trials in 2026 ³²; suppression, not cure.

Melanin is a skin pigment. The pigmented neurons of the substantia nigra and locus coeruleus carry a pigment of their own, made from the neurotransmitters they use, that rises with age, binds iron, can be imaged, sets a threshold for a Parkinson's-like syndrome in a rat ⁵⁰, and whose skin-pigment receptor, MC1R, predicts faster Parkinson's progression in people ⁵¹.

Levodopa improves vision in albinism. Dopamine is an intermediate of melanin synthesis, and animal work had suggested that supplying the precursor would strengthen the retina's wiring; in the one adequately powered randomized trial, 45 people with genetically confirmed albinism took levodopa at either of two doses or placebo for 20 weeks with complete follow-up, and visual acuity did not improve at either dose ⁹⁵.

Going gray is simply damage accumulating. In mice, the stem cells that gray a hair do so by differentiating and becoming senescent in response to DNA double-strand breaks, a response that uses them up and protects against melanoma, while carcinogens suppress it so that the damaged stem cells renew themselves instead ⁷⁷; graying and melanoma are, in that model, the two outcomes of one injury, and nothing of the kind has yet been shown in a person.

What is not known

The structure of the pigment itself. After a century, the arrangement of the units in eumelanin and pheomelanin, the size of the polymers, and how they are packed inside a melanosome are described by the specialist reviews as not fully understood ²; the casing model of a pheomelanin core under a eumelanin shell is inferred from reaction kinetics rather than seen ¹.

How pigment crosses to the keratinocyte, and what the keratinocyte does with it. Four models coexist and the 2021 review suggests they may all be true under different conditions; the processing of melanin inside keratinocytes, which decides whether granules sit singly or in clusters, is called ill-characterized ⁶. Since that packaging is half of skin color and much of its photoprotection ^7,10, this is a large gap.

Whether dark dimers happen in a living person's skin after a day outdoors, and how much of human melanoma they account for. The finding is from cultured cells and mice ¹², and the proposals to intervene with nitric oxide synthase inhibitors or triplet quenchers ⁸⁴ are ahead of any human measurement.

What a red-haired person's ultraviolet-independent risk is in numbers. The mouse is clean ³⁹ and the human tumor genomics are associations with wide confidence intervals ⁴⁰; no cohort has estimated how much melanoma a red-haired population would have with no sun at all.

Why Black patients with acral melanoma die at twice the rate at the same severity of disease at diagnosis ⁷⁶, and why people with albinism get squamous and basal cell carcinoma in about a fifth of cases and almost never melanoma ^17,92; the acral-melanoma authors note the lack of large genetic studies ⁷⁶, and the albinism meta-analysis asks for longitudinal studies with standardized diagnostic criteria ¹⁷.

Whether correcting a low 25-hydroxyvitamin D level in a dark-skinned person changes any outcome. The guideline that looked for randomized evidence found none and advised against routine testing ⁷⁵; the bone paradox, higher bone density with lower measured vitamin D in Black Americans ²², survived the correction of the assay that was supposed to explain it ²⁴ and has no accepted explanation.

What pulse oximeter bias has cost. The delays are measured ⁷³ and the association with death is measured ⁷², but no study has shown that a device without the bias, or a correction factor, reduces mortality; and the 2026 device study warns that passing a new standard may depend as much on who is recruited as on the device ^19,118.

Whether neuromelanin is cause or marker in Parkinson's disease. The rat with human tyrosinase says a threshold exists ⁵⁰; the primate says the pigment can be reduced from outside the skull ¹²³; whether lowering it in a person would slow anything is untested, and the MC1R association with faster progression ⁵¹ has no mechanism yet.

Whether gray hair can be reversed on purpose. The stranded-stem-cell model and the sympathetic-nerve model are mouse findings ^34,36; the human reversals are spontaneous ¹⁰²; and no intervention has been shown to restore pigment to a human hair that has lost it.

The long-term safety of the Janus kinase drugs in a disease that is not life-threatening but is lifelong and can be psychologically devastating ¹⁶. The 48-week trials saw no adjudicated cardiovascular or thrombotic events ³², the cream's trialists asked for larger and longer studies ³⁰, and the treatment suppresses rather than cures, so the exposure may be measured in decades.

What melanotan II does to the people who take it. There is no cohort, no registry and no dose; there are case reports ^130,131,132, and the population using it is the one in which melanoma is already commonest.

Where it connects

Sources

136 sources, numbered as they are cited. Every one was checked against PubMed or its publisher before it was cited here; the note under each says what it shows and what it does not.

1. 1
   Ito S, Wakamatsu K. Chemistry of mixed melanogenesis--pivotal roles of dopaquinone.Photochem Photobiol · 2008 · 84(3):582-92doi:10.1111/j.1751-1097.2007.00238.x · PMID 18435614

   The chemistry in one review: tyrosinase makes dopaquinone from tyrosine; without thiols it cyclizes to cyclodopa, then dopachrome, then 5,6-dihydroxyindole and (less) its 2-carboxylic acid, whose ratio is set by dopachrome tautomerase, and these polymerize to eumelanin; with cysteine present the chain is diverted to cysteinyldopas and pheomelanin. Pulse radiolysis puts mixed melanogenesis in three stages and the authors propose a casing model, a pheomelanin core under a eumelanin surface. Chemistry review, not a clinical study.

2. 2
   Simon JD, Peles D, Wakamatsu K, Ito S. Current challenges in understanding melanogenesis: bridging chemistry, biological control, morphology, and function.Pigment Cell Melanoma Res · 2009 · 22(5):563-79doi:10.1111/j.1755-148X.2009.00610.x · PMID 19627559

   States that the molecular structure of melanins, the mechanism of their production and their organization inside the melanosome remain to be fully understood, and that most natural melanins are mixtures of eumelanin and pheomelanin. The reason this page treats melanin as a process rather than a molecule.

3. 3
   Raper HS. The Tyrosinase-tyrosine Reaction: Production from Tyrosine of 5: 6-Dihydroxyindole and 5: 6-Dihydroxyindole-2-carboxylic Acid-the Precursors of Melanin.Biochem J · 1927 · 21(1):89-96doi:10.1042/bj0210089 · PMID 16743827

   The founding paper of the Raper-Mason pathway, identifying the two dihydroxyindoles as melanin's precursors; the PubMed record carries no abstract, so it is cited for the finding named in its title and its date.

4. 4
   Mason HS. The chemistry of melanin; mechanism of the oxidation of dihydroxyphenylalanine by tyrosinase.J Biol Chem · 1948 · 172(1):83-99PMID 18920770

   Mason's third paper on melanin chemistry, which completed the Raper-Mason scheme of dopa oxidation by tyrosinase. The PubMed record carries no abstract and no registered DOI, so it is cited for its title and date only.

5. 5
   Raposo G, Marks MS. Melanosomes--dark organelles enlighten endosomal membrane transport.Nat Rev Mol Cell Biol · 2007 · 8(10):786-97doi:10.1038/nrm2258 · PMID 17878918

   Review defining melanosomes as tissue-specific lysosome-related organelles in which melanins are synthesized and stored, formed by novel pathways from early endosomal intermediates; genetic disorders of their biogenesis illuminate endosomal sorting. Cell biology review.

6. 6
   Moreiras H, Seabra MC, Barral DC. Melanin Transfer in the Epidermis: The Pursuit of Skin Pigmentation Control Mechanisms.Int J Mol Sci · 2021 · 22(9):4466doi:10.3390/ijms22094466 · PMID 33923362

   Reviews the competing models of how melanin moves from melanocyte to keratinocyte, finds recent evidence favors exocytosis followed by phagocytosis and shedding of vesicles, proposes that several mechanisms coexist, and calls melanin processing inside the keratinocyte poorly understood. Review.

7. 7
   Thong HY, Jee SH, Sun CC, Boissy RE. The patterns of melanosome distribution in keratinocytes of human skin as one determining factor of skin colour.Br J Dermatol · 2003 · 149(3):498-505doi:10.1046/j.1365-2133.2003.05473.x · PMID 14510981

   Electron microscopy of photoprotected forearm skin: melanosomes in keratinocytes of dark skin were predominantly individual (88.9%), in light skin predominantly clustered (84.5%), and in Asian skin mixed (62.6% individual, 37.4% clustered); mean melanosome size was largest in dark skin (1.44 x 10^-2 square micrometers), then Asian (1.36) and Caucasian (0.94). Human tissue, small numbers.

8. 8
   Valverde P, Healy E, Jackson I, Rees JL, Thody AJ. Variants of the melanocyte-stimulating hormone receptor gene are associated with red hair and fair skin in humans.Nat Genet · 1995 · 11(3):328-30doi:10.1038/ng1195-328 · PMID 7581459

   The first human MC1R paper: sequence variants were found in over 80% of people with red hair and/or fair skin that tans poorly, in fewer than 20% with brown or black hair and in less than 4% of good tanners, establishing the receptor as a control point of human pigmentation. Genetic association in a British sample.

9. 9
   Raimondi S, Sera F, Gandini S, et al. MC1R variants, melanoma and red hair color phenotype: a meta-analysis.Int J Cancer · 2008 · 122(12):2753-60doi:10.1002/ijc.23396 · PMID 18366057

   Meta-analysis of 11 studies on MC1R and melanoma and 9 on phenotype: seven variants were associated with melanoma with odds ratios from 1.42 (95% CI 1.09-1.85) for R163Q to 2.45 (1.32-4.55) for I155T; R151C carried an odds ratio of 8.10 (5.82-11.28) for red hair; some variants were tied to melanoma but not to red hair, suggesting non-pigmentary pathways. Observational, pooled.

10. 10
    Kaidbey KH, Agin PP, Sayre RM, Kligman AM. Photoprotection by melanin--a comparison of black and Caucasian skin.J Am Acad Dermatol · 1979 · 1(3):249-60doi:10.1016/s0190-9622(79)70018-1 · PMID 512075

    Measured ultraviolet transmission through skin samples with biological and spectroscopic methods: on average five times as much ultraviolet A and B reached the upper dermis of Caucasian skin as of Black skin, with far smaller differences at the stratum corneum; the main filter in Caucasian skin is the stratum corneum and in Black skin the living layers beneath. Melanin behaved as a neutral density filter, and the authors attribute the extra protection to melanosome packaging as well as amount. Small ex vivo human study; the abstract gives no SPF number.

11. 11
    Tadokoro T, Kobayashi N, Zmudzka BZ, et al. UV-induced DNA damage and melanin content in human skin differing in racial/ethnic origin.FASEB J · 2003 · 17(9):1177-9doi:10.1096/fj.02-0865fje · PMID 12692083

    Biopsies before and after one minimal erythema dose of ultraviolet light in volunteers of different ethnic groups: every group suffered significant DNA damage, the amount fell as constitutive melanin rose, individual variation was large, and the erythema dose predicted damage better than ethnic origin or phototype. Human in vivo.

12. 12
    Premi S, Wallisch S, Mano CM, et al. Photochemistry. Chemiexcitation of melanin derivatives induces DNA photoproducts long after UV exposure.Science · 2015 · 347(6224):842-7doi:10.1126/science.1256022 · PMID 25700512

    In melanocytes, cyclobutane pyrimidine dimers, the DNA lesions of sunlight, went on forming for more than 3 hours after ultraviolet A exposure; these dark dimers were the majority and included the cytosine-containing ones that start the signature mutations. They arise when reactive oxygen and nitrogen species excite an electron in melanin fragments to a triplet state that hands its energy to DNA without light. Cells and mice; the authors conclude melanin may be carcinogenic as well as protective.

13. 13
    Cress RD, Holly EA. Incidence of cutaneous melanoma among non-Hispanic whites, Hispanics, Asians, and blacks: an analysis of california cancer registry data, 1988-93.Cancer Causes Control · 1997 · 8(2):246-52doi:10.1023/a:1018432632528 · PMID 9134249

    California registry, 1988-93: age-adjusted melanoma incidence per 100,000 was 17.2 in men and 11.3 in women for non-Hispanic Whites, 2.8 and 3.0 for Hispanics, 0.9 and 0.8 for Asians and 1.0 and 0.7 for non-Hispanic Blacks; among men the tumor sat on the lower limb in 50% of Blacks against 9% of Whites, and late-stage diagnosis was commoner in every non-White group. Registry data from one state, now three decades old.

14. 14
    Dawes SM, Tsai S, Gittleman H, Barnholtz-Sloan JS, Bordeaux JS. Racial disparities in melanoma survival.J Am Acad Dermatol · 2016 · 75(5):983-991doi:10.1016/j.jaad.2016.06.006 · PMID 27476974

    SEER cohort of 96,953 patients, 1992-2009: White patients survived longest, then Hispanic, then Asian American/Native American/Pacific Islander, then Black; within stage I the hazard ratio for Black patients was 3.037 (95% CI 2.335-3.951) and within stage III 1.864 (1.211-2.87), and a larger share of Black patients presented at stages II-IV. Registry observational data.

15. 15
    Grønskov K, Ek J, Brondum-Nielsen K. Oculocutaneous albinism.Orphanet J Rare Dis · 2007 · 2:43doi:10.1186/1750-1172-2-43 · PMID 17980020

    Reference review: prevalence of all albinism about 1 in 17,000, so about 1 in 70 people carry a gene; OCA1A the most severe with no melanin for life, OCA1B, OCA2, OCA3 and OCA4 accumulating some pigment; nystagmus, iris translucency, foveal hypoplasia, acuity usually 20/60 to 20/400, misrouted optic nerves; four genes (TYR, OCA2, TYRP1, MATP); autosomal recessive; normal lifespan and intelligence. Review, 2007 gene list.

16. 16
    Ezzedine K, Eleftheriadou V, Whitton M, van Geel N. Vitiligo.Lancet · 2015 · 386(9988):74-84doi:10.1016/S0140-6736(14)60763-7 · PMID 25596811

    Lancet seminar: vitiligo is the commonest cause of depigmentation worldwide with an estimated prevalence of 1%, is psychologically devastating especially in darker skin, and divides into segmental and the commoner non-segmental form with different prognosis and treatment. Review.

17. 17
    Tranquillini F, Soares VG, Mendes BX, et al. Skin cancer and actinic keratosis in people with albinism: a systematic review and meta-analysis.An Bras Dermatol · 2026 · 101(4):501374doi:10.1016/j.abd.2026.501374 · PMID 42364360

    Twelve cross-sectional and cohort studies, 1,747 people with albinism: pooled prevalence of actinic keratosis 38% (95% CI 17.7-60.3) and of skin cancer 19% (13.5-24.4); within the cancers, squamous cell carcinoma 55.4%, basal cell carcinoma 33.7% and melanoma 0% (0.00-0.56). Heterogeneity was very high (I² 89-99%) and most data were cross-sectional.

18. 18
    Sjoding MW, Dickson RP, Iwashyna TJ, Gay SE, Valley TS. Racial Bias in Pulse Oximetry Measurement.N Engl J Med · 2020 · 383(25):2477-2478doi:10.1056/NEJMc2029240 · PMID 33326721

    Letter, full text read: 10,789 paired readings from 1,333 White and 276 Black inpatients at Michigan and 37,308 pairs from 7,342 White and 1,050 Black intensive-care patients at 178 hospitals; with an oximeter reading of 92-96%, arterial saturation below 88% occurred in 11.7% (95% CI 8.5-16.0) of Black and 3.6% (2.7-4.7) of White measurements in the first cohort and 17.0% (12.2-23.3) against 6.2% (5.4-7.1) in the second. Retrospective, race self-identified, no skin measurement.

19. 19
    Hughes C, Chen D, Law T, et al. Pulse Oximeter Performance and Skin Pigment: Comparison of 34 Oximeters Using Current and Emerging Regulatory Frameworks.Anesth Analg · 2026 · 143(5):1137-47doi:10.1213/ANE.0000000000008048 · PMID 42012940

    Thirty-four oximeters tested by controlled desaturation in cohorts of at least 24 with at least 25% darkly pigmented participants, pigment graded by the Monk Skin Tone scale and individual typology angle: 28 passed the 2017 ISO standard, 22 the 2013 FDA guidance, 21 the anticipated ISO criteria and 1 the anticipated FDA criteria; 11 devices read higher in dark than light skin (median difference 1.42%), and 18 could pass or fail depending on which participants were analyzed. Whether new rules will reduce disparities is, the authors say, undetermined.

20. 20
    Clemens TL, Adams JS, Henderson SL, Holick MF. Increased skin pigment reduces the capacity of skin to synthesise vitamin D3.Lancet · 1982 · 1(8263):74-6doi:10.1016/s0140-6736(82)90214-8 · PMID 6119494

    Two lightly pigmented and three heavily pigmented volunteers: one minimal erythema dose of ultraviolet raised serum vitamin D up to 60-fold within 24-48 hours in the light-skinned subjects and not significantly in the dark-skinned; a dose six times larger brought one dark-skinned subject to the same level. Five people; the origin of the 'dark skin needs more sun' teaching.

21. 21
    Young AR, Morgan KA, Ho TW, et al. Melanin has a Small Inhibitory Effect on Cutaneous Vitamin D Synthesis: A Comparison of Extreme Phenotypes.J Invest Dermatol · 2020 · 140(7):1418-1426.e1doi:10.1016/j.jid.2019.11.019 · PMID 31883961

    102 healthy volunteers of Fitzpatrick types II to VI given the same sub-erythemal ultraviolet dose to 85% of the body five times: 25-hydroxyvitamin D rose linearly with dose in every group, only type II had a steeper slope, and the melanin inhibition factor between types II and VI was about 1.3-1.4, small compared with melanin's effect on sunburn. Human experiment that revised the 1982 inference.

22. 22
    Powe CE, Evans MK, Wenger J, et al. Vitamin D-binding protein and vitamin D status of black Americans and white Americans.N Engl J Med · 2013 · 369(21):1991-2000doi:10.1056/NEJMoa1306357 · PMID 24256378

    2,085 Black and White adults in Baltimore: total 25-hydroxyvitamin D 15.6 against 25.8 ng/mL and vitamin D-binding protein 168 against 337 micrograms/mL in Black against White participants, bone density higher in Black participants (1.05 against 0.94 g/cm²), two binding-protein polymorphisms explaining 79.4% of the protein's variation, and estimated bioavailable vitamin D similar (2.9 against 3.1 ng/mL, P = 0.71). The binding-protein assay was monoclonal, which later work showed was the problem.

23. 23
    Hoofnagle AN, Eckfeldt JH, Lutsey PL. Vitamin D-Binding Protein Concentrations Quantified by Mass Spectrometry.N Engl J Med · 2015 · 373(15):1480-2doi:10.1056/NEJMc1502602 · PMID 26397952

    Letter, full text read: a liquid chromatography-tandem mass spectrometry assay measuring peptides common to all isoforms, applied to 187 genotyped ARIC participants, gave concentrations by genotype similar to a polyclonal immunoassay, while the monoclonal assay read much lower whenever the Gc1f haplotype, commoner in Black Americans, was present; the authors conclude it is unclear whether binding protein differs by race at all.

24. 24
    Nielson CM, Jones KS, Chun RF, et al. Free 25-Hydroxyvitamin D: Impact of Vitamin D Binding Protein Assays on Racial-Genotypic Associations.J Clin Endocrinol Metab · 2016 · 101(5):2226-34doi:10.1210/jc.2016-1104 · PMID 27007693

    1,057 men in the United States, United Kingdom and The Gambia: by monoclonal assay, binding protein in men of African ancestry was about 50% lower than in Whites, but by polyclonal antibody or proteomics it was not lower; directly measured free 25-hydroxyvitamin D correlated with total (r = 0.84) and was lower in African Americans. The reported racial difference is attributed to monoclonal assay bias, and total 25-hydroxyvitamin D is reinstated as the useful marker.

25. 25
    Sulzer D, Cassidy C, Horga G, et al. Neuromelanin detection by magnetic resonance imaging (MRI) and its promise as a biomarker for Parkinson's disease.NPJ Parkinsons Dis · 2018 · 4:11doi:10.1038/s41531-018-0047-3 · PMID 29644335

    Review: neuromelanin is an autophagic product of catecholamine oxidation that increases linearly with age in the substantia nigra and locus coeruleus, is lost when those neurons die in Parkinson's disease, and because it binds iron avidly gives a paramagnetic signal that MRI can detect, as Zecca's group showed about 25 years earlier. Review.

26. 26
    Cho SJ, Bae YJ, Kim JM, et al. Diagnostic performance of neuromelanin-sensitive magnetic resonance imaging for patients with Parkinson's disease and factor analysis for its heterogeneity: a systematic review and meta-analysis.Eur Radiol · 2021 · 31(3):1268-1280doi:10.1007/s00330-020-07240-7 · PMID 32886201

    Twelve studies, 403 patients and 298 controls: pooled sensitivity 89% (95% CI 86-92%) and specificity 83% (76-88%), better with disease longer than 5 or 10 years, volume rather than intensity and automated segmentation. Case-control studies against healthy controls, not against the diseases that mimic Parkinson's.

27. 27
    Langendonk JG, Balwani M, Anderson KE, et al. Afamelanotide for Erythropoietic Protoporphyria.N Engl J Med · 2015 · 373(1):48-59doi:10.1056/NEJMoa1411481 · PMID 26132941

    Two randomized placebo-controlled trials of a 16 mg subcutaneous implant every 60 days, 74 patients in Europe and 94 in the United States: pain-free sun exposure at 6 months was a median 69.4 against 40.8 hours in the US trial (P = 0.04) and at 9 months 6.0 against 0.8 hours in the European trial (P = 0.005), with 77 against 146 phototoxic reactions; adverse events mostly mild. Funded by Clinuvel.

28. 28
    U.S. Food and Drug Administration. FDA approves first treatment to increase pain-free light exposure in patients with a rare disorder (afamelanotide implant, Scenesse, for erythropoietic protoporphyria), news release, October 2019.FDA · October 2019 · news release, 'FDA approves first treatment to increase pain-free light exposure in patients with a rare disorder' (Scenesse, afamelanotide implant)

    Primary regulatory document, not indexed in PubMed; the agency's website could not be reached from the session in which this page was written, so it is cited for the approval and its month and no URL is given. Secondary summaries date the approval to 8 October 2019.

29. 29
    Minder EI, Barman-Aksoezen J, Schneider-Yin X. Pharmacokinetics and Pharmacodynamics of Afamelanotide and its Clinical Use in Treating Dermatologic Disorders.Clin Pharmacokinet · 2017 · 56(8):815-823doi:10.1007/s40262-016-0501-5 · PMID 28063031

    Review of the first alpha-melanocyte-stimulating hormone analogue, which states that afamelanotide was approved by the European Medicines Agency in 2014 for the prevention of phototoxicity in adult patients with erythropoietic protoporphyria, five years before the FDA approval, and that no late effects were reported in volunteers 25 years after first exposure. Cited here for the European authorization and its year; the authors are from a porphyria clinic and laboratory in Zurich that used the drug.

30. 30
    Rosmarin D, Passeron T, Pandya AG, et al. Two Phase 3, Randomized, Controlled Trials of Ruxolitinib Cream for Vitiligo.N Engl J Med · 2022 · 387(16):1445-1455doi:10.1056/NEJMoa2118828 · PMID 36260792

    TRuE-V1 and TRuE-V2, 674 patients aged 12 and over with non-segmental vitiligo on 10% or less of body surface: at week 24 a 75% improvement in the facial score occurred in 29.8% on 1.5% ruxolitinib cream against 7.4% on vehicle (relative risk 4.0, 95% CI 1.9-8.4) and 30.9% against 11.4% (2.7, 1.5-4.9); application-site acne in 6.3% and 6.6%. Funded by Incyte; the authors call for larger and longer trials.

31. 31
    U.S. Food and Drug Administration. Approval of ruxolitinib cream 1.5% (Opzelura) for the treatment of nonsegmental vitiligo in adult and pediatric patients 12 years of age and older, July 2022 (supplemental approval letter and prescribing information).FDA · July 2022 · supplemental approval of Opzelura (ruxolitinib) cream 1.5% for nonsegmental vitiligo in patients 12 years of age and older

    Primary regulatory documents, not indexed in PubMed; the agency's website could not be reached from the session in which this page was written, so the approval is cited for its existence and month and no URL is given. The manufacturer announced the approval on 18 July 2022; that date comes from trade reporting of the announcement rather than from the agency's own page, and some outlets give 19 July. The first drug approved for repigmentation in vitiligo.

32. 32
    Passeron T, Prajapati VH, Sivamani RK, et al. Efficacy and safety of upadacitinib in adults and adolescents for treatment of non-segmental vitiligo (Viti-Up): results of two phase 3 randomised controlled studies.Lancet · 2026 · 408(10555):636-648doi:10.1016/S0140-6736(26)00896-2 · PMID 42594914

    Two phase 3 trials, 308 and 306 patients aged 12 and over, oral upadacitinib 15 mg daily versus placebo: at week 48 a 50% total-body improvement in 19% and 21% against 6% and 6%, and a 75% facial improvement in 25% and 23% against 6% and 7%; no adjudicated major cardiovascular or thromboembolic events, and herpes zoster the only opportunistic infection. Funded by AbbVie; no systemic drug yet approved for vitiligo at publication.

33. 33
    Gudelunas MK, Lipnick M, Hendrickson C, et al. Low Perfusion and Missed Diagnosis of Hypoxemia by Pulse Oximetry in Darkly Pigmented Skin: A Prospective Study.Anesth Analg · 2024 · 138(3):552-561doi:10.1213/ANE.0000000000006755 · PMID 38109495

    Prospective laboratory study, 146 healthy volunteers (25 light, 78 medium, 43 dark by Fitzpatrick class), 9,763 matched readings on two oximeters during controlled hypoxemia: pigment, perfusion index and depth of hypoxemia all added error; under low perfusion the missed-hypoxemia rate was 1.1% for light, 8.2% for medium and 21.1% for dark skin; sex and age did not matter. Shows the mechanism behind the retrospective findings.

34. 34
    Zhang B, Ma S, Rachmin I, et al. Hyperactivation of sympathetic nerves drives depletion of melanocyte stem cells.Nature · 2020 · 577(7792):676-681doi:10.1038/s41586-020-1935-3 · PMID 31969699

    In mice, acute stress grayed hair by fast loss of melanocyte stem cells, independent of immune attack or adrenal hormones: sympathetic nerves in the niche released a burst of norepinephrine, the quiescent stem cells proliferated, differentiated, migrated and were permanently lost; blocking proliferation prevented it. Mouse study.

35. 35
    Nishimura EK, Granter SR, Fisher DE. Mechanisms of hair graying: incomplete melanocyte stem cell maintenance in the niche.Science · 2005 · 307(5710):720-4doi:10.1126/science.1099593 · PMID 15618488

    In melanocyte-tagged mice and aging human hair follicles, graying was caused by defective self-maintenance of melanocyte stem cells, accelerated by Bcl2 deficiency, with aging stem cells differentiating and pigmenting in the wrong place, a process sped by Mitf mutation. Mouse plus human follicles.

36. 36
    Sun Q, Lee W, Hu H, et al. Dedifferentiation maintains melanocyte stem cells in a dynamic niche.Nature · 2023 · 616(7958):774-782doi:10.1038/s41586-023-05960-6 · PMID 37076619

    Live imaging, single-cell sequencing and lineage tracing in mice: melanocyte stem cells toggle between stem and transit-amplifying states, move between niche compartments under local cues such as WNT, and the system is maintained by reverted cells rather than a reserved pool; with age, stranded stem cells accumulate that no longer regenerate pigment cells. Mouse study.

37. 37
    Kang HY, Picardo M. Melasma: When Dermal Photoageing Precedes Melanocyte Ageing.Exp Dermatol · 2026 · 35(9):e70364doi:10.1111/exd.70364 · PMID 42746716

    Proposes melasma as persistent melanocyte activation sustained by a photoaged dermis of senescent fibroblasts, ultraviolet-activated sebocytes and vessels, arising in a window when the dermis has aged but melanocytes still respond, which would explain midlife onset, relapse and late attenuation. Hypothesis review.

38. 38
    Kaufman BP, Aman T, Alexis AF. Postinflammatory Hyperpigmentation: Epidemiology, Clinical Presentation, Pathogenesis and Treatment.Am J Clin Dermatol · 2018 · 19(4):489-503doi:10.1007/s40257-017-0333-6 · PMID 29222629

    Review defining post-inflammatory hyperpigmentation as reactive hypermelanosis after cutaneous inflammation from acne, eczema, burns or procedures, commoner and more distressing in darker skin, treated with topicals, peels and devices. Review.

39. 39
    Mitra D, Luo X, Morgan A, et al. An ultraviolet-radiation-independent pathway to melanoma carcinogenesis in the red hair/fair skin background.Nature · 2012 · 491(7424):449-53doi:10.1038/nature11624 · PMID 23123854

    Mice carrying an inactivating Mc1r mutation and a melanocyte-targeted BRAF V600E developed invasive melanomas at high incidence with no ultraviolet exposure; adding an albino allele that abolishes all pigment was protective, and the red-hair skin carried more oxidative DNA and lipid damage than albino skin. The oncogene was switched on by tamoxifen painted on the shaved back at 6 to 10 weeks of age, and the tumors arose mostly in that treated skin. The discussion puts the estimated sun protection factor of darker skin at only SPF 2.0-4.0, against SPF 20-40 for a sunscreen, and notes that sunscreen's evidence against melanoma, unlike against squamous cell carcinoma, is weak. Mouse study, full text read; the human inference is the authors' and is discussed in the page.

40. 40
    Robles-Espinoza CD, Roberts ND, Chen S, et al. Germline MC1R status influences somatic mutation burden in melanoma.Nat Commun · 2016 · 7:12064doi:10.1038/ncomms12064 · PMID 27403562

    In sequenced human melanomas, carrying one disruptive MC1R allele was associated with an estimated 42% (95% CI 15-76%) more sun-signature C-to-T mutations, comparable to 21 additional years of age, and with enrichment of non-C-to-T mutation classes too, supporting additional mutagenic processes in red-hair carriers. Human tumor genomics; association.

41. 41
    Alaluf S, Atkins D, Barrett K, Blount M, Carter N, Heath A. Ethnic variation in melanin content and composition in photoexposed and photoprotected human skin.Pigment Cell Res · 2002 · 15(2):112-8doi:10.1034/j.1600-0749.2002.1o071.x · PMID 11936268

    Epidermal melanin from photoprotected and photoexposed skin of African, Indian, Mexican, Chinese and European volunteers: the lighter skin types held about half as much epidermal melanin and up to three times as much of the lightly colored, alkali-soluble fraction, melanosome size fell progressively from African to European skin, and spheroidal, pheomelanin-type melanosomes made up under 10% in every group, so pheomelanin is a very minor component of human epidermal melanin even in the lightest skin. Human tissue; the authors worked for Unilever Research.

42. 42
    Chen Q, Zhou D, Abdel-Malek Z, et al. Measurement of Melanin Metabolism in Live Cells by [U-13C]-L-Tyrosine Fate Tracing Using Liquid Chromatography-Mass Spectrometry.J Invest Dermatol · 2021 · 141(7):1810-1818.e6doi:10.1016/j.jid.2021.01.007 · PMID 33549605

    A method paper: traditional melanin measurement detects chemical degradation products by liquid chromatography and cannot separate synthesis from degradation; tracing labeled tyrosine measures new eumelanin and pheomelanin directly, confirmed the known differences between melanocytes of different skin colors and MC1R genotypes, and showed that different ways of changing melanosome pH drive the two pigments differently. Cell culture.

43. 43
    Zucca FA, Vanna R, Cupaioli FA, et al. Neuromelanin organelles are specialized autolysosomes that accumulate undegraded proteins and lipids in aging human brain and are likely involved in Parkinson's disease.NPJ Parkinsons Dis · 2018 · 4:17doi:10.1038/s41531-018-0050-8 · PMID 29900402

    Proteomics and lipidomics of human substantia nigra organelles: they carry lysosomal membrane and matrix proteins at lower number than true lysosomes, lipid bodies rich in dolichols, double membranes and autophagy proteins, and appear to be a slow-turnover compartment of final destination for molecules other systems cannot degrade. Postmortem human tissue.

44. 44
    Berro Z, Zucca FA, Subirana MA, et al. High-Resolution Elemental Analysis of Neuromelanin-Containing Organelles in Human Locus Coeruleus.J Neurochem · 2026 · 170(7):e70531doi:10.1111/jnc.70531 · PMID 42509598

    Seven postmortem locus coeruleus samples mapped at 5-10 nanometer resolution: metals (aluminum, calcium, iron) localized mainly to the pigment compartment identified by its sulfur-rich, pheomelanin-like portion; sulfur and iron rose with age and the right hemisphere held more aluminum, calcium and iron. Seven brains; the asymmetry is flagged for further study.

45. 45
    Lai X, Wichers HJ, Soler-Lopez M, Dijkstra BW. Structure of Human Tyrosinase Related Protein 1 Reveals a Binuclear Zinc Active Site Important for Melanogenesis.Angew Chem Int Ed Engl · 2017 · 56(33):9812-9815doi:10.1002/anie.201704616 · PMID 28661582

    Crystal structure of human TYRP1, one of three tyrosinase-like enzymes of the melanocyte: its active site holds two zinc ions rather than the copper of tyrosinase, which explains why it lacks tyrosinase activity, and most albinism mutations of the protein are predicted to affect its stability or activity. Structural biology.

46. 46
    Cui R, Widlund HR, Feige E, et al. Central role of p53 in the suntan response and pathologic hyperpigmentation.Cell · 2007 · 128(5):853-64doi:10.1016/j.cell.2006.12.045 · PMID 17350573

    Tanning needs keratinocytes to secrete alpha-melanocyte-stimulating hormone, a cleavage product of pro-opiomelanocortin; p53 directly drives that gene after ultraviolet light, p53-knockout mice do not tan, the same pathway yields beta-endorphin, and several ultraviolet-independent pigmentary conditions involve p53 mimicking the tan. Mouse and human tissue.

47. 47
    Liem EB, Lin CM, Suleman MI, et al. Anesthetic requirement is increased in redheads.Anesthesiology · 2004 · 101(2):279-83doi:10.1097/00000542-200408000-00006 · PMID 15277908

    Ten natural redheads and ten dark-haired women: the desflurane concentration needed to block movement was 6.2 vol% (95% CI 5.9-6.5) against 5.2 (4.9-5.5), P = 0.0004, and nine of ten redheads carried two MC1R mutations. Small human experiment; the authors present it as a genotype linked to inhaled anesthetic requirement.

48. 48
    Brenner M, Hearing VJ. The protective role of melanin against UV damage in human skin.Photochem Photobiol · 2008 · 84(3):539-49doi:10.1111/j.1751-1097.2007.00226.x · PMID 18435612

    Review from the National Cancer Institute laboratory: melanin is a broadband ultraviolet absorber with antioxidant and radical-scavenging properties, darker skin has lower skin-cancer incidence in many epidemiological studies, and yet the authors call the role of melanin in photoprotection still controversial. Review.

49. 49
    Napolitano A, Panzella L, Monfrecola G, d'Ischia M. Pheomelanin-induced oxidative stress: bright and dark chemistry bridging red hair phenotype and melanoma.Pigment Cell Melanoma Res · 2014 · 27(5):721-33doi:10.1111/pcmr.12262 · PMID 24814217

    Chemistry review of how pheomelanin can create a pro-oxidant state without light, including oxygen-dependent depletion of glutathione, and of pheomelanin as a living polymer that can grow and catalyze autoxidation. Review.

50. 50
    Carballo-Carbajal I, Laguna A, Romero-Giménez J, et al. Brain tyrosinase overexpression implicates age-dependent neuromelanin production in Parkinson's disease pathogenesis.Nat Commun · 2019 · 10(1):973doi:10.1038/s41467-019-08858-y · PMID 30846695

    Rats made to overexpress human tyrosinase in the substantia nigra built up human-like neuromelanin with age up to elderly human levels; above a threshold they developed hypokinesia, Lewy-body-like inclusions and nigrostriatal degeneration, which boosting lysosomal function prevented. The first rodent model that makes the pigment; the paper states that common laboratory animals lack neuromelanin, not that all animals do.

51. 51
    Schumacher JG, Zhang X, Wang J, et al. Common MC1R Variants and Parkinson Disease Progression.JAMA Neurol · 2026 · 83(10):981-989doi:10.1001/jamaneurol.2026.2998 · PMID 42709445

    809 people with Parkinson's disease in the PPMI cohort (505 MC1R loss-of-function carriers, 304 non-carriers) and 587 in three trial cohorts: among 383 with sporadic disease, carriers declined 30% faster on the motor scale (0.57 points a year, 95% CI 0.16-0.98), 50% faster in the replication cohort (1.37, 0.28-2.46), and in 53 prodromal participants carriers had a 4.75-fold (1.48-15.27) risk of conversion. Observational, European-ancestry cohorts; the pigment receptor as a prognostic marker.

52. 52
    Seiji M, Fitzpatrick TB, Birbeck MS. The melanosome: a distinctive subcellular particle of mammalian melanocytes and the site of melanogenesis.J Invest Dermatol · 1961 · 36:243-52doi:10.1038/jid.1961.42 · PMID 13749804

    The paper that named the melanosome and placed melanin synthesis inside it; the record carries no abstract, so it is cited for the finding named in its title and its date.

53. 53
    Seiji M, Shimao K, Fitzpatrick TB, Birbeck MS. The site of biosynthesis of mammalian tyrosinase.J Invest Dermatol · 1961 · 37:359-68doi:10.1038/jid.1961.132 · PMID 13910340

    The companion paper locating the synthesis of tyrosinase; the record carries no abstract, so it is cited for the finding named in its title and its date.

54. 54
    Lerner AB, McGuire JS. Effect of alpha- and betamelanocyte stimulating hormones on the skin colour of man.Nature · 1961 · 189:176-9doi:10.1038/189176a0 · PMID 13761067

    The demonstration that alpha- and beta-melanocyte-stimulating hormone darken human skin, the observation on which the melanocortin account of pigmentation rests; the record carries no abstract, so it is cited for the finding named in its title. The British spelling in the title is the title's own.

55. 55
    Loomis WF. Skin-pigment regulation of vitamin-D biosynthesis in man.Science · 1967 · 157(3788):501-6doi:10.1126/science.157.3788.501 · PMID 6028915

    Argues that the correlation of human skin color with latitude reflects selection between two dangers, rickets where ultraviolet light is weak and vitamin D toxicity where it is strong, and estimates that black skin prevents up to 95 percent of incident ultraviolet from reaching the deeper layers where vitamin D is made. A hypothesis, not a measurement; Jablonski and Chaplin tested a version of it with data in 2000.

56. 56
    Jablonski NG, Chaplin G. The evolution of human skin coloration.J Hum Evol · 2000 · 39(1):57-106doi:10.1006/jhev.2000.0403 · PMID 10896812

    Using satellite ultraviolet data, skin reflectance of indigenous peoples correlated strongly with latitude and ultraviolet level, females were lighter than males in every population with data, and the cline is argued to be a compromise between photoprotection (including of folate) and vitamin D synthesis. Correlational anthropology; the folate and vitamin D mechanisms are hypotheses the paper argues for.

57. 57
    Szabó G, Gerald AB, Pathak MA, Fitzpatrick TB. Racial differences in the fate of melanosomes in human epidermis.Nature · 1969 · 222(5198):1081-2doi:10.1038/2221081a0 · PMID 5787098

    An early electron-microscopy paper reporting that melanosomes are packaged differently in the keratinocytes of darker and lighter skin; the record carries no abstract, so it is cited for the finding named in its title and its date.

58. 58
    Holick MF, MacLaughlin JA, Doppelt SH. Regulation of cutaneous previtamin D3 photosynthesis in man: skin pigment is not an essential regulator.Science · 1981 · 211(4482):590-3doi:10.1126/science.6256855 · PMID 6256855

    Human skin exposed to simulated solar ultraviolet light: the formation of previtamin D3 plateaus at about 10-15% of the skin's 7-dehydrocholesterol, after which further light converts the product to inert isomers; more melanin, like a higher latitude, lengthened the time needed to reach the maximum but did not change the total formed, and the authors rank the limiting factors as photochemical regulation first, pigmentation second, latitude third. Published a year before the five-person study that became the teaching.

59. 59
    Ito S, Fujita K. Microanalysis of eumelanin and pheomelanin in hair and melanomas by chemical degradation and liquid chromatography.Anal Biochem · 1985 · 144(2):527-36doi:10.1016/0003-2697(85)90150-2 · PMID 3993914

    The degradation assay the field still uses: permanganate oxidation of eumelanin yields pyrrole-2,3,5-tricarboxylic acid and hydriodic acid hydrolysis of pheomelanin yields aminohydroxyphenylalanine, both quantified by liquid chromatography, in 2 mg of hair or 5 mg of tissue and without isolating the melanin. Method paper.

60. 60
    Fitzpatrick TB. The validity and practicality of sun-reactive skin types I through VI.Arch Dermatol · 1988 · 124(6):869-71doi:10.1001/archderm.124.6.869 · PMID 3377516

    Fitzpatrick's own account of the six sun-reactive types, originally a tool for dosing ultraviolet treatment; the record carries no abstract and its subject headings are psoralen-ultraviolet A therapy, psoriasis, erythema and sunburn, so it is cited for the scale's authorship, date and stated purpose.

61. 61
    Jubran A, Tobin MJ. Reliability of pulse oximetry in titrating supplemental oxygen therapy in ventilator-dependent patients.Chest · 1990 · 97(6):1420-5doi:10.1378/chest.97.6.1420 · PMID 2347228

    54 ventilator-dependent patients: an oximeter target of 92% reliably indicated adequate oxygenation in White patients, but in Black patients such a reading was commonly associated with an arterial oxygen tension as low as 49 mm Hg, so a 95% target was required; readings differed from directly measured arterial saturation by more than 4% in 27% of Black against 11% of White patients. The clinical report of the bias, fifteen years before the 2005 laboratory study.

62. 62
    Mountjoy KG, Robbins LS, Mortrud MT, Cone RD. The cloning of a family of genes that encode the melanocortin receptors.Science · 1992 · 257(5074):1248-51doi:10.1126/science.1325670 · PMID 1325670

    Cloning of the melanocortin receptor family, including the mouse and human melanocyte-stimulating hormone receptor now called MC1R; without it the human variants reported in 1995 could not have been sought. Molecular biology.

63. 63
    Bickler PE, Feiner JR, Severinghaus JW. Effects of skin pigmentation on pulse oximeter accuracy at low saturation.Anesthesiology · 2005 · 102(4):715-9doi:10.1097/00000542-200504000-00004 · PMID 15791098

    Controlled desaturation in 11 darkly and 10 lightly pigmented volunteers with three finger oximeters: at 60-70% arterial saturation the oximeters overestimated by 3.56 ± 2.45% in dark skin against 0.37 ± 3.20% in light skin (P < 0.0001), with the bias rising roughly in proportion to desaturation and differing by device. The laboratory warning fifteen years before the clinical data.

64. 64
    Lamason RL, Mohideen MA, Mest JR, et al. SLC24A5, a putative cation exchanger, affects pigmentation in zebrafish and humans.Science · 2005 · 310(5755):1782-6doi:10.1126/science.1116238 · PMID 16357253

    The golden zebrafish mutant has fewer, smaller, less dense melanosomes, like light human skin; the gene is a cation exchanger on an internal membrane, the human version works in fish, its ancestral allele predominates in African and East Asian populations and the variant is nearly fixed in Europeans and tracks lighter skin in admixed people. Fish genetics plus human association.

65. 65
    Crawford NG, Kelly DE, Hansen MEB, et al. Loci associated with skin pigmentation identified in African populations.Science · 2017 · 358(6365):eaan8433doi:10.1126/science.aan8433 · PMID 29025994

    Genome-wide association in ethnically diverse Africans: the light variant at one locus entered East Africa by gene flow from non-Africans, while at the other loci the dark-pigmentation variants are identical by descent in South Asian and Australo-Melanesian populations; one gene encodes a lysosomal protein that changes melanogenesis in zebrafish and mice. Human genetics with functional follow-up.

66. 66
    Sasaki M, Shibata E, Tohyama K, et al. Neuromelanin magnetic resonance imaging of locus ceruleus and substantia nigra in Parkinson's disease.Neuroreport · 2006 · 17(11):1215-8doi:10.1097/01.wnr.0000227984.84927.a7 · PMID 16837857

    First neuromelanin-sensitive MRI at 3 tesla: high-signal areas matched the locus coeruleus and substantia nigra pars compacta of gross specimens, and in Parkinson's disease the signal in both was greatly reduced. Small imaging series; the technique's origin.

67. 67
    Lalueza-Fox C, Römpler H, Caramelli D, et al. A melanocortin 1 receptor allele suggests varying pigmentation among Neanderthals.Science · 2007 · 318(5855):1453-5doi:10.1126/science.1147417 · PMID 17962522

    Two Neanderthal specimens carried an MC1R mutation absent from about 3,700 modern humans; in functional assays it reduced receptor activity to a level that alters hair or skin color in humans, so inactive MC1R variants evolved independently in both lineages. Ancient DNA plus cell assays.

68. 68
    Eilers S, Bach DQ, Gaber R, et al. Accuracy of self-report in assessing Fitzpatrick skin phototypes I through VI.JAMA Dermatol · 2013 · 149(11):1289-94doi:10.1001/jamadermatol.2013.6101 · PMID 24048361

    270 ethnically diverse participants against reflectance spectrophotometry: 42% (114 of 270) gave burning and tanning answers that could not be classified by the standard definitions; dermatologist-assigned types separated III from IV and IV from V on the melanin index, self-assigned types separated III from IV but not IV from V (P = 0.90), and tanning answers did not correlate with the dermatologist's type. Observational.

69. 69
    Wensink D, Wagenmakers MAEM, Barman-Aksözen J, et al. Association of Afamelanotide With Improved Outcomes in Patients With Erythropoietic Protoporphyria in Clinical Practice.JAMA Dermatol · 2020 · 156(5):570-575doi:10.1001/jamadermatol.2020.0352 · PMID 32186677

    Post-authorization cohort required by the European Medicines Agency, 117 patients at one center over a median 2.0 years: 98% continued, time outside rose by 6.1 hours a week (95% CI 3.62-8.67), quality-of-life score by 14.01%, phototoxic reactions were less painful but not fewer or shorter; adverse events minor. Uncontrolled real-world cohort.

70. 70
    Cassidy CM, Zucca FA, Girgis RR, et al. Neuromelanin-sensitive MRI as a noninvasive proxy measure of dopamine function in the human brain.Proc Natl Acad Sci U S A · 2019 · 116(11):5108-5117doi:10.1073/pnas.1807983116 · PMID 30796187

    Neuromelanin MRI signal in postmortem midbrain correlated with measured neuromelanin even without neurodegeneration; in living people a voxelwise signal related to dopamine release in the dorsal striatum on PET and to psychosis severity in schizophrenia and at-risk individuals. Validation plus small clinical correlations.

71. 71
    U.S. Food and Drug Administration. Pulse Oximeter Accuracy and Limitations: FDA Safety Communication (February 2021); Anesthesiology and Respiratory Therapy Devices Panel meeting on pulse oximeters and skin pigmentation (November 2022); draft guidance, Pulse Oximeters for Medical Purposes: Non-Clinical and Clinical Performance Testing, Labeling, and Premarket Submission Recommendations (January 2025).FDA · 19 February 2021 (safety communication); 1 November 2022 (Anesthesiology and Respiratory Therapy Devices Panel meeting); January 2025 (draft guidance) · regulatory documents

    Primary regulatory documents, not indexed in PubMed. The agency's website could not be reached from the session in which this page was written, so no URL is given and the documents are cited for their existence and dates. The 19 February 2021 and 1 November 2022 dates are confirmed by secondary reporting; the day in January 2025 on which the draft guidance issued could not be verified, so only the month is given. the content of the anticipated criteria is taken from the 2026 Hughes paper that tested devices against them.

72. 72
    Wong AI, Charpignon M, Kim H, et al. Analysis of Discrepancies Between Pulse Oximetry and Arterial Oxygen Saturation Measurements by Race and Ethnicity and Association With Organ Dysfunction and Mortality.JAMA Netw Open · 2021 · 4(11):e2131674doi:10.1001/jamanetworkopen.2021.31674 · PMID 34730820

    87,971 encounters across 215 hospitals: hidden hypoxemia (oximeter 88% or more, artery below 88%) in 5.5% overall, 6.8% of Black, 6.0% of Hispanic, 4.8% of Asian and 4.9% of White patients, and it was followed by higher organ-failure scores and higher in-hospital mortality (among Black patients 21.1% against 15.0%). Retrospective, race self-identified.

73. 73
    Fawzy A, Wu TD, Wang K, et al. Racial and Ethnic Discrepancy in Pulse Oximetry and Delayed Identification of Treatment Eligibility Among Patients With COVID-19.JAMA Intern Med · 2022 · 182(7):730-738doi:10.1001/jamainternmed.2022.1906 · PMID 35639368

    1,216 COVID-19 patients with 32,282 concurrent pairs: occult hypoxemia in 28.5% of Black, 29.8% of non-Black Hispanic and 30.2% of Asian against 17.2% of White patients; oximetry overestimated arterial saturation by 1.2% (95% CI 0.6-1.9) in Black patients; Black patients had a 29% lower hazard of having treatment eligibility recognized (0.71, 0.63-0.80) and a median delay of 1.0 hour, and 54.8% of those never recognized were Black. Retrospective cohort.

74. 74
    Balwani M, Bonkovsky HL, Levy C, et al. Dersimelagon in Erythropoietic Protoporphyrias.N Engl J Med · 2023 · 388(15):1376-1385doi:10.1056/NEJMoa2208754 · PMID 37043653

    Phase 2, 102 patients (93 erythropoietic and 9 X-linked protoporphyria) on an oral MC1R agonist for 16 weeks: time to first symptom in sunlight lengthened by 53.8 minutes at 100 mg (P = 0.008) and 62.5 minutes at 300 mg (P = 0.003) against placebo; nausea, freckles, headache and hyperpigmentation were the commonest adverse events. Funded by Mitsubishi Tanabe; phase 2 only.

75. 75
    Demay MB, Pittas AG, Bikle DD, et al. Vitamin D for the Prevention of Disease: An Endocrine Society Clinical Practice Guideline.J Clin Endocrinol Metab · 2024 · 109(8):1907-1947doi:10.1210/clinem/dgae290 · PMID 38828931

    GRADE-based guideline: no clinical-trial evidence supports routine 25-hydroxyvitamin D screening in the general population, in obesity or in people with dark complexion, so the panel suggests against routine testing in all populations considered, while suggesting empiric supplementation for children, adults over 75, pregnancy and high-risk prediabetes. Guideline; optimal doses called unclear.

76. 76
    Grant S, Revan D, Tang L, et al. Disparities in acral lentiginous melanoma: Factors beyond delayed diagnosis.J Am Acad Dermatol · 2025 · 93(4):1018-1026doi:10.1016/j.jaad.2025.06.048 · PMID 40562077

    SEER analysis finding no significant difference in disease severity at diagnosis between Black and White patients with acral melanoma, yet a doubled risk of melanoma death in Black patients (hazard ratio 2.04, 95% CI 1.31-3.11) independent of clinicopathological and socioeconomic covariates; the authors say this counters the assumption that the disparity is mainly delayed diagnosis. Registry data; the authors note limited data outside the United States and no large genetic studies.

77. 77
    Mohri Y, Nie J, Morinaga H, et al. Antagonistic stem cell fates under stress govern decisions between hair greying and melanoma.Nat Cell Biol · 2025 · 27(10):1647-1659doi:10.1038/s41556-025-01769-9 · PMID 41053225

    Melanocyte stem cell fate tracking in mice: the stem cells answer DNA double-strand breaks with senescence-coupled differentiation, which depletes them, grays the hair and protects against melanoma, while carcinogens suppress that response even in cells carrying double-strand breaks, by activating arachidonic acid metabolism and niche-derived KIT ligand, so the damaged cells renew themselves instead. Mouse study; graying and melanoma as antagonistic outcomes of one injury.

78. 78
    Nordlund JJ. The melanocyte and the epidermal melanin unit: an expanded concept.Dermatol Clin · 2007 · 25(3):271-81doi:10.1016/j.det.2007.04.001 · PMID 17662893

    Review of the epidermal melanin unit, the melanocyte with the keratinocytes it supplies, arguing it should include Langerhans cells and be called the KLM unit because inflammatory mediators and melanogenic hormones act on all three. Review.

79. 79
    Tadokoro T, Yamaguchi Y, Batzer J, et al. Mechanisms of skin tanning in different racial/ethnic groups in response to ultraviolet radiation.J Invest Dermatol · 2005 · 124(6):1326-32doi:10.1111/j.0022-202X.2005.23760.x · PMID 15955111

    Human skin of several racial backgrounds before and up to a week after one minimal erythema dose: melanocyte density at the junction of epidermis and dermis was remarkably similar across groups and did not change significantly after exposure; the melanocyte-specific proteins rose in expression, total melanin rose only slightly, and the largest change was the movement of melanin from the lower to the middle layer of the epidermis, more dramatic in darker skin. Human in vivo.

80. 80
    Onojafe IF, Megan LH, Melch MG, et al. Minimal Efficacy of Nitisinone Treatment in a Novel Mouse Model of Oculocutaneous Albinism, Type 3.Invest Ophthalmol Vis Sci · 2018 · 59(12):4945-4952doi:10.1167/iovs.16-20293 · PMID 30347088

    In a Tyrp1-null mouse, nitisinone raised plasma tyrosine about 7-fold but changed the color of neither fur nor pigmented ocular structures after a month, and the only significant difference was in the number of pigmented melanosomes in the iris, so the authors judge the drug unlikely to be therapeutic in OCA3; it had earlier raised pigment in a mouse model of OCA1B. Animal study.

81. 81
    Mujahid N, Liang Y, Murakami R, et al. A UV-Independent Topical Small-Molecule Approach for Melanin Production in Human Skin.Cell Rep · 2017 · 19(11):2177-2184doi:10.1016/j.celrep.2017.05.042 · PMID 28614705

    Small-molecule inhibitors of salt-inducible kinase, optimized to cross human epidermis, raised MITF and induced pigment in Mc1r-deficient mice and in normal human skin explants, a route to a eumelanin tan without ultraviolet light. Laboratory and explant work; no clinical trial.

82. 82
    Sánchez AG, Gabrielli A, Keszenman DJ. Impact of ecological UV radiation on the photochemistry of nuclear DNA.Biophys Rev · 2025 · 17(2):537-545doi:10.1007/s12551-025-01275-0 · PMID 40376424

    2025 review of ultraviolet DNA photochemistry: ultraviolet upregulates nitric oxide synthase and NADPH oxidase in melanocytes, their products combine to peroxynitrite, and this oxidant drives melanin chemiexcitation that yields dimers in the dark hours after exposure ends. Review.

83. 83
    Srivastava J, Cooke MS, Premi S. Revisiting the action spectrum of DNA damage: Chemiexcitation and the expanded biological spectrum of cyclobutane pyrimidine dimer formation.Photochem Photobiol · 2026 · online ahead of printdoi:10.1111/php.70154 · PMID 42779255

    Argues that the action spectrum for dimer formation is no longer the DNA absorption spectrum alone, because ultraviolet A and redox chemistry make dimers after exposure, so erythema is an incomplete surrogate for mutagenic risk and sunscreens should address indirect damage. Perspective, 2026.

84. 84
    Srivastava J, Premi S. Melanin as a Redox Hub: Linking Energy Flow to Genome Stability and Immune Control.Antioxid Redox Signal · 2026 · 44(10-12):618-635doi:10.1177/15230864261421542 · PMID 41680087

    2026 review by the group that found dark dimers, framing melanin as a redox-active system rather than a passive filter and proposing nitric oxide synthase inhibition and triplet-state quenching as interventions; it admits that how pigment oxidation state and enzyme flux coordinate in vivo remains a gap. Review by authors with a stake in the hypothesis.

85. 85
    Waresiak J, Pięta E, Sagan F, et al. FTIR-based characterization of photodegradation in synthetic eumelanin and pheomelanin models supported by molecular electrostatic potential maps.J Photochem Photobiol B · 2026 · 283:113548doi:10.1016/j.jphotobiol.2026.113548 · PMID 42659740

    Synthetic eumelanin and pheomelanin bleached under aerobic light, losing radical content and hydroxyl and amino groups; pheomelanin shifted from benzothiazine to benzothiazole rings and lost intermolecular interactions, eumelanin became more conjugated. In vitro chemistry on model pigments, not skin.

86. 86
    Galván I, García-Guerra M, Araujo-Roque M. MC1R depalmitoylation inhibition reveals a physiological role for pheomelanin.PNAS Nexus · 2026 · 5(1):pgaf391doi:10.1093/pnasnexus/pgaf391 · PMID 41502961

    Male zebra finches given dietary cysteine together with ML349, an inhibitor of the enzyme that removes palmitate from MC1R and so blocks pheomelanin synthesis, showed impaired pheomelanin-based feather pigmentation and raised systemic oxidative damage in the pheomelanin-producing follicles but not in the eumelanin-producing ones; females, which make no pheomelanin, were unaffected. Bird study; evidence that pheomelanin synthesis serves cysteine homeostasis.

87. 87
    Galván I, Martínez-González J, Gómez-Vicioso J, et al. Melanin pigments protect nucleic acid integrity.J R Soc Interface · 2026 · 23(236)doi:10.1098/rsif.2025.1111 · PMID 41818795

    DNA-melanin and RNA-melanin complexes formed under physiological conditions from chicken follicular melanocytes: eumelanin protected DNA from strand breaks and from oxidative modification measured as 8-hydroxy-2'-deoxyguanosine, and pheomelanin protected RNA. In vitro, and the opposite sign from the ultraviolet-driven damage reported elsewhere on this page.

88. 88
    Nazzal Y, Najeeb M, Dagher J, et al. Vitiligo and skin cancer risk across racial/ethnic groups: A retrospective propensity score-matched study on 123,179 patients.Eur J Cancer · 2026 · 245:116949doi:10.1016/j.ejca.2026.116949 · PMID 42447511

    123,179 patients with vitiligo in a United States network of health records, propensity-matched to as many comparators: lower risk of any skin cancer (hazard ratio 0.83, 95% CI 0.78-0.88) and of melanoma (0.69, 0.58-0.81), with an exploratory subgroup suggesting higher squamous cell carcinoma risk in Black patients with vitiligo (2.16, 1.09-4.28). Retrospective and drawn from records, so an association; people with vitiligo are also watched and counseled about the sun more than their comparators.

89. 89
    Carter TM, Strassle PD, Ollila DW, et al. Does acral lentiginous melanoma subtype account for differences in patterns of care in Black patients?Am J Surg · 2021 · 221(4):706-711doi:10.1016/j.amjsurg.2020.12.040 · PMID 33461732

    National Cancer Database, 251,864 patients of whom 1,453 were Black: Black patients had 27.6-fold higher odds (95% CI 24.4-31.2) of an acral melanoma than non-Hispanic Whites, and still had higher odds of major amputation at every stage after adjusting for acral histology. Observational.

90. 90
    Huang K, Fan J, Misra S. Acral Lentiginous Melanoma: Incidence and Survival in the United States, 2006-2015, an Analysis of the SEER Registry.J Surg Res · 2020 · 251:329-339doi:10.1016/j.jss.2020.02.010 · PMID 32208196

    SEER 2006-2015: acral lentiginous melanoma had an incidence of 2.0 per million person-years, made up 32.6% of melanomas in Black patients, and carried a 5-year melanoma-specific survival of 80.6% against 93.0% for other cutaneous melanoma; by race it was 84.3% in non-Hispanic Whites and 66.9% in Blacks, who presented older, with thicker and more ulcerated tumors. Registry data.

91. 91
    Ansbro B, Riley JM. Trends by race and ethnicity in incidence and mortality of acral lentiginous melanoma: analysis of Surveillance, Epidemiology, and End Results 2000-2020.Arch Dermatol Res · 2024 · 316(7):456doi:10.1007/s00403-024-03167-x · PMID 38967822

    4,188 acral melanomas, 2000-2020: age-adjusted incidence rose 2.48% a year, driven by Hispanics and non-Hispanic Whites and stable in Black and Asian or Pacific Islander patients; adjusted hazard of acral-melanoma death was 2.19 (95% CI 1.74-2.76) in Black, 1.58 (1.23-2.04) in Asian or Pacific Islander and 1.38 (1.14-1.66) in Hispanic patients against Whites. Registry letter.

92. 92
    Burge P, Madaree A. Primary cutaneous malignancies in black African patients with oculocutaneous albinism in KwaZulu-Natal Province, South Africa.S Afr Med J · 2026 · 116(2):e3821doi:10.7196/SAMJ.2026.v116i2.3821 · PMID 42246818

    Twenty years of one tertiary center: 221 people with albinism contributed 618 histologically confirmed skin cancers, 59.7% presenting before age 40, with squamous (50.2%) and basal cell (49.7%) carcinoma nearly equal and a single melanoma. Retrospective single-center series.

93. 93
    Hayward NK, Wilmott JS, Waddell N, et al. Whole-genome landscapes of major melanoma subtypes.Nature · 2017 · 545(7653):175-180doi:10.1038/nature22071 · PMID 28467829

    Whole-genome sequencing of cutaneous, acral and mucosal melanoma: the heavily mutated cutaneous tumors resolved novel ultraviolet mutation signatures, while acral and mucosal melanomas were dominated by structural changes and by mutation signatures of unknown cause not previously identified in melanoma, some of the carcinogenic processes being unrelated to sun exposure. Human tumor genomics; the reason acral melanoma is not treated on this page as a disease of sunlight.

94. 94
    Adams DR, Menezes S, Jauregui R, et al. One-year pilot study on the effects of nitisinone on melanin in patients with OCA-1B.JCI Insight · 2019 · 4(2):e124387doi:10.1172/jci.insight.124387 · PMID 30674731

    Open-label pilot, five adults with OCA-1B on oral nitisinone 2 mg a day for 12 months: iris transillumination did not change (mean 1.0 ± 1.54 points), letters read improved by 4.2 (right eye) and 5 (left eye), inner-arm skin pigment rose (M index 1.72, 95% CI 0.03-3.41) and hair pigment rose by reflectometry (M index 17.3, 4.4-30.2) and chemically. Five patients, no control group; National Eye Institute intramural.

95. 95
    Summers CG, Connett JE, Holleschau AM, et al. Does levodopa improve vision in albinism? Results of a randomized, controlled clinical trial.Clin Exp Ophthalmol · 2014 · 42(8):713-21doi:10.1111/ceo.12325 · PMID 24641678

    45 people with albinism, mean age 14.5 years, every one carrying a mutation in a known albinism gene, randomized double-masked to levodopa 0.76 mg/kg, levodopa 0.51 mg/kg or placebo for 20 weeks with 100% follow-up: no statistically significant improvement in best-corrected visual acuity at either dose, and no serious adverse events. The one adequately powered randomized trial in albinism, and it is negative.

96. 96
    Seid K, Jaleta CG, Lorato SS, et al. Prevalence of vitiligo in East Africa: systematic review and meta-analysis.Skin Health Dis · 2026 · 6(5):680-689doi:10.1093/skinhd/vzag069 · PMID 42813184

    Eleven studies, 8,444 people reported between 2000 and 2025: pooled vitiligo prevalence 6% (95% CI 3-8); 3% in children, 9% in adults, 7% in general-population samples; 1% in Tanzania and 8% in Ethiopia. Few studies, mostly Ethiopian, and not directly comparable with the 1% global estimate.

97. 97
    Rashighi M, Agarwal P, Richmond JM, et al. CXCL10 is critical for the progression and maintenance of depigmentation in a mouse model of vitiligo.Sci Transl Med · 2014 · 6(223):223ra23doi:10.1126/scitranslmed.3007811 · PMID 24523323

    Lesional skin from vitiligo patients showed an interferon-gamma signature including CXCL10, raised in skin and serum, with its receptor CXCR3 on the pathogenic T cells; in mice, blocking or deleting CXCL10 prevented depigmentation and neutralizing it in established disease produced repigmentation. Human tissue plus mouse model; the mechanistic basis of the Janus kinase drugs.

98. 98
    Passeron T, Lebwohl M, Lynde C, et al. Vitiligo Biomarker CXCL10 Correlates With Clinical Response to Ruxolitinib Cream in the Phase 2 Randomized TRuE-V Mechanism-of-Action Study.J Invest Dermatol · 2026 · online ahead of printdoi:10.1016/j.jid.2026.09.857 · PMID 42869214

    Phase 2, 41 patients on cream and 19 on vehicle for 24 weeks: serum interferon-gamma-related proteins including CXCL10 fell at every time point, CXCL10 in serum and lesional skin fell from week 12 to 40, facial scores fell 32.9% at week 12 and 49.1% at week 24, and the fall in CXCL10 correlated with total-score improvement. Mechanism study funded by Incyte.

99. 99
    Pandya AG, Lynde CW, Shayesteh Alam M, et al. Efficacy and safety of ruxolitinib cream combined with narrow-band UVB phototherapy for treatment of vitiligo.J Invest Dermatol · 2025 (online) · 146(7):1867-1875.e4doi:10.1016/j.jid.2025.11.016 · PMID 41386328

    Exploratory phase 2, 55 patients: 35 of 55 (63.6%) had less than 25% improvement after 12 weeks of cream alone and added narrow-band ultraviolet B; their mean total score change went from -7.9% at week 12 to -50.4% at week 48, with 57.7% (15 of 26) reaching a 50% improvement. Uncontrolled add-on; Incyte employees among the authors.

100. 100
     Passeron T, Pandya AG, Ezzedine K, et al. Post Hoc Subgroup and Body Region Analyses of VASI Outcomes Among Patients with Vitiligo in a Phase 2 Povorcitinib Trial.Adv Ther · 2026 · 43(10):4602-4616doi:10.1007/s12325-026-03673-6 · PMID 42440041

     Post hoc analysis of 103 patients on oral povorcitinib for 52 weeks: 34.0% reached a 50% total-body improvement and 45.6% a 75% facial improvement, across age, sex, race and Fitzpatrick type, with response rates by region from 25.8% for the feet to 57.3% for the head and neck. Post hoc, uncontrolled extension; Incyte authors.

101. 101
     Lukic T, Ghosh P, Napatalung L, et al. Description of the Tranquillo Phase 3 Clinical Trial Designs/Study Protocols to Assess Ritlecitinib in Adults and Adolescents with Nonsegmental Vitiligo.Dermatol Ther (Heidelb) · 2026 · 16(7):3709-3727doi:10.1007/s13555-026-01724-y · PMID 42171987

     Protocol paper for three phase 3 studies of oral ritlecitinib 50 and 100 mg (about 600, 400 and 1,450 participants) with a 75% facial improvement at week 52 as primary end point. Design only, no results; Pfizer authors.

102. 102
     Rosenberg AM, Rausser S, Ren J, et al. Quantitative mapping of human hair greying and reversal in relation to life stress.eLife · 2021 · 10:e67437doi:10.7554/eLife.67437 · PMID 34155974

     Profiling pigment along single human hairs showed white or gray hairs that naturally regained pigment across sexes, ethnicities, ages and body sites, gray hairs with raised mitochondrial and antioxidant proteins, and graying and reversal running in parallel with psychological stressors; a model suggests a threshold mechanism. Human observational.

103. 103
     Ogawa T, Chen M, Lee SH, Gay DL, Ito M. Dissecting the boundary of quiescence and activation of murine melanocyte stem cells in the hair follicle niche.NPJ Regen Med · 2026 · 11(1)doi:10.1038/s41536-026-00460-3 · PMID 41932944

     2026 review from the group behind the 2023 finding, on quiescence and activation of melanocyte stem cells and their interaction with hair follicle stem cells, offered as a model of cellular aging. Review of mouse work.

104. 104
     Yang W, Meng X, Liu T, et al. The Exposome in Melasma: A Comprehensive Review of Etiology, Mechanisms, and Implications for Management.Clin Cosmet Investig Dermatol · 2026 · 19:594944doi:10.2147/CCID.S594944 · PMID 41930363

     Review organizing melasma triggers as endogenous (hormones, oxidative stress, thyroid, vitamin D metabolism, stress) and exogenous (ultraviolet, visible light, pollution, drugs, skincare) converging on MITF, oxidative cycles and the pro-opiomelanocortin pathway; argues traditional therapy treats formed pigment and relapses. Narrative review.

105. 105
     Leung JH, Leung HWC, Wang SY, Jang YC, Chan ALF. Efficacy and Safety of Different Treatments for Melasma: Network Meta-Analysis of Updated Data.Diseases · 2025 · 13(10):316doi:10.3390/diseases13100316 · PMID 41149049

     Network meta-analysis of 14 randomized trials (738 women, 15 modalities, 2022-2025): intradermal platelet-rich plasma alone or with oral tranexamic acid ranked best on severity index against hydroquinone 4%, intradermal tranexamic acid and laser combinations; most adverse events mild. Small trials; the first author is at an aesthetic clinic.

106. 106
     Perper M, Eber AE, Fayne R, et al. Tranexamic Acid in the Treatment of Melasma: A Review of the Literature.Am J Clin Dermatol · 2017 · 18(3):373-381doi:10.1007/s40257-017-0263-3 · PMID 28283893

     Review of 15 studies of tranexamic acid for melasma which states the mechanism as inhibition of ultraviolet-induced plasmin activity in keratinocytes, and finds the drug about as effective as standard therapies with fewer side effects. A review of small studies; cited here for the mechanism.

107. 107
     Mahmoud BH, Ruvolo E, Hexsel CL, et al. Impact of long-wavelength UVA and visible light on melanocompetent skin.J Invest Dermatol · 2010 · 130(8):2092-7doi:10.1038/jid.2010.95 · PMID 20410914

     20 volunteers of skin types IV-VI irradiated on the lower back with long-wave ultraviolet A (340-400 nm) or visible light (400-700 nm): both produced pigmentation, but the pigmentation from visible light was darker and more sustained and differed in quality and quantity; no pigmentation appeared in type II skin. Small human experiment; the basis of visible-light photoprotection.

108. 108
     Castanedo-Cazares JP, Hernandez-Blanco D, Carlos-Ortega B, Fuentes-Ahumada C, Torres-Álvarez B. Near-visible light and UV photoprotection in the treatment of melasma: a double-blind randomized trial.Photodermatol Photoimmunol Photomed · 2014 · 30(1):35-42doi:10.1111/phpp.12086 · PMID 24313385

     68 melasma patients randomized to a sunscreen containing iron oxide, which absorbs visible light, or a conventional ultraviolet-only sunscreen of protection factor 50 or more, both groups also using 4% hydroquinone for 8 weeks; 61 finished, and the visible-light group did better by 15% on the severity index, 28% on colorimetry and 4% on measured melanin. Randomized, double-blind, single center.

109. 109
     Boukari F, Jourdan E, Fontas E, et al. Prevention of melasma relapses with sunscreen combining protection against UV and short wavelengths of visible light: a prospective randomized comparative trial.J Am Acad Dermatol · 2015 · 72(1):189-90.e1doi:10.1016/j.jaad.2014.08.023 · PMID 25443629

     Randomized comparison of a sunscreen protecting against ultraviolet and short-wavelength visible light with an ultraviolet-only sunscreen for the prevention of melasma relapse. Published as a letter and the record carries no abstract, so it is cited for the comparison named in its title; one author works for the sunscreen manufacturer.

110. 110
     Bulengo-Ransby SM, Griffiths CE, Kimbrough-Green CK, et al. Topical tretinoin (retinoic acid) therapy for hyperpigmented lesions caused by inflammation of the skin in black patients.N Engl J Med · 1993 · 328(20):1438-43doi:10.1056/NEJM199305203282002 · PMID 8479462

     54 Black patients completed 40 weeks of randomized, double-blind, vehicle-controlled treatment of post-inflammatory hyperpigmentation with 0.1% tretinoin cream or vehicle: treated lesions were significantly lighter clinically (P < 0.001), colorimetry showed 40% lightening toward normal skin color against 18% on vehicle (P = 0.05), epidermal melanin fell 23% against 3% (P = 0.24), normal skin lightened minimally, and retinoid dermatitis occurred in 12 of the 24 treated patients and diminished as the study went on. The oldest good trial in the condition.

111. 111
     Hakozaki T, Minwalla L, Zhuang J, et al. The effect of niacinamide on reducing cutaneous pigmentation and suppression of melanosome transfer.Br J Dermatol · 2002 · 147(1):20-31doi:10.1046/j.1365-2133.2002.04834.x · PMID 12100180

     Niacinamide did not inhibit tyrosinase or melanin production in cultured melanocytes, but inhibited melanosome transfer by 35-68% in a keratinocyte-melanocyte coculture, reduced pigmentation in a reconstructed pigmented epidermis, and in 18 subjects with facial hyperpigmentation reduced it against vehicle over 4 weeks, with a second clinical study of 120 subjects testing 2% niacinamide with sunscreen. Laboratory work with small, short clinical arms, done at Procter and Gamble.

112. 112
     Wongdama S, Yenyuwadee S, Li JB, et al. Interventions to Prevent Postinflammatory Hyperpigmentation After Laser and Energy-Based Device Treatments: A Systematic Review and Network Meta-Analysis.Lasers Surg Med · 2026 · 58(3):157-168doi:10.1002/lsm.70098 · PMID 41535730

     Fourteen randomized trials, 11 in the network: intradermal tranexamic acid (risk ratio 0.02, 95% CI 0.00-0.53), topical corticosteroids, vasoconstrictors, oral tranexamic acid and epidermal cooling all beat sunscreen alone at preventing laser-induced hyperpigmentation, and sunscreen alone was no better than placebo. Small trials; the authors ask for caution.

113. 113
     Minasyan M, Hogan S, Lal K. Oral Tranexamic Acid for Prevention and Treatment of Postinflammatory Hyperpigmentation.Dermatol Surg · 2024 · 50(12S):S219-S224doi:10.1097/DSS.0000000000004400 · PMID 39347626

     Narrative review of oral tranexamic acid, an antifibrinolytic borrowed from bleeding control, for post-inflammatory hyperpigmentation: the studies reviewed all reported improvement without serious adverse effects. A review of small studies, not a trial.

114. 114
     Blaney G, Fantini S. Broad-linewidth sources result in a skin-tone bias in noninvasive optical measurement of oxygen saturation.J Biomed Opt · 2026 · 31(7):070501doi:10.1117/1.JBO.31.7.070501 · PMID 42428948

     Physics model (diffusion theory and Monte Carlo) in which oximeter calibration curves depend on melanin concentration when broad-spectrum light-emitting diodes are used and become melanin-independent with monochromatic laser diodes; proposed as one contribution to the bias and one way to remove it. Simulation, not a device trial.

115. 115
     Martin DS, Doidge JC, Gould D, et al. The impact of skin tone on performance of pulse oximeters used by NHS England COVID Oximetry @home scheme: measurement and diagnostic accuracy study.BMJ · 2026 · 392:e085535doi:10.1136/bmj-2025-085535 · PMID 41534914

     903 critically ill adults in 24 English intensive care units, 11,018 paired oximeter and arterial co-oximetry measurements across five devices, skin tone measured by handheld spectrophotometer: readings averaged 0.6-1.5 percentage points higher at the darkest than at the lightest skin tone, and the proportion of readings above 94% despite a paired arterial saturation of 92% or less was 5.3-35.3 percentage points higher in darker skin (rate ratio 2.3-7.1), while false-positive rates fell. Prospective observational, with objective pigment measurement.

116. 116
     Siefkes H, Holla I, Giusto E, Tancredi D, Lakshminrusimha S. Neonatal Pulse Oximetry Accuracy and Disparities by Skin Pigmentation (NeoPODS): A Prospective Study.J Pediatr · 2026 · 115114doi:10.1016/j.jpeds.2026.115114 · PMID 42025790

     70 newborns with arterial catheters in two neonatal intensive care units, 40% of them Black, 136 tightly paired measurements: mean bias -0.98 ± 2.8%, that is slight underestimation of arterial saturation, with no statistically significant difference by categorical pigment measure or by race, a small continuous relationship with individual typology angle, and occult hypoxemia in one lightly pigmented infant. Prospective; a negative study, in newborns.

117. 117
     Barros KB, Pinheiro JS, Sampaio MS, et al. Accuracy and Bias of Pulse Oximetry in the Intensive Care Unit: A Prospective Observational Study.Nurs Crit Care · 2026 · 31(3):e70433doi:10.1111/nicc.70433 · PMID 41856803

     100 intensive care patients, three oximeters, pigment measured by calibrated colorimeter at two sites: two of the three devices exceeded the 2013 FDA accuracy threshold, agreement with arterial saturation was weak, overestimation grew as saturation fell, no association was found between individual typology angle and error, and older age predicted greater overestimation. Prospective; the authors note that clinical variability limits the detection of small biases.

118. 118
     Lipnick MS, Chen D, Law T, et al. Comparison of methods for characterizing skin pigment diversity in research cohorts.Br J Dermatol · 2026 · 194(1):135-145doi:10.1093/bjd/ljaf397 · PMID 41073884

     789 adults, 33,856 assessments by perceived Fitzpatrick, Monk Skin Tone and Von Luschan scales and two instruments, a Konica Minolta CM-700d spectrophotometer and a Delfin Skin Color Catch colorimeter, which correlated strongly with each other on individual typology angle and lightness (rho above 0.7): with dark defined as an individual typology angle below -30 degrees, 14% of the cohort was dark, against 26% by perceived Fitzpatrick, 16% by Monk and 23% by self-identification as Black or African American; about half the dark group was below -50 degrees, where only 7% of the cohort fell. Each self-assigned race category spanned a wide range of pigment. The authors conclude that published thresholds for dark skin are too light.

119. 119
     de Boer F, Kezic S, van der Lelie G, et al. Effect of Repeated Low-Dose UVR Exposure on Skin Inflammation Threshold, Skin Biomarkers, and Vitamin D in Healthy Adults.J Invest Dermatol · 2025 · 145(9):2304-2312.e5doi:10.1016/j.jid.2025.01.024 · PMID 39956428

     31 adults exposed to 0.8 standard erythema doses three times a week for 9 weeks: the minimal erythema dose rose 50% (more in lighter phototypes), melanin index 12%, stratum corneum thickness 34%, and 25-hydroxyvitamin D 21% after adjusting for seasonal decline, with local but not systemic immunosuppression. Small human experiment.

120. 120
     Goddard A, Watson A, Tilbury R, Corfe BM, Fairley A. Circannual prevalence of vitamin D insufficiency in older and minoritized ethnic adults in Northern Britain: screening outcomes from a clinical trial.Eur J Clin Nutr · 2026 · 80(8):855-858doi:10.1038/s41430-026-01760-z · PMID 42156552

     299 people screened in northern Britain between December 2024 and August 2025 by mass spectrometry on dried blood spots: 25-hydroxyvitamin D below 50 nmol/L in 54.8% of adults aged 65 and over and 72.1% of adults of Fitzpatrick types IV-VI, with no decline in summer. Screening data, one region, 2026.

121. 121
     Biesemeier A, Eibl O, Eswara S, et al. Elemental mapping of Neuromelanin organelles of human Substantia Nigra: correlative ultrastructural and chemical analysis by analytical transmission electron microscopy and nano-secondary ion mass spectrometry.J Neurochem · 2016 · 138(2):339-53doi:10.1111/jnc.13648 · PMID 27121280

     Human substantia nigra sections: single neuromelanin granules and lipid bodies measured about 200-600 nanometers, granules held significant sulfur, iron and copper, with iron at about 0.15 atomic percent, and lipid bodies held phosphorus consistent with phospholipids. Postmortem analytical microscopy.

122. 122
     Zucca FA, Segura-Aguilar J, Ferrari E, et al. Interactions of iron, dopamine and neuromelanin pathways in brain aging and Parkinson's disease.Prog Neurobiol · 2017 · 155:96-119doi:10.1016/j.pneurobio.2015.09.012 · PMID 26455458

     Review: the neuromelanin-iron complex is the main iron compound of dopamine and norepinephrine neurons, neuromelanin traps iron and converts excess dopamine into a stable form, but organelles overloaded with iron and toxins during aging can trigger degeneration, and neuromelanin released from dying neurons activates microglia in a self-propelling loop. Review.

123. 123
     Esteban-García N, Martín-Flórez C, Pineda-Pardo JA, et al. Blood-brain barrier opening with focused ultrasound reduces neuromelanin in a non-human primate model.Brain · 2026 · online ahead of printdoi:10.1093/brain/awag343 · PMID 42831534

     Primates overexpressing human tyrosinase accumulated neuromelanin with extracellular release, iron, inflammation and inclusions but without substantial dopamine cell loss; repeated low-intensity focused ultrasound with microbubbles reduced intracellular and extracellular neuromelanin and tended to lower free ferric iron. Animal study, 2026; a proposed route to a therapy, not one.

124. 124
     Isaias IU, Trujillo P, Summers P, et al. Neuromelanin Imaging and Dopaminergic Loss in Parkinson's Disease.Front Aging Neurosci · 2016 · 8:196doi:10.3389/fnagi.2016.00196 · PMID 27597825

     18 patients and 18 controls: substantia nigra volume and contrast on neuromelanin MRI were reduced in Parkinson's disease and correlated with striatal dopamine transporter density on SPECT, volume being the best predictor; iron-sensitive measures did not differ. Small cross-sectional study.

125. 125
     Wang X, Zhang Y, Zhu C, et al. The diagnostic value of SNpc using NM-MRI in Parkinson's disease: meta-analysis.Neurol Sci · 2019 · 40(12):2479-2489doi:10.1007/s10072-019-04014-y · PMID 31392640

     Meta-analysis: pooled sensitivity 0.82 (95% CI 0.74-0.87) and specificity 0.82 (0.73-0.89) of neuromelanin MRI of the substantia nigra pars compacta for Parkinson's disease, with volume a better marker than signal. The authors' recommendation of wide use goes beyond the data, which are case-control.

126. 126
     Nikparast F, Danesh-Doust M, Zare H. Combining neuromelanin-sensitive MRI and quantitative susceptibility mapping for enhanced diagnosis and differentiation of parkinson's disease: A systematic review.Psychiatry Res Neuroimaging · 2026 · 363:112289doi:10.1016/j.pscychresns.2026.112289 · PMID 42537381

     Seventeen studies, 2,228 participants, searched to July 2026: combining neuromelanin MRI with quantitative susceptibility mapping of iron gave areas under the curve of 0.86-0.99 and separated Parkinson's disease from controls and some mimics, with simultaneous sequences and automated segmentation shortening scans. Systematic review without pooled estimate.

127. 127
     Haller S, He N, Schwarz S, Haacke EM. Clinical MRI Biomarkers to Differentiate Parkinson Disease from Its Mimics.Radiology · 2025 · 317(1):e242044doi:10.1148/radiol.242044 · PMID 41085393

     Review of standard and emerging magnetic resonance markers against the differential diagnosis: nigrosome 1, seen as the swallow-tail sign, and neuromelanin imaging are abnormal in Parkinson's disease and in most atypical parkinsonian syndromes but normal in essential tremor, drug-induced parkinsonism and vascular parkinsonism; dopamine transporter imaging remains the reference standard and cannot reliably separate Parkinson's disease from the atypical syndromes either. Review; proposes adding the sequences to a standard scan to select patients for nuclear imaging.

128. 128
     Kim JR, Sohn B, Jo JW, et al. Neuromelanin-sensitive MRI for identifying dopamine transporter imaging abnormality risk in idiopathic rapid eye movement sleep behavior disorder.Brain Imaging Behav · 2026 · 20(4)doi:10.1007/s11682-026-01176-0 · PMID 42406004

     66 people with idiopathic rapid eye movement sleep behavior disorder and 30 healthy controls: 37 (56.1%) had abnormal dopamine transporter imaging and reduced neuromelanin measures throughout the substantia nigra, and neuromelanin volume and signal-to-noise ratio in the associative and limbic subregions separated them from those with normal scans with areas under the curve of 0.73 and 0.75. Combined retrospective and prospective study; the sequence used before diagnosis rather than after.

129. 129
     Dorr RT, Lines R, Levine N, et al. Evaluation of melanotan-II, a superpotent cyclic melanotropic peptide in a pilot phase-I clinical study.Life Sci · 1996 · 58(20):1777-84doi:10.1016/0024-3205(96)00160-9 · PMID 8637402

     Single-blind, alternating-day, placebo-controlled pilot in three healthy men at the University of Arizona, escalating in 0.005 mg/kg steps from 0.01 to 0.03 mg/kg in two subjects and 0.025 mg/kg in the third, subcutaneously: two developed pigmentation of the face, upper body and buttock by reflectance a week after dosing ended, mild nausea occurred at most doses, one of the two men at the top dose had grade II somnolence and fatigue, and a stretching-and-yawning complex accompanied erections lasting 1-5 hours. Three people; the whole of the controlled human evidence for the drug now sold illicitly.

130. 130
     Yassin Alsabbagh A, Bhujel N, Singh RP. Melanotan II nasal spray: a possible risk factor for oral mucosal malignant melanoma?Int J Oral Maxillofac Surg · 2025 · 54(9):806-808doi:10.1016/j.ijom.2025.03.014 · PMID 40210573

     A 22-year-old woman who used an unlicensed melanotan II nasal spray for tanning developed a mucosal malignant melanoma of the upper jaw, resected and treated with immunotherapy; the authors note the product's sale is illegal in the United Kingdom and many countries. One case; causation cannot be inferred.

131. 131
     Vadner DJ, Smith S. Five primary melanomas in situ in a patient with recent tanning bed use, melanotan exposure, and anabolic hormone use.JAAD Case Rep · 2026 · 73:111-114doi:10.1016/j.jdcr.2026.05.009 · PMID 42328529

     Case report of five primary melanomas in situ in a person combining tanning beds, melanotan and anabolic hormones; the record carries no abstract, so it is cited for the finding named in its title. One case with three exposures, so attribution to melanotan is impossible.

132. 132
     Bonchev A. Changes in Oral Mucosa Associated with Melanotan II Injections: A Case Report.Life (Basel) · 2026 · 16(2):265doi:10.3390/life16020265 · PMID 41752902

     After 64 days of self-injected melanotan II, symmetrical brown pigmentation of the gums and cheek lining; cheek pigment nearly cleared a month after stopping, gum pigment persisted at three months. One case; shows the drug pigments mucosa, not only skin.

133. 133
     Resnick G, Khajeh-Afzaly M, Yousefian F, Raza A, Issa NT. Insights into Tanning Biology and Tanning Products.J Clin Aesthet Dermatol · 2026 · 19(2):33-42PMID 41890775

     Systematic review of 68 studies of self-tanning agents: dihydroxyacetone pigments skin by the Maillard reaction, with unresolved concerns about cytotoxicity, genotoxicity and systemic absorption, while unregulated melanotan I and II use has caused rhabdomyolysis, renal infarction and priapism; forskolin and carotenoids are little studied in people. Review; the record carries no registered DOI.

134. 134
     Okoji UK, Taylor SC, Lipoff JB. Equity in skin typing: why it is time to replace the Fitzpatrick scale.Br J Dermatol · 2021 · 185(1):198-199doi:10.1111/bjd.19932 · PMID 33666245

     Commentary arguing that a scale built to grade sunburn in lighter skin should not be used as a proxy for skin color or race; the record carries no abstract, so it is cited for the argument in its title.

135. 135
     Weir VR, Li Y, Gillis MC, et al. Evaluating skin tone scales for dermatologic dataset labeling: a prospective-comparative study.NPJ Digit Med · 2025 · 8(1):787doi:10.1038/s41746-025-02245-2 · PMID 41429926

     Prospective comparison of Fitzpatrick, Monk Skin Tone and Pantone scales against a colorimeter, described as the gold standard for color measurement: the Monk scale showed slightly tighter clustering in color space and high repeatability, colors extracted from dermoscopic images correlated poorly with colorimetry, and Monk ratings captured differences in an artificial-intelligence melanoma classifier's scores more effectively than Fitzpatrick types. Observational; the authors state that Fitzpatrick type is not a proxy for skin tone.

136. 136
     Bernat AS, Bromberg L, Sinefeld D, Arad B. Unifying the spectrum: a framework for the meta-analysis of human skin tone scales.J Biomed Opt · 2026 · 31(9):095005doi:10.1117/1.JBO.31.9.095005 · PMID 42812670

     Overlaying the major scales on 15,256 spectral reflectance measurements from a global archive, the authors find Fitzpatrick and Monk scales fail to capture the continuous, multidimensional nature of pigmentation and that physical color targets fail under changing light. All authors work for a spectral-imaging company.