The Vitamin D Gradient

Low vitamin D is common and unevenly distributed. Higher deficiency tracks with greater asthma severity and selected autoimmune disease in darker-skinned children, yet routine testing is rare, controller therapy is used less, and maintenance spending does not rise with the burden.

Low vitamin D is common, and it is not evenly shared. In NHANES 2005–2006, deficiency was a serum 25-hydroxyvitamin D at or below 20 ng/mL. 30.9% of non-Hispanic White adults are deficient (70.3% of the U.S. adult population). 69.2% of Hispanic adults are deficient (13.9% of the U.S. adult population). 82.1% of non-Hispanic Black adults are deficient (9.9% of the U.S. adult population).

Children carry the same split, and the illnesses that follow a thin reservoir show it in percentages. In the 2021 National Health Interview Survey, 12.5% of non-Hispanic Black children had current asthma, against 5.7% of non-Hispanic White children and 5.4% of Hispanic children. Autoimmune disease is not one rate. Type 1 diabetes in youth is still a little more common in non-Hispanic White children, about 3.9 per 1,000, than in non-Hispanic Black children, about 2.6 per 1,000, in the National Health Interview Survey for 2019–2022. Childhood lupus runs the other way. Among children on Medicaid from 2000 to 2004, systemic lupus erythematosus was 4.86 per 100,000 in White children, 11.51 in Hispanic children, more than 2 times the White rate, 14.08 in African American children, nearly 3 times the White rate, and 23.79 in Asian children, nearly 5 times the White rate. NHANES through 2010 did not publish Asian Americans as their own group. They were folded into a mixed other category, and that sample was too small to stand alone. The survey did not start oversampling Asian Americans until 2011, after the counts used here. Medicaid could separate them because race was already on the claim. Under both patterns is the vitamin. Kumar and colleagues, in Pediatrics in 2009, used NHANES 2001–2004 and found 9% of Americans aged 1 to 21 deficient, a 25-hydroxyvitamin D under 15 ng/mL, and 61% insufficient. A non-Hispanic Black child was about 22 times as likely to be in that deficient group as a non-Hispanic White child. Among Black girls aged 13 to 21, 59% were deficient. Among White girls the same age, 5%. Immune cells use that storage form. When they meet a microbe they turn 25-hydroxyvitamin D into the active hormone on the spot and make cathelicidin, the peptide that kills what got inside the cell. A low level does not invent asthma or lupus by itself. It leaves the antimicrobial step the immune system already knows with less to draw on. The diseases of an empty reservoir land hardest on the children whose blood is most often empty.

The bone paradox

It is published, and it is large, and the public vitamin D message almost never says it. It was not buried. It is just not in the advertisement. African Americans have the lowest 25-hydroxyvitamin D and the lower fracture rate. The same surveys that call the deficiency highest in Black girls are the ones that find the stronger bone.

It is published, and it is large, and the public vitamin D message almost never says it.

The label and the bone

The girl figures are Kumar, already stated: 25-hydroxyvitamin D under 15 ng/mL, 5% of White girls aged 13 to 21 and 59% of Black girls the same age. Black women in the Medicare window had a hip-fracture rate of 412 per 100,000. White women had 975. The age-adjusted rates for 2008–2009 are 975.2 and 411.7 (Wright and colleagues, Journal of Bone and Mineral Research, 2012). In the Women’s Health Initiative the hip-fracture hazard for Black women was 0.29 compared with White women, confidence interval 0.25 to 0.34 (Wright and colleagues, 2024). John Aloia’s review in the American Journal of Clinical Nutrition, 2008, is the plain statement of the paradox. It reports the National Osteoporosis Risk Assessment: among 197,848 postmenopausal women, the 7,784 Black participants had about half the osteoporosis and about half the fracture risk. In that cohort, after bone density, weight, and other factors were accounted for, the fracture relative risk in Black women was 0.52 compared with White women (Barrett-Connor and colleagues, Journal of Bone and Mineral Research, 2005). A 2025 review of NHANES 2005–2008 put femoral-neck bone density about 9% to 10% higher in Black adults than in White adults.

The cutoff that labels most Black Americans deficient was built mainly on White bone data, then applied to everyone. Qiu, Divine, and Rao put the bone itself under the microscope. Bone Reports, 22 December 2022: 25 Black women and 25 White women, matched for age, menstrual status, and body mass index. Their conclusion is one sentence. Lower serum 25-hydroxyvitamin D in the Black women did not impair bone structure or remodeling, and did not decrease bone mineralization. They also found higher 1,25-dihydroxyvitamin D in the Black women, which may help keep bone and, in the same biopsies, moderately slow the last step of mineralization.

Two limits, so it does not get overread. The osteoporosis figures in the table are age-adjusted, femur neck or lumbar spine, women 50 and older, NHANES 2013–2014 (Looker and colleagues, Osteoporosis International, 2017). Hispanic women in that window were 20.5%. Non-Hispanic Asian women were 40.0%, the highest, so low vitamin D is not a shield. And a lower fracture rate is not zero. Hip fracture is still a real injury in Black women. The point is the gap between the lab label and the bone.

A pill cannot do the rest of what sunlight does

Sunlight on skin and a D3 pill make the same molecule. The test measures that shared molecule. Vitamin D synthesis in skin starts only when the UV index is above 3. The band that does the work is UVB, about 290–315 nm. It opens 7-dehydrocholesterol in the epidermis into previtamin D3, which body heat turns into vitamin D3. Below a UV index of 3, that band is too weak at the ground.

A D3 pill is that same molecule, cholecalciferol, usually made from lanolin. A D2 pill is not. Ergocalciferol comes from fungi, and the sun never makes D2. After either kind of D3, sun or capsule, the liver builds 25-hydroxyvitamin D, the storage form, which lasts about 2 to 3 weeks. The kidney, and some immune cells, turn a little of that into 1,25-dihydroxyvitamin D, the active hormone, which lasts hours. The chemical identity stops there. The delivery does not.

Skin lets D3 out slowly, over days. If the ultraviolet keeps arriving, surplus previtamin D3 and vitamin D3 are broken into inactive lumisterol and tachysterol. Sunlight does not fill the store without a limit, and sunlight alone does not produce vitamin D toxicity. A pill is a bolus through the gut. Nothing in the capsule shuts the dose off. Enough of it can push blood calcium too high.

The rest of daylight is not in the bottle. UVA does not make vitamin D. It releases nitric oxide already stored in the upper skin. In 24 healthy volunteers, 2 standard erythemal doses of UVA widened arteries and lowered blood pressure, and the vitamin D level did not have to change for that to happen (Liu, Weller, and colleagues, Journal of Investigative Dermatology, 2014). Light through the eye, especially blue light near 480 nm, sets the clock: melatonin, the day’s cortisol, sleep. That path does not use vitamin D. Skin that is actually being lit also makes some of the active hormone on the spot and spends it there, including on cathelicidin, the peptide named above. Ultraviolet also raises beta-endorphin in skin. In mice, that signal was enough that the animals sought the light (Fell and colleagues, Cell, 2014). That result is in mice, not a measured human addiction. A mood that follows outdoor light is still not in the capsule, and neither is the clock. A pill can raise the blood store. It cannot shine.

The ordinary blood test reads total 25-hydroxyvitamin D. That is the storage form. It is not the active hormone, not the vitamin D still sitting in the skin, and not the nitric oxide or the clock. Ordering 1,25-dihydroxyvitamin D to ask whether someone got enough sun is the wrong test. That hormone is often normal when the store is low. 20 ng/mL is 50 nmol/L. 30 ng/mL is 75 nmol/L. The number is the last several weeks of sun, food, and pills added together, not this morning.

Sun D3 and a D3 pill both become 25-hydroxyvitamin D3. A careful laboratory cannot tell those two apart. Only a D2 pill shows up as its own chemical, 25-hydroxyvitamin D2. Liquid chromatography with tandem mass spectrometry lists D2 and D3 on separate lines. Many antibody tests print 1 total, and they are weak at seeing D2. A draw the morning after one short exposure may barely move. The liver has not finished, and the store turns over in weeks, not hours. After a large pill, vitamin D3 itself rises within hours, and 25-hydroxyvitamin D rises over about 1 to 2 weeks. The report still names the same D3 storage form the sun would have made. The test also does not read the other half of sunlight. Blood pressure movement from UVA, and a clock set by the eye, can change while 25-hydroxyvitamin D stays the same. A normal store does not prove that daylight happened. A low store does not prove that the only missing thing is a capsule. Darker skin needs more of the same UVB to build the same store. The test still sees only the store.

What the vitamin D test can and cannot name

Melanoma’s recorded rise and the marketing of sunscreen are not the same event. From the 1930s to the 1960s, when a tan was the product and Coppertone was still a beach lotion, melanoma incidence in the United States rose by more than 300% in men and more than 400% in women, far faster than cancer overall. Coppertone began as Benjamin Green’s wartime petrolatum and was sold after the war as a way to tan, not as a cancer drug. Franz Greiter’s SPF number dates to 1962. The Food and Drug Administration’s sunscreen rules, including a label that the product could prevent skin cancer, date to 1978. The cancer-prevention campaign is the 1980s: broad-spectrum bottles, Australia’s Slip, Slop, Slap in 1981, and a public line that there is no safe tan. In the SEER registry the rate of new melanoma cases was 8.8 per 100,000 in 1975 and 28.3 in 2019. The death rate did not follow. It was 2.06 per 100,000 in 1975 and 2.06 again in 2019. More tumors were found. People were not dying of them at 3 times the rate. The early rise tracks fashion, travel, and more bare skin. A large share of the later rise is more looking, and thinner tumors. The Nambour trial in Queensland still found that daily sunscreen reduced squamous-cell cancers and, on later follow-up, melanoma. That result stands. It does not turn the sales chart into a medical chart.

The Nambour numbers, stated once: 1,621 adults, daily versus discretionary sunscreen on the head, neck, arms, and hands. Squamous-cell tumor rates were lower in the daily arm, rate ratio 0.61 (Green, Lancet 1999). Basal-cell carcinoma was not reduced. On follow-up to 2006 the same cohort had 11 new melanomas in the daily arm and 22 in the discretionary arm, hazard ratio 0.50, and the confidence interval still touched 1. Invasive melanomas were 3 versus 11 (Green, Journal of Clinical Oncology 2011). The authors wrote that melanoma may be preventable. That “may” belongs beside the result.

What did rise, year after year, was the money. Global sunscreen sales were about 13.2 billion dollars in 2024, with published forecasts climbing toward about 22 billion by 2033, on the order of 6% a year. A worldwide profit figure for sunscreen in 1996 is not published. The companies do not break sunscreen profit out as its own line. What is published is sales. Euromonitor put global sun care sales at 5.6 billion dollars in 2005. Datamonitor’s 2013 pieces add to about 8.7 billion: 5.6 billion in sun protection, 1.7 billion after-sun, and 1.4 billion self-tanning. Euromonitor had the category near 10 billion in 2016. Research and Markets put sun care products at 11.9 billion in 2022. From 5.6 billion in 2005 to 13.2 billion in 2024 is 7.6 billion dollars more, about 2.4 times the 2005 sales, across 19 years. The one filed profit line that can be named is Edgewell’s Sun and Skin Care segment, which also includes skin care: 108.8 million dollars of segment profit in 2022, 137.4 million in 2023, and 131.3 million in 2024.

Sales, not profit. Billion dollars. About 20 years.

5.6

2005

8.7

2013

10

2016

11.9

2022

13.2

2024

2005 and 2016, Euromonitor sun care. 2013, Datamonitor sun care pieces added. 2022, Research and Markets sun care products. 2024, sunscreen sales. Houses do not use the same basket.

Melanoma, over the same years, did not follow the sales chart, and it did not follow the vitamin D order. These are new cases of melanoma per 100,000 people, age-adjusted, both sexes, from the SEER Cancer Statistics Review, 1975–2016. Basal-cell and squamous-cell skin cancers are not in this registry. White and Black rates are SEER 9. Hispanic, Asian/Pacific Islander, and American Indian/Alaska Native rates are SEER 21, so the baskets are not identical.

Melanoma incidence per 100,000

White incidence rose from 27.9 to 31.6. Black incidence fell from 1.1 to 0.8. In 2016 the White rate was about 40 times the Black rate. Hispanic stayed near 4.7. Asian/Pacific Islander stayed near 1.3. The next SEER window, 2015–2019, uses a non-Hispanic cut and splits by sex, so it is not added to the table. Non-Hispanic White incidence was 38.6 in men and 25.5 in women. Hispanic was 4.6 and 4.5. Non-Hispanic American Indian/Alaska Native was 8.7 and 8.3. Non-Hispanic Asian/Pacific Islander was 1.4 and 1.2. Non-Hispanic Black was 1.0 and 0.9. New cases rose in the group with the highest vitamin D. They did not rise in the group with the lowest.

Melanoma, new cases per 100,000

A dash means that source has no earlier year. Hispanic, Asian/Pacific Islander, and American Indian/Alaska Native are not in the 1975 file, so 2005 is their first year here. White and Black are SEER 9. Those three are SEER 21. Connecticut and England are not split by race. England uses the European standard. Australia uses its own. The birthplace rows use the world standard. These rates cannot be stacked as if they were one yardstick.

A 2025 meta-analysis of 5 studies (1,814 adults) found lower odds of bloodstream infection and shorter hospital stays with vitamin D sufficiency.

The advertisement that carries that growth is that daylight on skin is the hazard. The Washington voice of the trade is the Personal Care Products Council. OpenSecrets, using Senate lobbying filings, records 840,000 dollars of federal lobbying by that council in 2015, and 270,000 dollars in the first half of 2026 alone, spent on which filters the FDA may sell. The same filings, read quarter by quarter so an amendment is not counted twice, put the Council’s own reported expenses at 1,020,000 dollars in 2023, 430,000 in 2024, and 480,000 in 2025. The 2015 text that was read does not mention sunscreen. From 2023 it does. The Public Access to Sunscreens Coalition, through Holland & Knight, reported 360,000 dollars in 2015 and 280,000 in each of 2024 and 2025. Those totals cover the issues named on the form. They are not a sunscreen-only budget, and the payments to outside firms are already inside the Council’s own total.

Member stock trades are not named here. Johnson & Johnson is a widely held large company. Disclosures are ranges, often a spouse or a child, and no record reviewed here ties a trade to a sunscreen vote. One revolving door is documented, and it is the cosmetics office, not the sunscreen review desk. John Bailey directed the FDA Office of Cosmetics and Colors, left in 2002, and became the Council’s scientist. A document showing that a company or the agency buried a benzene result, or barred a paper, was not found. Disclosed lobbying is advocacy. It is not, by itself, suppression.

The damage that has been measured is in the bottles. In May 2021 the laboratory Valisure told the FDA it had found benzene, a carcinogen, in a large share of aerosol sunscreens. Of 294 batches from 69 companies, benzene was detectable in 27%. 14 lots, from 4 brands, measured 2.78 to 6.26 parts per million. 217 batches had none. Johnson & Johnson recalled Neutrogena and Aveeno sprays that July after its own tests agreed. In January 2023 a federal judge in Brooklyn gave final approval to a Coppertone settlement: Beiersdorf and Bayer to pay 2.3 million dollars and to test for benzene. The SPF number was always a burn test. It measures UVB. UVA, which also causes skin cancer, was not in that number, and the people who printed the labels knew the difference. In June 2011 the FDA said more than a third of the SPF products then on the shelf might fail a real test for both rays, and that a lotion had been allowed to say “broad spectrum” if it blocked any UVA at all. A bottle that does not say broad spectrum SPF 15 or higher, the agency said, has not been shown to prevent skin cancer. It stops the burn. The ads kept blaming the sun. The final rule of 17 June 2011 is the one still on the books: only a broad-spectrum product of SPF 15 or higher may claim it decreases the risk of skin cancer and early aging, and only if it is used as directed with shade, clothing, a hat, and sunglasses.

The first national vitamin D count that listed Asian Americans alone is NHANES 2011–2014, ages 1 and older (Herrick and colleagues). It uses a stricter low line than the 2005–2006 adult figures above, so the percents cannot be laid on top of 30.9, 69.2, and 82.1. Under 12 ng/mL the shares were 2.1% White, 5.9% Hispanic, 7.6% Asian, and 17.5% Black. From 12 to just under 20 ng/mL they were 11.8%, 26.3%, 29.1%, and 35.8%. Black Americans are the lowest. Asian Americans sit near Hispanic Americans, not at the bottom. The Medicaid lupus order does not follow that map. The group with the lowest vitamin D does not have the highest childhood lupus rate. The group with the highest lupus rate does not have the lowest vitamin D.

A higher store is linked with quieter lupus, and the size of that help is limited. Immune cells carry the vitamin D receptor. The active hormone turns down the type I interferon signal, favors regulatory T cells, and slows the B cells that make autoantibodies. In 32 women with lupus, interferon activity was 3.5 where the store was under 20 ng/mL and 0.3 where it was not (Ritterhouse, 2011). That is a small, cross-sectional study. In a 2025 review of supplement trials, activity scores on the SLEDAI scale sat about 1 point lower than with placebo. Of 21 studies, 12 reported a benefit and 9 did not. A 1-point easing is a help. It is not remission, and it does not replace lupus treatment. People who already have lupus are often low because they are told to stay out of the sun, because steroids change the handling of the vitamin, and because kidney disease can waste the binding protein. The arrow runs both ways.

Trials find that restoring a low level, daily or weekly, lowers the chance of an acute respiratory infection. Immune cells use that same 25-hydroxyvitamin D. When they detect bacteria they activate it on site and make cathelicidin, an antibiotic peptide. Martineau’s individual-patient meta-analysis, 25 trials and 11,321 people, put the overall adjusted odds at 0.88. Among daily or weekly doses, not rare large boluses, the odds were 0.30 when the starting level was under 10 ng/mL and 0.75 when it was already higher. The people who started lowest gained the most. In older adults who did not already have autoimmune disease, 2,000 IU a day in the VITAL trial was followed by fewer new autoimmune diagnoses, hazard ratio 0.78, about a 22% reduction. After the pills stopped, 2 more years of watching erased that vitamin D difference.

The skin’s own limit is already stated above. Food will not close a deep deficit by itself. From the National Institutes of Health table: cod liver oil, 1 tablespoon, 1,360 IU; farmed rainbow trout, 3 ounces, 645 IU; sockeye salmon, 3 ounces, 570 IU; ultraviolet-exposed white mushrooms, ½ cup, 366 IU; fortified 2% milk, 1 cup, 120 IU; 1 egg, 44 IU, almost all of it in the yolk. The adult recommended intake is 600 IU a day through age 70 and 800 IU after that. The tolerable upper level for ages 9 and older, for the general population and not as a treatment course, is 4,000 IU a day.

These patterns of deficiency and disease activity sit against measured differences in care intensity and spending. Non-Hispanic Black children experience current asthma prevalence roughly 1.6 times the overall rate and substantially higher rates of emergency-department visits and hospitalizations. Medicaid analyses have shown adjusted odds of hospitalization and emergency visits approximately 2.4-fold higher for Black children with asthma. Childhood-onset SLE shows a similar skew in the Medicaid data cited above. Institute for Health Metrics and Evaluation estimates for ages 0–19 years placed per-person health care spending lower for non-Hispanic Black children (approximately $2,791) than for non-Hispanic White children (approximately $3,423), with less ambulatory utilization but more inpatient and emergency spending. Among children with treated asthma in 2016–2021 Medical Expenditure Panel Survey data, Black children generated roughly $2,721 less in total annual medical expenditures and $803 less in office-based visits than White children with treated asthma. Controller inhaler use remains lower among Black, Hispanic, and Asian adults with asthma than among White adults. Combination controllers carry high average wholesale prices. The result is relatively more rescue and acute care, with Medicaid absorbing a larger share of the elevated acute-care volume while spending on continuous controllers stays comparatively restrained. Historical Agency for Healthcare Research and Quality figures placed the average cost of a pediatric asthma hospitalization near $3,600.

Routine 25-hydroxyvitamin D testing is uncommon and is discouraged by major guidelines in the absence of specific indications. The American Academy of Pediatrics Choosing Wisely recommendation advises against ordering vitamin D concentrations routinely in otherwise healthy children, including those who are overweight or obese. The Endocrine Society’s 2024 guideline suggests empiric supplementation for children and adolescents ages 1–18 to prevent rickets and potentially lower respiratory-infection risk, but it advises against routine testing in the general population, including those with dark complexion or obesity. Testing is more often reserved for disorders associated with low bone mass, rickets, or certain chronic conditions. In practice the large majority of children across racial groups do not receive annual assays. Supplement use is low. Recent National Health and Nutrition Examination Survey data indicate that only about 3 percent of youth reported vitamin D supplement use in the past 30 days. Mean dietary intakes have been lower among non-Hispanic Black children. Clinic studies of urban African American youth with asthma have documented high rates of deficiency or insufficiency without evidence of systematic testing or intensified repletion as standard care. The prevailing approach has favored low-dose empiric supplementation recommendations (400 IU for infants and 600 IU for older children) over measurement and targeted repletion. For most children this is inexpensive and safe. For subgroups with darker skin, limited sun exposure, or established asthma or lupus, the same empiric dose frequently fails to reach sufficiency, and the absence of routine testing means the shortfall is not systematically identified. Some cost-utility analyses have found vitamin D add-on therapy cost-effective for mild-to-moderate asthma. Absolute supplement costs remain far below those of controller inhalers. Uncontrolled asthma is projected to generate roughly $300 billion in direct medical costs and nearly $964 billion in total economic burden, including productivity losses, over 20 years in the United States. Global modeling attributes hundreds of billions in cumulative output losses through 2050.

Higher deficiency prevalence, higher asthma and selected autoimmune morbidity, lower controller utilization, and low rates of testing produce a coherent gap. Guidelines correctly note that universal screening of healthy children lacks strong trial support. They do not show that high-risk groups—darker-skinned children with asthma living at higher latitudes—are currently receiving either adequate repletion or proportional intensity of maintenance therapy. Targeted testing in these groups is low-risk and low-cost relative to documented acute-care and productivity losses. Raising documented low levels into a sufficient range can be achieved with safe daily or weekly supplementation at doses higher than the standard recommended dietary allowance when needed, alongside continued controller therapy.

Improving vitamin D status will not erase disparities driven by housing, air quality, access to specialty care, or socioeconomic position. It does address a measurable, modifiable contributor to immune dysregulation that is disproportionately present in the populations already carrying higher burdens of exacerbations and selected autoimmune disease. For clinicians the practical step is to measure 25-hydroxyvitamin D in high-risk patients, replete documented deficiency, and monitor response with standard therapy. For payers the arithmetic favors identifying and correcting deficiency early so that larger downstream costs of uncontrolled disease can be reduced. For families the benefit is fewer exacerbations and a clearer path to more stable immune regulation. Higher disease incidence linked to lower vitamin D status has not been matched by higher maintenance spending or by routine testing and targeted repletion. Emphasizing measurement and correction in the groups at greatest biologic risk is a direct response to that mismatch.

Selected Sources and References

Forrest KYZ, Stuhldreher WL. Nutr Res. 2011;31:48–54. NHANES 2005–2006. 30.9% of non-Hispanic White adults (70.3% of the U.S. adult population), 69.2% of Hispanic adults (13.9%), 82.1% of non-Hispanic Black adults (9.9%), at or below 20 ng/mL. PubMed 21310306

Liu D, Fernandez BO, Hamilton A, et al., including Weller RB. J Invest Dermatol. 2014;134:1839–1846. UVA, 24 volunteers, blood pressure down without a change in vitamin D. PubMed 24445737

Looker AC et al. Osteoporos Int. 2017. NHANES 2013–2014, adults 50 and older, osteoporosis at the femur neck or lumbar spine, age-adjusted: non-Hispanic White women 17.0%, non-Hispanic Black women 8.2%, Hispanic women 20.5%, non-Hispanic Asian women 40.0%. PubMed 28315954

Wright NC et al. J Bone Miner Res. 2012. Medicare, women 65 and older, 2008–2009 hip fracture: 975 per 100,000 White, 412 per 100,000 Black. JBMR 2012

Wright NC et al. J Bone Miner Res. 2024;39:1296–1305. Women’s Health Initiative. Hip-fracture hazard for Black women versus White women, 0.29 (0.25 to 0.34), adjusted for age and trial arm. JBMR 2024

Aloia JF. Am J Clin Nutr. 2008;88:545S–550S. Reports the National Osteoporosis Risk Assessment: 197,848 postmenopausal women. The 7,784 Black participants had about half the osteoporosis and about half the fracture risk. PubMed 18689399

Barrett-Connor E, Siris ES, Wehren LE, et al. J Bone Miner Res. 2005;20:185–194. Same cohort. After bone density, weight, and other factors, fracture relative risk in Black women was 0.52 (0.38 to 0.70) compared with White women. PubMed 15647811

Qiu S, Divine G, Rao SD. Bone Rep. 2022 Dec 22;18:101650. 25 Black women and 25 White women, matched for age, menstrual status, and body mass index. Lower 25-hydroxyvitamin D did not impair bone structure or remodeling, and did not decrease mineralization. Higher 1,25-dihydroxyvitamin D may help preserve bone and may moderately slow the last step of mineralization. DOI 10.1016/j.bonr.2022.101650

NHANES 2005–2008. Femoral-neck bone density about 9% to 10% higher in Black adults than in White adults, age- and sex-adjusted, as summarized in J Clin Endocrinol Metab. 2025;110:3309. JCEM 2025

Fell GL, Robinson KC, Mao J, Woolf CJ, Fisher DE. Cell. 2014;157:1527–1534. In mice, ultraviolet raised beta-endorphin, and the animals sought the light. DOI 10.1016/j.cell.2014.04.032

Kumar J, et al. Prevalence and associations of 25-hydroxyvitamin D deficiency in US children: NHANES 2001–2004. Pediatrics. 2009.

Herrick KA et al. Am J Clin Nutr. 2019;110:150–157. First national Asian American vitamin D figures, NHANES 2011–2014. PubMed 31076739

Paulose-Ram R et al. Am J Public Health. 2017;107:916–921. Asian American oversampling began in 2011.

Hiraki LT et al. Arthritis Rheum. 2012;64:2669–2676. Medicaid children, 2000–2004. PubMed 22847366

National Health Interview Survey (NHIS) 2021 asthma prevalence and 2019–2022 type 1 diabetes estimates.

Institute for Health Metrics and Evaluation (IHME) health care spending estimates for ages 0–19 by race.

Medical Expenditure Panel Survey (MEPS) 2016–2021 treated asthma expenditure data.

American Academy of Pediatrics Choosing Wisely recommendation on routine vitamin D testing in healthy children.

Endocrine Society Clinical Practice Guideline: Vitamin D for the Prevention of Disease (2024).

Agency for Healthcare Research and Quality (AHRQ) pediatric asthma hospitalization cost data.

UK Biobank and meta-analyses of vitamin D levels in children with asthma (Asia/Africa cohorts).

Global Burden of Disease and economic modeling projections for uncontrolled asthma costs (US and global, through 2050).

Martineau AR et al. BMJ. 2017;356:i6583. Odds 0.30 below 10 ng/mL, 0.75 above that, on daily or weekly dosing. BMJ

El Kababi S et al. Nutrients. 2025;17:2794. SLEDAI about 1 point lower. PubMed 40944182

Ritterhouse LL et al. Ann Rheum Dis. 2011;70:1569–1574. Interferon activity and vitamin D in lupus.

Hahn J et al. BMJ. 2022;376:e066452. VITAL, 2,000 IU a day. Costenbader KH et al. Arthritis Rheumatol. 2024;76:973–983. The difference did not hold after the pills stopped.

NIH Office of Dietary Supplements. Vitamin D fact sheet, food table. ods.od.nih.gov

Valisure laboratory findings on benzene in aerosol sunscreens (2021); Johnson & Johnson recall, 14 July 2021; Coppertone settlement, 10 January 2023, 2.3 million dollars.

SEER cancer registry melanoma incidence and mortality trends; Nambour trial (Green, Lancet 1999; Green, J Clin Oncol 2011).

SEER Cancer Statistics Review, 1975–2016, Figure 16.2. Melanoma incidence per 100,000, observed, age-adjusted to the 2000 U.S. standard population. White and Black, SEER 9. Hispanic, Asian/Pacific Islander, and American Indian/Alaska Native, SEER 21. SEER CSR

Henrikson NB et al. USPSTF evidence review, 2023. SEER melanoma incidence and mortality, 2015–2019, by race and sex. Table 2

Geller AC et al. J Clin Oncol. 2013;31:4172–4178. Connecticut Tumor Registry, 1950–2007. Men 1.9 to 33.5 per 100,000. Women 2.6 to 25.3. PubMed 24043747

Memon A et al. Lancet Reg Health Eur. 2021. England melanoma, 1981–2018. Men 4.5 to 28.3 per 100,000. Women 7.3 to 25.3. European standard population. PMC8454583

Australian Institute of Health and Welfare. Cancer data in Australia. Melanoma age-standardised rate 30 per 100,000 in 1982 to an estimated 63 in 2025. Ages 0–39: 10.2, then 12.5 in 2005, then an estimated 7.2. AIHW

Euromonitor International. Global sun care sales 5.6 billion dollars in 2005; near 10 billion in 2016. Datamonitor, reported 2016: 2013 sun protection 5.6 billion, after-sun 1.7 billion, self-tanning 1.4 billion. Research and Markets, 14 March 2023: sun care products 11.9 billion in 2022. ResearchIntelo: sunscreen 13.2 billion in 2024.

Edgewell Personal Care, fiscal 2024 Form 10-K. Sun and Skin Care segment profit 108.8 million dollars in 2022, 137.4 million in 2023, 131.3 million in 2024. Sun care product sales inside that segment were 401.8, 450.7, and 483.6 million.

Senate Lobbying Disclosure Act filings and OpenSecrets. Personal Care Products Council, 2015 through the first half of 2026. PASS Coalition. lda.gov

FDA. 21 CFR 201.327. Sunscreen labeling rule, 17 June 2011.

Additional observational and meta-analytic evidence linking vitamin D status to respiratory infection risk, asthma exacerbations, and immune function.