Vitamin D — The Supplement That Isn't Sunlight

• Norway study (University of Tromsø): Long-term vitamin D intake only slightly above the recommended 400 IU per day was associated with increased myocardial infarction risk. Degenerate joint disease and arthritis were also cited as conditions promoted by increased vitamin D intake. This finding was published before the mass-supplementation era and was largely ignored as supplementation guidelines expanded.
• Arterial calcification in animal models: Rats fed 250 IU per day of supplemental vitamin D developed hardened arteries. The calcium mobilization mechanism by which vitamin D operates — moving calcium from the gut into circulation — requires adequate downstream processing. When that processing is impaired (magnesium deficiency, vitamin K2 insufficiency, insulin resistance), calcium deposits in arterial walls rather than in bone.
• The VITAL trial and cardiovascular outcomes: The large VITAL randomized trial (2019) found that 2,000 IU per day of vitamin D3 supplementation did not reduce cardiovascular events. It did not harm, but it also did not produce the cardiovascular benefits that had been hypothesized from observational studies — suggesting that what observational data was capturing was not the supplement effect but the sun exposure effect.

• K2 does not address the root cause. Isolated K2 supplementation does not lower serum 25-OH-D. It does not slow adipose tissue release. It does not reduce calcium absorption driven by excess vitamin D. It attempts to manage the downstream consequence — calcium in circulation — without addressing what put it there. The tissue burden of stored vitamin D continues accumulating on every subsequent supplement dose regardless of K2 intake.
• K2 has its own interaction profile. MK-7 (the long-chain menaquinone form most commonly sold) has a half-life of approximately 70 hours — it accumulates with daily dosing (Schurgers et al., Blood 2007). At doses as low as 10 µg/day — far below the 100–200 µg in typical D+K2 supplements — MK-7 significantly altered INR in approximately 50% of subjects on anticoagulant therapy; at 45 µg/day it decreased INR by an average of 37% (PMID 23530987). Patients on warfarin or any vitamin K antagonist are not uniformly safe to supplement with MK-7. The assumption that K2 is benign because "it's a vitamin" does not hold at the doses used in D+K2 protocols.
• The evidence base is thin and uses surrogate endpoints. The most-cited trial (Knapen et al., Thrombosis and Haemostasis 2015) found that 180 µg/day MK-7 reduced carotid-femoral pulse wave velocity and a vascular calcification marker (dp-ucMGP) in healthy postmenopausal women — but the significant local carotid effects appeared only in the subgroup with the highest baseline arterial stiffness, the study used a commercial supplement (MenaQ7, NattoPharma) with authors disclosing industry ties, and a 2023 systematic review flagged unclear allocation concealment. No randomized trial has demonstrated that K2 supplementation reduces hard cardiovascular endpoints — myocardial infarction, stroke, or cardiovascular mortality. The entire evidence base rests on surrogate biomarkers.
• It's a workaround, not a solution. The logic of "take D + K2 + magnesium to manage D toxicity" is the logic of adding supplements to manage the side effects of a supplement. The better answer is to source vitamin D as the body was designed to source it — from sun on skin — and let the body manage calcium disposition through the feedback systems that evolved for exactly that purpose.

The primary clinical justification for vitamin D supplementation is bone health — prevention of osteoporosis and fractures. A 2019 JAMA study found that high-dose vitamin D supplementation (4,000 IU per day) did not improve bone density and in some measures worsened it. This is the opposite of the stated therapeutic goal. The mechanism: at high concentrations, vitamin D activates osteoclast activity — bone resorption — rather than supporting bone building. The result can be net bone loss rather than gain. The supplement prescribed for bone health, at doses commonly recommended, may be accelerating exactly the process it is meant to prevent.

• Kidney calcification in infants: Documented in offspring of women with high vitamin D intake during pregnancy.
• Severe mental retardation: Linked in the literature Kime reviewed to excessive prenatal vitamin D.
• Supravalvular aortic stenosis: Congenital narrowing of the aorta above the valve — associated with excess prenatal vitamin D.
• Facial bone abnormalities: Described as "elfin faces" — characteristic facial structure changes. The animal model was striking: 70% of offspring of rabbits given large amounts of vitamin D during pregnancy had abnormalities of the facial bones.
• Hypercalcemia in infants: Elevated calcium in neonates from maternal supplementation — associated with multiple developmental problems.

Outcomes that correlate with formula feeding and excess early vitamin D

• Type 1 diabetes: Formula-fed infants have consistently higher rates of type 1 diabetes than breastfed infants. The mechanism has been attributed to cow's milk protein and early gut antigen exposure — but the vitamin D loading is a parallel variable that has received far less attention. Vitamin D receptors (VDR) are expressed on pancreatic beta cells and on regulatory T cells. The VDR-mediated immune calibration that occurs in early infancy may be disrupted by non-physiological vitamin D loading. Both deficiency and excess have been associated with autoimmune risk — the dose matters, and formula provides a dose the infant's pancreatic immune environment was not designed to receive continuously from birth.
• Obesity: Formula-fed infants have higher rates of childhood and adult obesity than breastfed infants. Vitamin D receptors are expressed in adipose tissue and regulate adipocyte differentiation. Early excess vitamin D may affect fat cell programming during the window when adipose set points are established. This is not the only mechanism — caloric density, feeding self-regulation, and gut microbiome effects of formula all contribute — but the vitamin D dose is an uncontrolled variable in every formula-fed infant.
• The wrong fraction entirely: Breastmilk delivers the sulfated form of vitamin D (25-OH-D3-sulfate) at approximately a 3:1 ratio to the fat-soluble form. Formula delivers only the fat-soluble form — the same fraction that accumulates, that the standard test measures, and that at excess drives calcium dysregulation. No infant formula currently provides the sulfated fraction. The infant receiving formula is getting more of the problematic fraction and none of the fraction their biology expects.

• In kidney disease, the standard protocol is wrong from the start. Physicians routinely prescribe activated vitamin D (calcitriol) or high-dose cholecalciferol to patients with chronic kidney disease on the rationale that impaired kidneys cannot convert 25-OH-D to active calcitriol. This is partially true. What it ignores: 25-OH-D itself drives calcium absorption independently of renal activation, and nephrocalcinosis is a known consequence of vitamin D toxicity. Supplementing vitamin D into a kidney that already cannot process it is adding calcium load to an organ already failing under calcium burden.
• Fortified foods matter. After stopping supplements, continued intake through fortified milk, cereals, and infant formula contributes to the ongoing load. Full resolution requires eliminating fortified sources, not only pills.

Why the supplement fails to correct what it's measuring

• The standard 25-hydroxyvitamin D test measures only the fat-soluble fraction of circulating vitamin D — not the sulfated fraction that sunlight produces and that the body preferentially uses (see The Wrong Fraction tab)
• A supplement raises the fat-soluble fraction measured by the test without addressing the underlying photonic deficit
• The non-vitamin-D benefits of sunlight — nitric oxide release, immune activation, circadian regulation, hormone production — are not addressed by any supplement and are not captured by any blood test
• Deficiency that persists despite supplementation at 400 IU suggests the test threshold may be inadequate, or the supplement is not being absorbed and processed efficiently — often due to magnesium deficiency, gut dysbiosis, or gallbladder insufficiency impairing fat-soluble vitamin absorption

Why LED and fluorescent light disrupt what sunlight coordinates

• Blue-dominant spectrum without infrared balance. Natural sunlight delivers the full spectrum — UV, visible light with blue balanced by red, orange, and yellow, and near-infrared from sunrise through sunset. LED and fluorescent lights spike in blue wavelengths with absent near-infrared. Infrared is not decorative: it is the wavelength range that drives mitochondrial cytochrome c oxidase (Complex IV), enables retinal regeneration after light exposure, and provides the repair signal that allows the photobiological system to reset. Driving the blue-sensitive melanopsin system without the infrared recovery signal is like running an engine with no coolant.
• Flicker below conscious perception. Fluorescent lights driven by 60 Hz AC power flicker at 120 Hz — twice the power frequency. LED lights with pulse-width modulation drivers flicker at frequencies ranging from 200 Hz to several kHz. Flicker at these rates is invisible to conscious vision but is detected by the retinal and skin melanopsin system. Sustained high-frequency flicker creates neurological stress in the visual processing pathway and disrupts the steady photonic input that melanopsin requires for stable circadian signaling.
• Vitamin A depletion at the chromophore level. Every opsin-based photopigment in the human body uses retinal (the aldehyde form of vitamin A) as its chromophore — not just the well-known visual opsins. This includes rhodopsin and cone opsins in the retina, melanopsin (OPN4) in retinal ganglion cells and skin, neuropsin (OPN5, expressed in brain and skin, UV-sensitive), encephalopsin (OPN3, expressed in liver, brain, and skin), and peropsin in the retinal pigment epithelium. The distribution of opsins throughout the body means that the retinol demand for photopigment cycling extends far beyond vision. The light-sensing cycle bleaches retinal to activate the receptor; retinal must then be regenerated from retinol to reset the pigment. Chronic blue-dominant artificial light drives this bleach-and-regenerate cycle across multiple tissue systems simultaneously, depleting the retinol pool at a rate that outpaces replenishment. This is one mechanism by which chronic artificial light exposure degrades vitamin A status — not through the gut, but through the opsin system in eye, skin, liver, and brain.
• Vitamin A and D are co-dependent at the nuclear receptor level. The vitamin D receptor (VDR) does not act alone. To bind DNA and regulate gene transcription, VDR must pair with the retinoid X receptor (RXR) to form a heterodimer. RXR is vitamin A's nuclear receptor. A depleted retinol pool impairs RXR availability — which means vitamin D cannot complete its receptor signaling even when serum 25-OH-D appears adequate. Low vitamin A from chronic artificial light exposure may be one reason supplemental vitamin D consistently fails to produce the clinical outcomes the blood number predicts.

How seed oils compromise every stage of vitamin D production

• The skin burns before it can produce. Omega-6 linoleic acid from seed oils incorporates into cell membranes throughout the body — including skin. The adipose half-life of linoleic acid is approximately six years, meaning current skin membrane composition reflects years of dietary fat history. When UV light hits skin loaded with PUFA-rich membranes, it initiates lipid peroxidation directly in the skin cells — generating 4-HNE and malondialdehyde at the site. This is the chemical basis of sunburn susceptibility in seed oil-dominant diets: the membranes oxidize under UV before adequate vitamin D synthesis can occur. The response — burning, inflammation, avoidance, sunscreen — eliminates the UV exposure the vitamin D pathway requires.
• The sulfation pathway is damaged at the enzyme level. The sulfotransferase enzymes (SULT2B1) that convert vitamin D to its sulfated form are membrane-associated proteins sitting in the same lipid environment that PUFA peroxidation products damage. 4-HNE and malondialdehyde generated in skin during UV exposure are directly cytotoxic to these enzymes. A person with a PUFA-loaded skin membrane may produce some cholecalciferol under UV, but the sulfation step — the pathway that produces the water-soluble, non-accumulating, biologically active fraction — is running on damaged machinery. The fat-soluble fraction accumulates; the sulfated fraction is not produced at the rate the body needs.
• The liver cannot complete the first hydroxylation cleanly. 4-HNE from seed oil oxidation accumulates in the liver and is directly hepatotoxic. Non-alcoholic fatty liver disease — extremely common in Western populations eating seed oil-heavy diets — impairs hepatic function broadly, including the CYP2R1 enzyme that performs the first hydroxylation step: converting cholecalciferol to 25-OH-D. A compromised liver produces less 25-OH-D from the same cholecalciferol input. The low storage test number may reflect not lack of sun but impaired hepatic processing of what the sun did produce.
• The kidney cannot complete the second hydroxylation cleanly. Chronic seed oil consumption drives systemic inflammation and oxidative stress that affects renal tubular function over time. The kidney's 1-alpha-hydroxylase enzyme converts 25-OH-D to 1,25-OH-D (calcitriol) — the active hormone. Impaired kidney function, even subclinical renal inflammation, reduces this conversion. The patient has adequate storage form but cannot activate it. Standard testing shows low 25-OH-D; supplementation raises 25-OH-D; active hormone status remains suboptimal because the conversion enzyme is running below capacity.

What the randomized trials found when they tested the hypothesis directly

• VITAL (2019): 2,000 IU of vitamin D3 per day for five years did not reduce cardiovascular events or cancer incidence in healthy adults. The observational associations that generated the supplementation recommendations were not reproduced when vitamin D was isolated as the intervention.
• Bone density (JAMA, 2019): 4,000 IU per day did not improve bone density and worsened it in some measures. The primary clinical justification for supplementation did not hold under controlled conditions.
• Autoimmune and cancer RCTs: Multiple trials testing vitamin D supplementation against MS, type 1 diabetes, and cancer have not replicated the protective associations from observational data.

• A "low" result may not reflect actual vitamin D status. A person with adequate sulfated vitamin D can test low on the standard fat-soluble fraction test. The "deficiency" diagnosis may be measuring a fraction rather than a deficit — which would explain why supplementation often fails to produce the expected clinical improvements despite raising the blood number.
• Supplements raise the tested fraction only. No evidence exists that cholecalciferol supplementation increases circulating 25(OH)D3-sulfate. A supplement that raises the test number without raising the biologically relevant fraction is not addressing the underlying deficiency.
• Sunlight produces both fractions. UVB on skin initiates the photochemical conversion that the body then processes through both the fat-soluble and sulfated pathways. This is the mechanism that supplements entirely bypass. The sulfated form is a product of the full sunlight pathway — not of oral cholecalciferol.
• Vitamin D Binding Protein (GcMAF precursor) preferentially handles the sulfated form. The primary transport protein for vitamin D in the bloodstream handles the sulfated form preferentially. The fat-soluble fraction that supplements raise and tests measure is not the primary circulated form.
• The entire supplementation paradigm was built on half the data. When researchers labeled the sulfated form "inert" in the 1960s because it did not mobilize calcium, they used a single functional marker to dismiss an entire class of vitamin D chemistry. That decision shaped 60 years of clinical practice. The question of what the sulfated form actually does — beyond calcium mobilization — has not been answered.

Seed oils and the sulfation pathway

• The sulfation pathway runs on cholesterol, not PUFA. The production of sulfated vitamin D (25-OH-D3-sulfate) and cholesterol sulfate in skin requires cholesterol as the substrate — specifically, cholesterol in the skin membrane that UVB can act on. Cholesterol is a product of the saturated fat and animal food pathway. Diets high in polyunsaturated seed oils and low in saturated fats and animal foods reduce the cholesterol substrate available in the skin for this photochemical reaction. The body cannot substitute omega-6 linoleic acid for cholesterol in the sulfation process.
• PUFA peroxidation damages the enzyme system. The sulfotransferase enzymes that add sulfate groups to vitamin D metabolites (primarily SULT2B1, expressed in skin, liver, and placenta) are membrane-associated proteins. Their activity depends on the lipid environment of the membrane. Polyunsaturated fats oxidize readily at body temperature, generating lipid peroxidation products (4-HNE, malondialdehyde) that damage enzyme function. A membrane environment dominated by peroxidation-prone PUFA from seed oils is a hostile environment for the enzymes that produce the sulfated vitamin D fraction.
• Vitamin A requires saturated fat for absorption and transport. Retinol is a fat-soluble vitamin that requires dietary fat for absorption from the gut and for transport in chylomicrons. Saturated fats and cholesterol-rich foods — egg yolks, liver, dairy fat, fatty fish — are both the primary dietary sources of preformed vitamin A (retinol) and the fat matrix that enables its absorption. Replacing these with seed oils removes both the source and the transport medium. The depletion of vitamin A further impairs VDR signaling as described above — a cascade in which poor dietary fat quality undermines both the substrate and the receptor function for vitamin D.
• Vitamin A mobilization depends on copper, ceruloplasmin, and zinc. Retinol stored in the liver cannot be released into circulation without retinol-binding protein (RBP). RBP synthesis requires zinc as a cofactor — zinc deficiency impairs the body's ability to mobilize liver retinol stores regardless of how much vitamin A has been consumed. Ceruloplasmin — the copper-dependent ferroxidase that carries approximately 95% of serum copper — also plays a direct role in retinol transport and mobilization from hepatic stores. Copper deficiency reduces ceruloplasmin activity, impairing retinol release. This means a person can have adequate liver stores of vitamin A and still present with functional vitamin A deficiency if zinc or copper status is compromised. Magnesium is required for the enzymatic conversions that activate retinol downstream. The vitamin A pathway is not self-sufficient — it runs on a mineral substrate that modern diets routinely fail to supply.

Isolated vitamin A supplementation does not solve this — and carries its own dangers

• Preformed retinol is teratogenic at supplemental doses. Isolated vitamin A (retinol acetate, retinol palmitate) is one of the most well-documented teratogens in human medicine. Doses above approximately 10,000 IU/day — a level routinely sold over the counter — cause neural crest cell defects, craniofacial abnormalities, and cardiac malformations. The risk window includes the period before a pregnancy is confirmed. Many practitioners who recognize the vitamin A/D co-dependency respond by recommending vitamin A supplements — which introduces a distinct teratogenic risk that the supplement industry does not adequately communicate.
• Isolated retinol accumulates in the liver. Vitamin A is fat-soluble and stored in hepatic stellate cells. Chronic supplementation produces hepatic accumulation leading to hypervitaminosis A: fatty liver, hepatic fibrosis, portal hypertension, and — at very high doses — pseudotumor cerebri (elevated intracranial pressure). These outcomes occur at doses far below what is required to produce clinical signs of toxicity in short-term studies.
• High-dose vitamin A can block vitamin D — the opposite of the intended effect. Vitamin A and vitamin D compete for RXR as a co-receptor. At high supplemental doses, retinol can displace vitamin D from the RXR heterodimer, functionally antagonizing vitamin D signaling. Supplementing isolated vitamin A to "help vitamin D work" at pharmacological doses may achieve the reverse — further impairing the VDR/RXR partnership the intervention was meant to restore.
• Beta-carotene is not the answer either. Beta-carotene (provitamin A from plants) does not accumulate to toxicity the way preformed retinol does — the body regulates conversion to retinol. But in people with common genetic variants in BCMO1 (the conversion enzyme), beta-carotene is poorly converted to retinol, making it an unreliable vitamin A source. Elevated serum beta-carotene in someone with poor conversion can create a false impression of adequate vitamin A status.

The correct answer is whole-food retinol in the amounts and matrix in which the body evolved to receive it: liver (beef, chicken, cod), pasture-raised egg yolks, and grass-fed butter. Cod liver oil delivers vitamins A and D together in their evolutionary ratio — but most commercial cod liver oils add synthetic cholecalciferol (the same isolated vitamin D3 discussed throughout this article) to standardize the label claim, because natural vitamin D content in the oil is too variable to control. This defeats the purpose entirely: you are now getting isolated supplemental vitamin D on top of whatever natural D the oil contains, with no feedback regulation. Fermented cod liver oil adds rancidity and processing transparency concerns on top of that. Eating the actual cod liver — canned, in pâté, or fresh — delivers the full fat-soluble vitamin matrix without the added synthetic vitamin D.

• 25-hydroxyvitamin D (25-OH-D) — the standard test. This measures the circulating storage form: the form the liver produces by hydroxylating cholecalciferol, before the kidneys convert it to the active hormone. It reflects how much vitamin D has been taken in (from supplements, food, or sun) and is sitting in reserve. It does not measure whether that reserve is being converted and used. This is the test used to diagnose deficiency, justify supplementation, and set dosing targets. It is a measurement of inventory, not function.
• 1,25-dihydroxyvitamin D (1,25(OH)2D, calcitriol) — the active hormone test. This measures the biologically active form — the hormone that actually binds vitamin D receptors, regulates calcium absorption, modulates immune function, and drives gene transcription. This test is available from most clinical labs but is rarely ordered alongside the standard test. The two numbers together tell a story that neither tells alone.

Seed oils make burning almost inevitable — dietary preparation comes first

A person eating a high seed oil diet (canola, soybean, sunflower, corn, safflower) cannot build a solar callus at a normal rate — because their skin membranes are loaded with polyunsaturated fatty acids that oxidize under UV before the skin's adaptive response can engage. The burning they experience is not sun sensitivity — it is lipid peroxidation in the skin cell membranes. The UV that would stimulate melanin upregulation is instead being spent generating 4-HNE and malondialdehyde in the membrane lipids. Adaptation cannot outpace the chemical damage.

Transitioning away from seed oils toward stable fats (butter, tallow, lard, olive oil, coconut oil) changes skin membrane composition over months to years — the adipose half-life of linoleic acid is approximately six years. Sun tolerance improves as the membrane composition shifts. This is the dietary prerequisite to building a functional solar callus, not a parallel project. Begin the fat transition before aggressively extending sun exposure.