How Your Skin Makes Vitamin D
Vitamin D is not really a vitamin — it is a hormone your own body manufactures, on an assembly line that runs through three organs. Watch UVB photons (290–315 nm) strike 7-dehydrocholesterol in your skin, follow the molecule to the liver where it becomes 25(OH)D — the storage form your blood test measures — and then to the kidney, where parathyroid hormone switches on the final step that makes the active hormone, calcitriol.
Try this: start in Summer sun and watch 25(OH)D climb to about 50 ng/mL over the simulated weeks. Then switch to Winter / high latitude and watch it fall — no matter how long you stay outside. Then add Chronic kidney disease and notice the trap: 25(OH)D still reads “sufficient” while the active hormone collapses.
Blood & hormone readout
What's happening
The Science in Plain Language
1. It is a hormone, not a vitamin
A vitamin, by definition, is something you must eat because your body cannot make it. Vitamin D fails that test. Given enough ultraviolet light, your skin manufactures it from cholesterol — and what it produces is a secosteroid hormone, chemically a cousin of cortisol and oestrogen, that travels in the blood, binds a nuclear receptor, and switches genes on and off. It was mislabelled a “vitamin” in the 1920s, when rickets was cured with cod-liver oil before anyone realised sunlight did the same job. The name stuck. The biology did not change.
That matters practically, because hormones are made on an assembly line. Vitamin D needs three organs in sequence — skin, liver, kidney — and each hand-off can fail independently. Most confusion about vitamin D comes from treating it as one substance when it is really three different molecules with three different jobs and three wildly different half-lives.
2. Step one — the skin, and why latitude and season decide almost everything
Deep in your epidermis, in the basal and spinous layers, sits a cholesterol precursor called 7-dehydrocholesterol (7-DHC). When a photon of ultraviolet B — wavelengths 290 to 315 nm, with the peak effect around 295–300 nm — hits it, the photon’s energy snaps open one of the molecule’s rings. The result is previtamin D3, which then rearranges itself over several hours, driven simply by your body heat (thermal isomerisation), into cholecalciferol — vitamin D3. That molecule slips into a dermal capillary and rides away on a carrier protein.
The catch is that UVB is fragile. It is absorbed by ozone far more strongly than the UVA that tans and ages your skin. When the sun is high, the light takes a short, steep path through the ozone layer and plenty of UVB survives. When the sun is low — in winter, at high latitude, in early morning, in late afternoon — the same light must cross a much longer slice of atmosphere, and the UVB is almost entirely stripped out before it reaches the ground. That is why, above roughly 35–37° latitude, there is a period each year that researchers call the “vitamin D winter”: at Boston’s latitude (42°N) essentially no vitamin D is made in the skin from about November to February, and further north the window is longer still. Classic experiments in the 1980s exposed 7-DHC in test tubes on rooftops at different latitudes and months and found exactly this — production simply stops.
This is the single most under-appreciated fact about vitamin D: in winter, at high latitude, standing outside longer does not help. The photons you need are not arriving. Switch the animation to Winter / high latitude and watch the UVB fizzle out inside the atmosphere — the synthesis meter collapses to under 1%, and 25(OH)D drifts down no matter what. A useful rule of thumb, shown by the little figure and its shadow: if your shadow is longer than you are tall, the sun is too low to make much vitamin D.
Three other things stop UVB before it reaches 7-DHC:
- Glass. Ordinary window glass transmits UVA but blocks essentially all UVB. Sitting in a sunny window, or driving with the sun on your arm, produces effectively no vitamin D — while still delivering the UVA. A conservatory is not a substitute for a walk.
- Melanin. Melanin is an extremely effective natural UVB filter — that is its evolutionary job. Deeply pigmented skin therefore needs substantially longer sun exposure — commonly cited as roughly three to six times longer — to make the same amount of vitamin D as fair skin under identical conditions. This is real, it is not a small effect, and it is a major reason why vitamin D deficiency is more common in dark-skinned people living at high latitudes. Switch to Darker skin and watch the photons being swallowed by the melanin granules before they reach the 7-DHC.
- Sunscreen. Sunscreen is designed to absorb UVB, and in the laboratory — applied at the full test thickness of 2 mg/cm² — SPF 30 blocks around 97% of it. The animation shows that idealised case. But be careful with the conclusion. In the real world people apply sunscreen far more thinly than the test protocol, miss patches, and do not reapply — and studies of everyday sunscreen users have generally not found reduced vitamin D levels. Meanwhile sunscreen’s protection against skin cancer and photoageing is genuinely well established. The honest summary: keep using sunscreen. If you are worried about vitamin D, take a supplement — do not get sunburnt instead.
3. You cannot overdose on vitamin D from the sun
Watch the grey particles drifting away from the skin. Sunlight does not just make previtamin D3, it also destroys it. The same UVB that creates previtamin D3 converts it onward into lumisterol and tachysterol, which are biologically inert, and photodegrades any vitamin D3 still sitting in the skin into suprasterols. Within about 10–20 minutes of bright sun the reactions reach a photoequilibrium in which no more net vitamin D accumulates — only a modest fraction of the available 7-DHC ever ends up as usable D3, and the rest is shunted into the inert products.
This is an elegant built-in safety valve. No amount of sunbathing can cause vitamin D toxicity. It also means the extra hours have no benefit — you get the sunburn and the skin-cancer risk with none of the vitamin D. Brief, regular, non-burning exposure is all the skin route can ever give you.
4. Step two — the liver: the storage form, and the number on your lab report
Vitamin D3 from skin (or from a capsule, or from oily fish) travels to the liver, where the enzyme 25-hydroxylase (CYP2R1) adds a hydroxyl group to give 25-hydroxyvitamin D — 25(OH)D. This step is barely regulated: the liver hydroxylates roughly whatever it is given. The product circulates bound to vitamin D binding protein with a half-life of about two to three weeks.
Because it is abundant, stable and long-lived, 25(OH)D is the molecule that reflects your total supply — sun plus diet plus supplements — averaged over the preceding weeks. That is why 25(OH)D is the vitamin D blood test, and why “my vitamin D level” always means this number. It is also why levels move slowly: start a supplement and it takes roughly 8 to 12 weeks to reach a new plateau. Watch the tank in the animation fill and drain over the simulated weeks — it never jumps.
5. Step three — the kidney: the active hormone, and a trap worth knowing
25(OH)D is a reservoir, not a worker. The active hormone is made in the kidney, where 1α-hydroxylase (CYP27B1) converts 25(OH)D into calcitriol, 1,25(OH)₂D. Unlike the liver step, this one is tightly regulated. The gate in the animation is pushed open by parathyroid hormone (PTH), released when calcium is low, and pushed shut by FGF23 (secreted by bone osteocytes) and by calcitriol itself — a classic negative-feedback loop, which also switches on CYP24A1, the enzyme that destroys both 25(OH)D and calcitriol. Calcitriol’s half-life is only a matter of hours, and normal levels are roughly 20–60 pg/mL — a thousand times lower than 25(OH)D.
Now the trap. Because PTH rises when vitamin D is low, and PTH drives the kidney enzyme harder, calcitriol often looks perfectly normal — or even high — in someone who is genuinely vitamin D deficient. Try it: set the scenario to Winter and watch 25(OH)D fall into the deficient band while calcitriol stubbornly stays in range and PTH climbs. That compensation is not free — the high PTH is quietly pulling calcium out of your bone to keep the blood level normal. This is precisely why ordering a calcitriol level to check someone’s vitamin D status is a mistake: it can be reassuringly normal in florid deficiency. Order 25(OH)D.
The mirror-image trap is chronic kidney disease. Turn it on. The 25(OH)D tank stays comfortably full — the skin and liver are fine, the blood test says “sufficient” — but the damaged kidney cannot perform the final activation, and FGF23 (which climbs steeply in CKD) slams the gate further shut. Calcitriol collapses, calcium absorption falls, and PTH runs away into the hundreds — secondary hyperparathyroidism, the core of CKD–mineral and bone disorder. Someone can be functionally vitamin D deficient with a completely normal vitamin D test. That is a genuinely under-appreciated and clinically important distinction, and it is why people with advanced CKD are sometimes given activated vitamin D (calcitriol or an analogue) rather than ordinary D3.
6. Magnesium — the cofactor nobody mentions
Every enzyme in this pathway — the liver’s 25-hydroxylase, the kidney’s 1α-hydroxylase, the CYP24A1 that switches it all off, and the vitamin D binding protein that carries the hormone around — is magnesium-dependent. So is the secretion of PTH, and so is the ability of your tissues to respond to PTH. Turn on Low magnesium: the enzyme sockets on the liver and kidney empty out, the gate half-closes, and calcitriol falls even though the substrate is there and the sun is shining.
Clinically this shows up as vitamin D supplementation that stubbornly “doesn’t work,” and in severe hypomagnesaemia as a low calcium that will not correct with calcium or vitamin D until the magnesium is replaced — because the parathyroid glands cannot secrete PTH properly and the tissues cannot hear it. If someone is not responding to vitamin D, magnesium is a reasonable thing to check.
7. What calcitriol actually does — and what the trials really show
Calcitriol enters a target cell, binds the vitamin D receptor (VDR), pairs with the retinoid X receptor, and the complex docks onto DNA to switch genes on. In the intestine it induces the calcium channel TRPV6 and the shuttle protein calbindin, and the result is measurable: fractional calcium absorption rises from roughly 10–15% when you are deficient to roughly 30–40% when you are replete. It also acts on bone, kidney, and immune cells — in macrophages the VDR switches on the antimicrobial peptide cathelicidin, which is well documented in the laboratory.
Here is where honesty matters, because vitamin D is surrounded by overclaiming. The evidence for bone is solid: severe deficiency causes rickets in children and osteomalacia in adults, and correcting deficiency fixes them. That is not in dispute.
The rest is far weaker than the internet suggests. Large, well-conducted randomised trials in already-replete people have been largely negative. VITAL (about 25,900 US adults, 2,000 IU/day of D3 for around five years) found no reduction in invasive cancer or in major cardiovascular events; its bone sub-study found no reduction in fractures in generally healthy adults who were not deficient. D-Health (about 21,000 older Australians given a large monthly dose for five years) found no reduction in all-cause mortality. Trials of vitamin D for preventing respiratory infection have been mixed at best, with the most credible signal confined to people who were deficient to start with. The pattern is consistent and worth internalising: correcting a deficiency helps; topping up someone who is already replete generally does not. Vitamin D is a repletion therapy, not a tonic.
And the danger runs the other way too. Vitamin D toxicity from high-dose supplements is real. Excess D3 — typically sustained megadoses, or a dosing error — causes hypercalcaemia: nausea, vomiting, confusion, excessive thirst and urination, kidney stones, and kidney damage. It is generally seen at 25(OH)D levels well above 100–150 ng/mL. The sun cannot do this to you. A bottle can. The IOM sets a tolerable upper intake of 4,000 IU/day for adults; there is no good reason for most people to exceed it without a specific medical indication and monitoring.
8. How much is enough — a real disagreement, honestly stated
The bands on the readout (deficient <20, insufficient 20–29, sufficient 30–50 ng/mL) follow the Endocrine Society, which targets ≥30 ng/mL. The Institute of Medicine / National Academies, reviewing the same literature, concluded that 20 ng/mL is sufficient for the bone health of 97.5% of the population, that most people need 600 IU/day (800 IU over 70), and that pushing everyone above 30 is not supported by the evidence. Both bodies are serious; they weigh the same trials differently. There is no consensus here, and anyone who tells you there is, is selling something. A pragmatic reading: below 20 ng/mL, act. Between 20 and 30, it depends on your risk. Above 30, more is not better.
Units are a common trap: ng/mL × 2.5 = nmol/L. A “level of 30” in the US and a “level of 75” in Europe are the same number. And if you do supplement, remember what the animation shows: the effect is slow (weeks), bypasses the skin entirely (entering at the liver step), and depends on a kidney that works and a magnesium status that is adequate.
This page is an educational model, not medical advice. Values are typical textbook ranges, and the simulation is deliberately simplified — a real person’s response depends on body fat, age, kidney and liver function, medicines, and genetics. Reference ranges vary between laboratories. If you are concerned about your vitamin D, ask for a 25(OH)D test and discuss the result with your clinician.