Is Vitamin D Actually a Hormone?

The reaction that starts the whole system happens in the deepest living layer of the skin, and it uses no enzyme at all. A photon of ultraviolet-B light, wavelength somewhere between 290 and 315 nanometres, arrives at a molecule of 7-dehydrocholesterol held in a cell membrane. That molecule is a steroid: four carbon rings fused edge to edge, the same chassis that carries cortisol, testosterone and oestradiol. The photon breaks one bond in it. The B ring springs open. What remains is previtamin D3, which then folds itself, over hours and without further light, into vitamin D3.

Leave the light on and the chemistry turns cautious. Previtamin D3 that keeps absorbing ultraviolet is not stockpiled; it is diverted into two biologically inert isomers, tachysterol and lumisterol, and the skin’s output flattens out. The vitamin D3 already made is drawn off the membrane onto a carrier in the blood, vitamin D binding protein, and delivered to the liver. Nothing in that sequence has done any biological work yet. The molecule is still inert.

None of this is how a vitamin behaves. A vitamin, by definition, is a compound the body cannot synthesise and must take in from food. This one is manufactured by an organ, exported on a carrier protein, activated by dedicated enzymes in two further tissues, read by a receptor inside the cell nucleus, and destroyed by an enzyme that its own product switches on. That is a hormone in every particular, and the distinction is not a semantic one. Vitamin implies topping up a shortfall, where more input means more effect. Hormone implies a regulated loop with a set point, where past a certain intake the body simply adjusts. Read the last fifteen years of vitamin D trials with the second model in mind and a run of apparent disappointments starts to look like confirmation.

Why it was ever called a vitamin

The name is a fossil of how the molecule was found. Rickets became a public health problem as populations moved from farms into industrial cities, and the cure arrived well before the mechanism: cod liver oil, and foods that had been exposed to ultraviolet light, prevented and reversed the disease. The active principle was isolated out of that work and catalogued as a vitamin, because from the outside that is precisely what it looked like — a substance in food whose absence caused a deficiency disease. That the skin builds it from cholesterol using sunlight, and that the compound circulating in blood is a precursor rather than the working molecule, came later. The classification never caught up.

Three organs, two hydroxylations, one hormone

Vitamin D3 reaching the liver is hydroxylated at carbon 25. Several enzymes can perform this step; CYP2R1 is the important one. The product, 25-hydroxyvitamin D, is the principal circulating form and the metabolite laboratories measure when they report a vitamin D level. It is a reservoir. It is not the hormone.

The second hydroxylation is the regulated one. It happens principally in the kidney, catalysed by CYP27B1, and it yields 1,25-dihydroxyvitamin D — calcitriol, the form responsible for most of vitamin D’s biological actions. Renal production is governed the way any endocrine output is governed: parathyroid hormone stimulates the enzyme, while calcium, phosphate and the bone-derived hormone FGF23 suppress it. Calcitriol also regulates its own disposal, inducing CYP24A1, the 24-hydroxylase that breaks down both the active hormone and its precursor. Supply rises; the off-switch is thrown.

Both metabolites travel bound to vitamin D binding protein and albumin. For most tissues it is the free, unbound fraction that crosses into the cell; kidney and parathyroid are the exceptions, taking up the protein-bound form through a megalin and cubilin mechanism. Inside the cell, calcitriol binds the vitamin D receptor, which partners with the retinoid X receptor and acts as a transcription factor, raising or lowering gene expression depending on the cofactors it recruits. That receptor is not confined to bone, gut and kidney. It is present in most cell types, which is why a molecule catalogued as a bone nutrient has a mechanistic claim on immunity, epithelial biology and a good deal else.

The kidney is not even the only site of activation. Keratinocytes, macrophages, parathyroid cells, intestinal epithelium, prostate and breast tissue all carry CYP27B1 and make calcitriol locally, for their own use. That local production answers to cytokines such as tumour necrosis factor alpha and interferon gamma rather than to parathyroid hormone: a paracrine and autocrine system running alongside the endocrine one, on different instructions.

Scoring it against the definition

Set the textbook criteria for an endocrine hormone beside what is known about this molecule.

The last row is the only place the older label holds. Vitamin D is a conditional dietary essential: a hormone that becomes a nutrient whenever ultraviolet exposure is inadequate — high latitude, winter, indoor work, sun avoidance, heavily pigmented skin, older skin that synthesises less of it. For a large share of people that condition is met for months at a time, which is why the misnomer has survived so comfortably in practice.

What the big trials found, and why it fits

If vitamin D behaved like a true vitamin, supplementing it should scale: more substrate, more effect. If it behaves like a hormone, then in people who are already replete the loop absorbs the extra and the clinical needle does not move. The trials read as the second story.

VITAL randomised 25,871 US adults to 2,000 IU of vitamin D3 daily or placebo, with a median 5.3 years of follow-up. It missed both primary endpoints: invasive cancer of any type (hazard ratio 0.96, 95% confidence interval 0.88 to 1.06) and major cardiovascular events (0.97, 0.85 to 1.12). Death from any cause was unchanged (0.99, 0.87 to 1.12). The secondary and ancillary analyses were more interesting than the headline. Advanced cancer, meaning metastatic or fatal disease, was lower in the vitamin D arm (0.83, 0.69 to 0.99), an effect concentrated in participants of normal body mass index (0.62, 0.45 to 0.86) and absent in those with obesity (1.05, 0.74 to 1.49). Confirmed incident autoimmune disease was lower as well (0.78, 0.61 to 0.99).

D-Health took the opposite approach to dosing and got a colder answer. It gave 21,315 Australians aged 60 and over a 60,000 IU capsule once a month for five years, without screening anyone first. In the narrow sense it worked: mean 25-hydroxyvitamin D reached 115 nmol/L against 77 in the placebo group. All-cause mortality did not move (1.04, 0.93 to 1.18). Cancer mortality ran numerically higher (1.15, 0.96 to 1.39), and in an exploratory analysis excluding the first two years of follow-up it reached 1.24 (1.01 to 1.54). The investigators concluded that the precautionary principle argues against this regimen in people who are already replete, and the placebo group’s own blood levels say how replete that was.

The clearest evidence that schedule matters independently of total dose comes from a trial of 2,256 community-dwelling women aged 70 and over at high fracture risk, given a single 500,000 IU dose of cholecalciferol each autumn or winter for three to five years. Falls rose (incidence rate ratio 1.15, 1.02 to 1.30) and so did fractures (1.26, 1.00 to 1.59). The timing is the tell: the rate ratio for falling was 1.31 in the three months after each dose and 1.13 across the following nine. In the substudy, median 25-hydroxyvitamin D climbed from 49 nmol/L at baseline to roughly 120 a month after dosing, then drifted down. A spike and a decay, imposed on a system built around a set point, produced harm where smaller daily doses have not.

The longevity result, kept in proportion

The finding that draws the most attention in this field is a telomere sub-study nested inside VITAL. Among 1,054 participants evaluated in person, across 2,571 leukocyte samples analysed at baseline, year two and year four, daily vitamin D3 reduced telomere attrition by 0.14 kilobase pairs over four years relative to placebo (95% confidence interval 0.007 to 0.27). Marine omega-3 fatty acids had no significant effect on telomere length at either time point. The trial team put the effect at 140 base pairs of erosion prevented; press coverage converted that into roughly three years of cellular ageing, a translation the paper’s own conclusion does not make.

The framing matters more than the number. Leukocyte telomere length is a surrogate marker, not a clinical outcome. This was an ancillary study inside a trial that missed its primary endpoints. The effect sits close to its lower confidence bound. And the same participants showed no reduction in cancer incidence or cardiovascular events. Slowing telomere attrition is a coherent biological observation about a hormone with a receptor in most cell types. It is not a demonstration that anyone lived longer.

What the guidelines did with all this

The 2024 Endocrine Society guideline on vitamin D for disease prevention reads like a document written about a regulated hormone rather than a nutrient with a deficit to fill. It suggests supplementation above the dietary reference intakes for four groups: children and adolescents aged 1 to 18, adults aged 75 and over, people who are pregnant, and people with high-risk prediabetes. It suggests against empiric supplementation above those reference intakes in healthy adults under 75. It suggests against routine 25-hydroxyvitamin D testing in every population it examined, including people with obesity or darker complexion, because the trial evidence did not identify a target level that reliably guides a decision. And for non-pregnant people over 50 who do have an indication, it prefers daily dosing to intermittent high doses.

Those are population-level recommendations about prevention in generally healthy people, and parts of them — the testing position above all — are contested. They say nothing about managing diagnosed deficiency, malabsorption, chronic kidney disease or the other clinical situations where measuring and treating is plainly the right thing to do.

The practical shape of the answer

Calling vitamin D a hormone changes the question worth asking. Not how much can be taken, but whether a person is replete and whether the schedule is one the system can absorb. Repletion is where the evidence of benefit lives. Past it, the extra is argued with by an enzyme. Daily dosing has the better safety record, the guidelines now say so explicitly for older adults, and the one trial that pushed bolus logic to its limit produced more falls and more fractures rather than fewer.

Which returns to the skin. The dose made by sunlight is the only one that cannot overshoot, because the reaction regulates itself: past a certain exposure the photons stop making the molecule and start making inert isomers instead. A capsule has no such mechanism. That asymmetry — a hormone the body knows how to stop producing, delivered in a form it cannot refuse — is the whole practical content of the naming argument.

Educational, not medical advice.