Vitamin D: Beyond Bone Health

Vitamin D occupies an odd position. It is one of the most-ordered tests in preventive medicine, one of the most commonly low results, one of the cheapest things to correct — and the intervention whose large trials have been among the most consistently disappointing. Both halves are true, and holding them together is what separates a useful understanding from the two popular caricatures: that vitamin D fixes everything, or that it does not matter.

Not really a vitamin

A vitamin is something the body cannot make and must obtain from food. Vitamin D fails that definition. It is synthesised in skin from a cholesterol precursor when UVB photons break a ring in the molecule, and it functions as a steroid hormone — binding a nuclear receptor and altering the transcription of a large number of genes. The name stuck because it was discovered while researchers hunted for a dietary factor that prevented rickets.

Activation takes two steps in two organs. The liver hydroxylates it to 25-hydroxyvitamin D, the main circulating form with a half-life measured in weeks. The kidney then adds a second hydroxyl to produce calcitriol, the active hormone, in a step tightly regulated by parathyroid hormone, calcium and phosphate. Many other tissues — immune cells, prostate, breast, colon — carry the same activating enzyme and make small amounts of calcitriol locally for their own use, which is the mechanistic basis for interest in effects beyond bone (Holick, N Engl J Med 2007).

Why the test measures the inactive form

It seems backwards to measure the storage form rather than the active hormone, but it is the correct choice. Calcitriol has a half-life of hours and is homeostatically defended: as stores fall, parathyroid hormone rises and drives more conversion, so calcitriol can look normal — or high — in someone frankly depleted. Measuring it would produce false reassurance at exactly the wrong moment.

25-hydroxyvitamin D behaves like a fuel gauge instead. It is not defended to a set point, it reflects cumulative supply over the preceding weeks, and it is what every reference range and every trial has used.

Why deficiency is so common

Skin synthesis has a hard physical constraint. UVB is filtered out at low solar angles, so above roughly 35-37 degrees of latitude essentially no vitamin D is made in skin through the winter, regardless of how long anyone stands outside. Sunscreen, glass, clothing and being indoors at midday remove more of what remains.

On top of that sit several biological modifiers. Melanin competes for the same photons, so darker skin requires substantially longer exposure for the same synthesis. Skin synthesis declines with age as the precursor becomes less abundant. And vitamin D is fat-soluble, so in obesity it distributes into a much larger adipose compartment and serum concentration falls for the same total body content — a dilution effect rather than a true deficiency of substrate, though the measured level is genuinely lower (Wortsman et al., Am J Clin Nutr 2000). Malabsorption, from coeliac disease or bariatric surgery, removes the dietary route as well. Winter deficiency rates in the region of 40-80% in temperate populations follow directly from those constraints (Holick et al., J Clin Endocrinol Metab 2011).

The part that is usually left out

Observational studies have linked low vitamin D to almost everything: cardiovascular disease, cancer, depression, autoimmune conditions, infection, low testosterone, all-cause mortality. The associations are consistent and biologically plausible, given that the receptor is present in most tissues.

Then the large randomised trials arrived, and mostly did not confirm them. The biggest of them tested supplementation against cancer and cardiovascular disease in a general adult population followed for years and found no reduction in either primary endpoint (Manson et al., N Engl J Med 2019). Trials of fracture prevention, depression and several other endpoints have been similarly unimpressive in replete populations. The comprehensive reviews of skeletal and extra-skeletal actions now describe a field where the mechanistic case remains strong and the interventional case is much weaker than the observational literature implied (Bouillon et al., Endocr Rev 2019).

The likeliest explanation is reverse causation plus confounding. Low vitamin D is a good marker of being ill, sedentary, indoors, heavier and eating poorly — all of which cause disease independently — so supplementing corrects the marker without touching the causes. There is also a floor effect: most trial participants started replete, and giving more of a nutrient to people who already have enough should not be expected to do much.

The defensible conclusion is narrower than the marketing and broader than the nihilism. Vitamin D has a genuine deficiency state with genuine consequences for bone and muscle. Correcting a deficiency is worth doing; treating it as a therapy for the non-deficient is not supported.

Reference versus optimal, and why bodies disagree

• Reference "normal": 30-100 ng/mL on most laboratory reports
• Commonly used optimal target: 40-60 ng/mL
• Below 30: deficient
• Below 20: severely deficient, the range at which skeletal consequences are well established
• Above 100: potentially excessive — achievable from supplements, essentially never from sun

Expert bodies place the threshold differently, and the disagreement is not arbitrary. One camp defines adequacy as the level at which parathyroid hormone stops rising and population bone health is protected — a lower cut-off. The other defines it by outcomes across the wider range of vitamin D-responsive tissues — a higher one. The first has better trial support, the second better mechanistic reasoning. This is the tension running through most laboratory interpretation, discussed in optimal versus normal lab ranges.

Supplementation, practically

• Typical maintenance dosing sits around 2,000-5,000 IU daily, adjusted to the measured result rather than guessed.
• Severe deficiency is repleted faster, in the region of 5,000-10,000 IU daily, then reduced once the level is restored.
• Take it with a meal containing fat — absorption is meaningfully better than on an empty stomach.
• D3 (cholecalciferol) raises 25-hydroxyvitamin D more effectively than D2 and is the preferred form (Tripkovic et al., Am J Clin Nutr 2012).
• Vitamin K2 is often paired on the argument that it directs calcium toward bone rather than arteries. The reasoning is plausible; the outcome evidence in people who are not deficient in K is thin. Reasonable to take, not something to claim benefit from.
• Magnesium is a required cofactor for the enzymes that activate vitamin D, and magnesium insufficiency is common — see magnesium.

Monitoring, and the two ways this goes wrong

Test at baseline. Retest at around three months after starting or changing a dose — long enough for the level to plateau given the multi-week half-life, short enough to fix an inadequate dose before another winter. Once stable, annually is enough, and drawing at the same time of year matters, since summer and winter results in the same person differ substantially. Sunlight entrains other systems too, covered in sunlight and hormones.

Two failure modes recur. The first is supplementing indefinitely without ever measuring, and either never reaching an adequate level or drifting well above it; sustained very high levels cause hypercalcaemia, and the route there is always supplementation, never sun. The second is treating vitamin D as the answer to symptoms it was never going to fix — correcting a level from 22 to 55 and being disappointed that fatigue or poor performance persisted, when the driver was sleep, iron, thyroid or training load. What vitamin D does covers the realistic scope.

Bottom line

Vitamin D is a hormone measured by its storage form, and deficiency is genuinely common because skin synthesis has physical limits that modern life makes worse. Correcting a real deficiency is cheap, safe and worth doing, particularly for bone and muscle. What the large randomised trials do not support is the wider claim that supplementing people who are already replete prevents cancer, cardiovascular disease or much else — the observational associations were substantially confounded. Test, correct what is low, retest at three months, and do not expect the correction to solve problems that belong to sleep, iron, thyroid or training.

Educational content, not medical advice. Laboratory interpretation and any treatment decision are made by a licensed physician after individual evaluation. Individual results vary.