Vitamin D

Best-in-class: Vitamin D-5,000

A pro-hormone governing calcium/bone metabolism and thousands of genes involved in immune and muscle function. D3 is the form your skin makes from sunlight; K2 helps direct calcium into bone rather than arteries.

How Vitamin D actually works

Vitamin D isn't really a vitamin, it's the precursor to a steroid hormone. What you swallow is inert: the liver converts it to 25-hydroxyvitamin D — the storage form, and the one your blood test measures — and the kidney then makes the active hormone calcitriol under tight parathyroid control. Calcitriol binds the vitamin D receptor, which is a transcription factor present in most tissues including immune cells, so it directly regulates hundreds of genes. That breadth is why observational studies link low D to almost everything, and also why supplementing rarely reproduces those associations: the deficiency is often a marker of illness rather than its cause.

Where to get Vitamin D

The evidence for Vitamin D

Graded by what exists behind each claim.

✅ Clinically validated

• Correcting deficiency reliably raises serum 25-OH-D and supports bone mineral density (established in RCTs).
• Meta-analyses show modest reduction in acute respiratory infection risk, strongest in deficient individuals dosed daily.
• Reduces falls/fractures in deficient older adults when combined with adequate calcium.

📊 Correlative data

• Low vitamin D status is associated with higher all-cause mortality, autoimmune disease, depression and metabolic dysfunction across large cohorts — though causality is debated.

🧪 Theoretical / extrapolated benefits

• Extensive immunomodulatory and anti-cancer mechanisms are proposed from the near-ubiquitous VDR receptor, but large RCTs (e.g., VITAL) have not confirmed broad disease-prevention in replete people.

How to read these tiers: they say how much human evidence exists, not how well something works — and ✗ flags harm, never a disappointing trial. How the evidence tiers work →

What Vitamin D actually does

The molecule in the capsule is not a hormone and does not become one until two hydroxylations later. Cholecalciferol is a secosteroid with no meaningful affinity for the vitamin D receptor. The liver adds a hydroxyl at carbon 25, mostly through the cytochrome CYP2R1, to give 25-hydroxyvitamin D — the circulating store, the thing a lab measures, and still not the active molecule. The kidney then adds a second hydroxyl at carbon 1 through CYP27B1 to give 1,25-dihydroxy­vitamin D, which is the ligand. Only the first of those three steps is under the control of the person swallowing the capsule.

The second step is under the control of parathyroid hormone, and that is the whole reason more substrate does not mean more signal. CYP27B1 is transcriptionally driven by parathyroid hormone and suppressed by fibroblast growth factor 23 and by its own product. A person whose calcium is normal has low parathyroid hormone drive, so the kidney converts a small and tightly regulated fraction of whatever 25-hydroxyvitamin D is presented to it. Raising the substrate pool from 25 to 55 ng/mL therefore raises the measured number by more than twofold and the active hormone by very little, which is the single most useful fact on this page.

What the active hormone actually does at the gene is switch on a calcium transport apparatus. The liganded vitamin D receptor heterodimerizes with the retinoid X receptor and binds vitamin D response elements in the promoters of TRPV6, calbindin-D9k and the basolateral calcium pump, which together move calcium across the duodenal enterocyte. That pathway matters most when dietary calcium is scarce; when calcium intake is generous, most absorption is passive and paracellular and does not need the receptor at all.

There is a disposal enzyme, and it is induced by the thing it disposes of. CYP24A1 is a 24-hydroxylase that converts both 25-hydroxyvitamin D and the active hormone into calcitroic acid for biliary excretion, and 1,25-dihydroxyvitamin D induces its own catabolic enzyme. That is a negative feedback loop with a gain, and it is why the dose-response for 25-hydroxyvitamin D flattens: each increment of intake meets a larger disposal capacity than the last Heaney 2003.

The receptor is expressed almost everywhere, which is the source of most of the claims and none of the evidence. Immune cells, myocytes, keratinocytes and beta cells all transcribe the vitamin D receptor, and several of them express CYP27B1 and make the active hormone locally in a paracrine fashion. That is a real and interesting biology. It is also the step where a mechanism paper becomes a marketing claim, because local production is regulated by local signals and not by the serum pool a supplement raises Bouillon 2022.

Cell, rodent, human — and where it stops

This is the best-tested supplement in the catalog, and that is exactly why it is in this cohort. The chain from cell to human is complete, the human end of it is enormous, and it disagrees with the cell end.

In the dish and in the animal, the signal is unambiguous. Vitamin D response elements in the TRPV6 and calbindin promoters were mapped in cell culture; receptor-null and CYP27B1-null mice develop rickets, hypocalcemia and secondary hyperparathyroidism that a high-calcium rescue diet largely reverses. The animal work establishes that the pathway is necessary. It does not establish that more of it is better, because a knockout tests the floor and a supplement tests the ceiling.

In humans the dose-to-marker relationship is settled to two decimal places. Extended oral dosing with cholecalciferol produces a predictable, saturable rise in serum 25-hydroxyvitamin D, with the increment per 1,000 IU falling as baseline rises Heaney 2003. Anyone can reproduce that. It is the most reliable result in nutritional supplementation and it is a measurement of the substrate pool, not of an outcome.

Then the outcome trials arrived and the transfer broke. VITAL randomized 25,871 adults to 2,000 IU daily or placebo for a median of 5.3 years and found no reduction in invasive cancer or major cardiovascular events Manson 2019; the ancillary fracture analysis in the same cohort found no reduction in total, nonvertebral or hip fracture LeBoff 2022. A meta-analysis of community-dwelling older adults had already reached the same place for calcium and vitamin D together Zhao 2017. The review of the mega-trials states the position plainly: the recent large trials in unselected, largely replete populations have not shown fracture or fall benefit Dawson-Hughes 2024.

The obstacle to transfer has a name and it is baseline. VITAL did not require a low 25-hydroxyvitamin D to enter, and mean baseline was around 30 ng/mL — a population that was, by the trial's own measurement, mostly not deficient. Prespecified subgroup work in the same trial finds the effect estimates shifting in populations defined by kidney function Limonte 2022. A null result in replete people is evidence about repletion, and the honest reading is not that vitamin D does nothing but that giving it to people who already have enough does nothing.

And where a high dose was given to replete people, the bone result went the wrong way. Three years of 400, 4,000 or 10,000 IU daily in healthy adults with normal baseline status produced a dose-dependent LOSS of volumetric bone mineral density at the radius and tibia, with no gain in bone strength Burt 2019. That is the cleanest available demonstration that the marker and the outcome can move in opposite directions.

Vitamin D — which form, and does it matter

D3 against D2 is a real difference and it is a half-life difference, not a potency difference. Cholecalciferol and ergocalciferol are both 25-hydroxylated, but 25-hydroxyvitamin D3 binds vitamin D binding protein more tightly than the D2 metabolite, so it is cleared more slowly and holds a higher steady-state concentration per unit dose. Both raise the number; D3 holds it. On a weekly or monthly schedule the difference is large, on a daily schedule it is modest.

The oral barrier is fat, and it is the one variable most people get wrong. Cholecalciferol is lipophilic, packaged into chylomicrons and absorbed through the lymphatic route, which bypasses the portal vein and therefore escapes hepatic first-pass metabolism in the usual sense. Oral bioavailability is materially higher when the capsule is taken with a fat-containing meal, and it is depressed by orlistat, bile acid sequestrants and fat malabsorption of any cause. Softgels in an oil carrier behave better than dry powder tablets for the same reason.

The two clocks that matter are two weeks and four hours. 25-hydroxyvitamin D has an elimination half-life on the order of two to three weeks, which is why a retest inside 8 weeks of a dose change reads a moving target and why a single missed day is irrelevant. 1,25-dihydroxyvitamin D has a half-life of roughly 4 to 6 hours, which is why it is useless as a status test and why the lab that measures it is answering a different question. Degradation of both runs through the cytochrome CYP24A1 to calcitroic acid, with biliary rather than renal clearance carrying most of the load.

Bolus dosing is a different drug. A 500,000 IU annual oral bolus and an intramuscular injection of a comparable amount both raise 25-hydroxyvitamin D and both have been associated with more falls and fractures rather than fewer in trial settings, which is the strongest hint in this literature that the shape of the exposure curve matters as much as the area under it Dawson-Hughes 2024. Daily dosing is not a convenience preference; it is the exposure pattern the favorable trials used.

The K2 pairing on the label is a hypothesis, not a finding. The argument is that vitamin D raises the supply of calcium and vitamin K-dependent carboxylation directs where it lands, so menaquinone should convert a calcium risk into a bone benefit. The carboxylation chemistry is real. A randomized trial in which vitamin D plus K2 beats vitamin D alone on fracture has not been run, and a combination product is not evidence that the combination was tested.

What would have to be true, and how you would know it was not

1. The number will move, and that prediction is nearly unfalsifiable. Vitamin D (25-hydroxy) rises with dose in a saturable curve; predict roughly 5 to 10 ng/mL per 1,000 IU per day at a low baseline and much less at a high one, with a plateau reached in about 8 to 12 weeks Heaney 2003. Retest at 12 weeks, not at 4. If the number does not move on 4,000 IU daily taken with fat, the question is absorption — fat malabsorption, bile acid sequestrants, or bariatric anatomy — and not dose.

2. Parathyroid hormone is the read-out that says whether the rise meant anything. The physiological purpose of repletion is to switch off compensatory parathyroid drive. Predict that PTH falls toward the lower half of the reference interval as vitamin D rises from a deficient baseline, and that it does not move at all in someone who starts above 30 ng/mL. A PTH that does not fall while 25-hydroxyvitamin D climbs is the signature of a person who did not need the supplement.

3. The prediction that cuts against the product. In an adult whose baseline is already above 30 ng/mL, predict no change in fracture rate, no change in cardiovascular events and no change in cancer incidence over five years, because that is what 25,871 people returned Manson 2019 LeBoff 2022. A trial in that population showing a fracture reduction would falsify this page. So would a high-dose trial showing a bone density gain rather than the dose-dependent loss that was reported Burt 2019.

4. The respiratory prediction is the one that survives, and it is conditional. Individual participant data across 25 trials showed protection against acute respiratory infection concentrated in people with baseline below 10 ng/mL and on daily or weekly rather than bolus dosing Martineau 2017. Predict a measurable reduction in infection episodes only in that subgroup, over one winter, and essentially none in a replete adult taking 5,000 IU.

5. Predict serum calcium stays flat, and treat it as the safety endpoint rather than the efficacy one. On 1,000 to 5,000 IU daily, predict no change in serum calcium. A rising calcium with a rising 25-hydroxyvitamin D is not a better result; it is the first sign of the toxicity described below, and it is measurable on a standard comprehensive metabolic panel months before anybody feels anything.

What nobody has tested yet

Nobody has run the trial this supplement actually needs. Every large trial enrolled without a status threshold, so the population that the biology predicts should respond — adults with 25-hydroxyvitamin D below 20 ng/mL — has never been randomized at scale to repletion against placebo with fracture as the endpoint. That study is entirely feasible and its absence is why the argument about vitamin D never ends Dawson-Hughes 2024.

Nobody knows whether free 25-hydroxyvitamin D is the better test. Roughly 85 to 90 percent of circulating 25-hydroxyvitamin D is bound to vitamin D binding protein, whose concentration varies with pregnancy, liver disease and common genetic variants. Free and bioavailable assays exist and have never been shown, prospectively, to predict an outcome that total 25-hydroxyvitamin D misses Bouillon 2022.

Nobody has settled whether the extraskeletal receptor does anything you can supplement into. Local CYP27B1 activity in immune cells is a documented paracrine system, and no trial has demonstrated that raising the serum substrate pool changes an immune outcome in a person who was not deficient. That is the gap between the mechanism papers and the outcome papers, stated as an experiment rather than as a disagreement.

And nobody has explained the bone density loss. A dose-dependent decline in volumetric bone mineral density at 4,000 and 10,000 IU daily was not predicted by anybody, has not been mechanistically resolved, and has not been replicated with fracture as the endpoint Burt 2019. Until it is, the upper end of the dosing range on this page rests on the absence of harm rather than on its demonstration.

Vitamin D — its own safety story, not its category's

The toxicity is hypercalcemia and it is a slow accumulation, not an overdose. Because 25-hydroxyvitamin D has a half-life of weeks and is stored in adipose tissue, the harmful exposure is months of a high dose rather than one large one. The presentation is the presentation of high calcium: thirst, polyuria, nausea, constipation, confusion, and nephrocalcinosis or stones at the end of it. It is detectable on a routine chemistry panel long before it is symptomatic.

The tolerable upper intake level for adults is 4,000 IU per day and the dose on many shelves is above it. The 5,000 IU capsule is widely sold and widely stacked on top of a multivitamin that already carries 1,000 to 2,000 IU, which is how people reach 8,000 IU without intending to. A page that recommends dosing to a target number owes the reader the arithmetic of everything else in the cabinet.

Three conditions turn an ordinary dose into a dangerous one, and all three do it by the same mechanism. Sarcoidosis, tuberculosis and some lymphomas contain granulomatous macrophages that express CYP27B1 without the parathyroid hormone feedback that restrains the kidney. In those people, substrate is converted to active hormone without a brake, and modest supplementation can produce frank hypercalcemia. Primary hyperparathyroidism belongs on the same list for the same arithmetic reason.

The interactions that matter are absorption interactions and one enzyme interaction. Orlistat, cholestyramine and colesevelam reduce uptake of a fat-soluble vitamin taken at the same time. Thiazide diuretics reduce urinary calcium excretion and therefore raise the calcium consequence of any given dose. Phenytoin, phenobarbital and rifampin induce CYP24A1 and accelerate catabolism, which lowers 25-hydroxyvitamin D at an unchanged intake. Digoxin deserves separate mention because hypercalcemia potentiates its cardiac effect.

What this page will not do is tell anyone their number. There is no consensus on an optimal 25-hydroxyvitamin D and the guideline bodies disagree by a factor of two on the threshold for sufficiency Bouillon 2022. Interpreting a result, deciding whether to treat and choosing a dose are clinical decisions. Nothing here is medical advice or diagnosis, and these statements have not been evaluated by the Food and Drug Administration.

Sources read for this page

• Manson JE, et al. Vitamin D Supplements and Prevention of Cancer and Cardiovascular Disease. The New England Journal of Medicine, 2019 · PMID 30415629
• LeBoff MS. Supplemental Vitamin D and Incident Fractures in Midlife and Older Adults. N Engl J Med 2022 · PMID 35939577
• Bouillon R. The health effects of vitamin D supplementation: evidence from human studies. Nat Rev Endocrinol 2022 · PMID 34815552
• Dawson-Hughes B. Effect of vitamin D on risk of falls and fractures - The contribution of recent mega-trials. Metabol Open 2024 · PMID 39100895
• Limonte CP. Effects of Vitamin D3 Supplementation on Cardiovascular and Cancer Outcomes by eGFR in VITAL. Kidney360 2022 · PMID 36591342
• Zhao JG, et al. Association Between Calcium or Vitamin D Supplementation and Fracture Incidence in Community-Dwelling Older Adults: A Systematic Review and Meta-analysis. JAMA, 2017 · PMID 29279934
• Burt LA, et al. Effect of High-Dose Vitamin D Supplementation on Volumetric Bone Density and Bone Strength: A Randomized Clinical Trial. JAMA, 2019 · PMID 31454046
• Heaney RP, et al. Human serum 25-hydroxycholecalciferol response to extended oral dosing with cholecalciferol. The American Journal of Clinical Nutrition, 2003 · PMID 12499343
• Martineau AR, et al. Vitamin D supplementation to prevent acute respiratory tract infections: systematic review and meta-analysis of individual participant data. BMJ, 2017 · PMID 28202713

How you would know if it worked

Vitamin D — safety & side effects

Bloodwork to run alongside Vitamin D

Baseline first, then again at 8–12 weeks.

The Full Micronutrient Screen panel covers these in one order — 11 markers, $363.60 with the discount applied.

Check results you already have → · All 103 markers A–Z

Vitamin D — frequently asked questions

Vitamin D inside a finished plan

One arm of 9 Protocol Blueprints, free to read in full.

What Vitamin D is used for

Vitamin D appears under 11 goals in the goal router.

Related Vitamins supplements

← Browse the full Supplement Vault

↑ Back to on this page