Vitamin D dosage and evidence

Overview

Mechanism of action

Cholecalciferol is hydroxylated in the liver to 25-hydroxyvitamin D, which is what a blood test measures, then hydroxylated again in the kidney to 1,25-dihydroxyvitamin D, the active hormone. That binds the vitamin D receptor, a nuclear receptor expressed in most tissues, and acts as a transcription factor. Its established role is calcium and phosphate homeostasis: it raises intestinal calcium absorption and, with parathyroid hormone, regulates bone remodelling. The vitamin D receptor's presence in immune, muscle and vascular tissue is the basis of the wider claims, and is a mechanistic observation rather than a demonstrated clinical effect.

Human evidence

Among the largest randomised evidence bases for any supplement, and predominantly negative for the outcomes people take it for. Tens of thousands of participants across multiple trials on multiple continents.

• VITAL, 25,871 participants, 2,000 IU daily, median 5.3 years: no reduction in invasive cancer and no reduction in major cardiovascular events. Both primary endpoints missed.
• D-Health, monthly high-dose in older adults: all-cause mortality hazard ratio 1.04.
• The widely quoted 22% reduction in autoimmune disease from VITAL was an ancillary result at p = 0.05 that weakened to a confidence interval crossing 1.0 once two further years of case adjudication were applied, with no persistence after stopping.
• 8,851 Mongolian schoolchildren who were 95.6% deficient: supplementation tripled blood levels and the primary endpoint was still null.
• A single annual 500,000 IU dose increased falls by 15% and fractures by 26% against placebo in 2,256 older women.
• Where it is genuinely established: preventing and treating rickets and osteomalacia, and calcium plus vitamin D together for fracture risk in institutionalised older adults, which is a different population and a different intervention than a supplement bottle bought by a healthy adult.

What this does not tell you: Both large trials enrolled mostly replete populations, so they answer the question of whether to top up someone who is not short, and they answer it no. They do not settle whether repletion helps a genuinely deficient person, which is the question most readers actually have. The honest state of that question is that the one large trial run specifically in a deficient population was also null. Nothing here speaks to pregnancy, malabsorption, kidney disease or documented deficiency syndromes, where dosing is a clinical decision.

Reading the research record

The evidence, charted

Key studies & citations

• Human2019
  Vitamin D supplements and prevention of cancer and cardiovascular disease

  Vitamin D was tested properly, at scale, and it did not prevent cancer or heart disease. Nearly 26,000 adults were split by chance to take a daily vitamin D pill or a dummy pill. They were followed for more than five years. Neither of the two questions the trial was built to answer came out ahead. Most of these people already had decent vitamin D levels. Only about one in eight started out low. So this does not settle what happens when someone is truly short of it.

  VITAL. 25,871 participants randomised to 2,000 IU daily or placebo, median 5.3 years. Neither primary endpoint was met: no reduction in invasive cancer, no reduction in major cardiovascular events. Mean baseline 25-hydroxyvitamin D was 30.8 ng/mL and only 12.7% of participants were below 20.

  PMID 30415629 New England Journal of Medicine
• Human2022
  The D-Health trial: a randomised controlled trial of the effect of vitamin D on mortality

  A second large test asked whether vitamin D helps people live longer, and it did not. Older Australian adults took a big dose once a month, or a dummy pill, for years. The same share died in each group, of any cause. If anything the treated group fared very slightly worse, by an amount too small to separate from chance. As in the trial above, the people getting the dummy already had reasonably good vitamin D levels. Nobody has tested a population that starts out genuinely deficient.

  Monthly high-dose vitamin D in older Australian adults. All-cause mortality hazard ratio 1.04. No benefit, again in a population whose placebo group averaged about 31 ng/mL.

  PMID 35026158 Lancet Diabetes and Endocrinology
• Human2010
  Annual high-dose oral vitamin D and falls and fractures in older women: a randomized controlled trial

  A single huge yearly dose of vitamin D made things worse, not better. Over two thousand women aged 70 and above were given one very large dose each year, or a dummy. The women on vitamin D fell more often and broke more bones than the women on the dummy. This is the clearest sign that more vitamin D is not automatically safer. It was one enormous annual dose, which is not how most people take it. It does not tell you what a normal daily amount does.

  2,256 women aged 70 and over given a single annual 500,000 IU dose. Falls incidence rate ratio 1.15 and fractures 1.26. The supplement group fell and broke more, not less. Evidence that more is not safer.

  PMID 20460620 JAMA
• Review2014
  Effect of vitamin D on mortality in adults: a systematic review and individual participant data meta-analysis

  The pooled individual-participant analysis behind the observational case. People in the lowest quartile of 25-hydroxyvitamin D had substantially higher mortality than the highest. Observational exposure, so it cannot establish that supplementing changes it.

  PMID 24938302 BMJ
• Human2020
  Vitamin D supplements for prevention of tuberculosis infection and disease

  This is the result that answers the usual excuse, and it is a hard one. Nearly nine thousand children were studied in a place where almost all of them were short of vitamin D. Supplements tripled their blood levels, so the pills clearly worked on the blood test. The thing the trial was built to answer still came out empty. Fixing a real shortage did not protect them from tuberculosis. This was the population where a benefit should have been easiest to see, and it was not there.

  The most uncomfortable result in this literature and the one that answers the usual defence. 8,851 children, 95.6% vitamin D deficient at baseline. Supplementation tripled their blood levels. The primary endpoint was null. Correcting real deficiency did not deliver the expected benefit in the population where it should have been easiest to show.

  PMID 32706534 New England Journal of Medicine
• Human2021
  Effect of marine omega-3 fatty acid and vitamin D supplementation on incident atrial fibrillation

  Vitamin D did not cause an irregular heartbeat, which is a rare clean safety answer here. This analysis was planned in advance inside the big vitamin D trial, so it is not a fishing expedition after the fact. Atrial fibrillation, a fast and irregular heart rhythm, happened no more often on vitamin D than on the dummy pill. That is worth knowing on its own. It says nothing about whether vitamin D helps with anything. It only says this particular harm did not show up.

  The prespecified atrial fibrillation analysis within VITAL. Vitamin D did not increase atrial fibrillation. Included here because it is one of the few clean safety answers in the file.

  PMID 33724323 JAMA

What users report

Self-reported experiences, not evidence. Nothing below was measured under controlled conditions, and reports like these cannot separate a real effect from placebo, from the training or diet change that accompanied it, or from what the product actually contained. They are here because knowing what people describe, including what goes wrong, is worth reading alongside the studies.

• Amounts described run from about 2,000 to 10,000 IU daily, with prescription weekly amounts around 50,000 IU also mentioned, and a repeated observation that people need very different amounts to reach the same blood level. Some describe supplementing only in winter.
• The blood level target is the central and unresolved disagreement. Some people aim at around 50 ng/mL or higher, others argue the mortality data favours lower, and a persistent minority argue that a low reading is a marker of poor health rather than something worth treating in itself.
• Pairing with K2, magnesium and sometimes vitamin A is close to universal advice in these threads, with the stated worry being calcium ending up in soft tissue rather than bone.
• Constipation is the side effect described in most detail, by people who report it persisting across many brands, amounts and timings and not responding to fibre, water or magnesium, while others at the same amounts report nothing at all. Digestive upset at higher amounts, hypercalcaemia at very high blood levels, and questions about insomnia after injections also appear.
• Reports of no subjective change at any amount are common and are made by people who kept supplementing anyway on the basis of the blood number. Several note that they changed other things at the same time and cannot separate them.
• Scepticism about the literature is loud and comes from both directions. One group argues most trials fail because they never measure the blood level achieved or account for cofactors, and another uses the newer large trials showing no mortality benefit to argue the whole practice is misplaced.

Sources: Vitamin D threads on ExcelMale covering high dose use, reported side effects and disagreement over target blood levels, found by site-restricted search and read directly, September 2026. Nothing above is attributed to any specific post and no post is reproduced.