Correcting a real deficiency is worthwhile. But the largest randomized trial deflated the idea that routine vitamin D pills meaningfully prevent cancer, heart disease, or fractures in already-sufficient adults. Fix a shortfall; don't expect a miracle.
The one-paragraph version
Vitamin D matters for bone health and is genuinely important if you're deficient — common in people with little sun exposure, darker skin, older age, obesity, or malabsorption S1. But the VITAL trial, which randomized ~25,000 adults to 2000 IU/day for a median 5.3 years, did not reduce its primary cancer or cardiovascular endpoints S2. Some subgroup and secondary signals are intriguing (cancer mortality, diabetes progression in people with prediabetes) S2S3, but the population-wide "take it, it prevents everything" story didn't survive contact with the data. A modest daily dose is cheap and safe; mega-doses are not better and can be harmful.
What it is and how it works
Vitamin D is a fat-soluble hormone-precursor your skin makes from sunlight and you also get from a few foods and supplements. It regulates calcium and phosphate — hence its central role in bone — and has receptors across many tissues, which is the basis for the broader (and largely unconfirmed) disease-prevention hopes S1.
What the evidence actually supports
Bone health and deficiency correction — the solid core. Vitamin D (with adequate calcium) supports bone mineralization; severe deficiency causes rickets and osteomalacia S1. Treating a documented deficiency is standard, evidence-based care.
Cancer and cardiovascular prevention — largely negative at the population level. VITAL's primary endpoints were null: 2000 IU/day did not lower total cancer incidence or major cardiovascular events versus placebo over ~5 years S2. A secondary signal suggested reduced cancer mortality when early follow-up was excluded, and possible benefit in people with normal BMI and in African-American participants — hypothesis-generating, not practice-changing S2.
Type 2 diabetes — a modest subgroup signal. In a VITAL ancillary analysis, vitamin D did not prevent diabetes across the whole group, though related trial evidence suggests a small benefit in people with prediabetes S3. Interesting, not a general recommendation.
Aging/telomeres — early and preliminary. A 2025 VITAL sub-study reported vitamin D3 slowed leukocyte telomere attrition over 4 years S4. Mechanistically intriguing; nowhere near a reason to promise anti-aging effects.
Who actually benefits
The people with the most to gain are those who are actually low: limited sun exposure, darker skin, older adults, higher body weight, malabsorptive conditions, and northern winters S1. For a sun-exposed, healthy adult already in the sufficient range, routine supplementation buys little based on the trial data S2.
Dosing (standard, well-established)
The RDA is 600 IU/day for adults up to 70 and 800 IU/day for those over 70 S1. Many clinicians use 1000–2000 IU/day for maintenance or mild insufficiency — the dose VITAL tested for safety and (largely null) efficacy was 2000 IU/day S1S2. Because it's fat-soluble, take it with a meal containing fat. Deficiency treatment sometimes uses higher short-term doses, but that is a clinician's call with monitoring — not a DIY move.
Safety
At typical doses vitamin D is very safe. The tolerable upper limit for adults is 4000 IU/day S1. Toxicity is rare and comes from chronic very-high dosing (often ≥10,000 IU/day for long periods), causing hypercalcemia — nausea, kidney stones, and in extreme cases kidney damage S1. This is the clearest "more is not better" nutrient in the common-supplement world: the megadose trend carries real downside and no proven extra benefit. People with sarcoidosis, hyperparathyroidism, or kidney disease need medical guidance S1.
The marketing myths
- "Everyone is deficient and should megadose." Insufficiency is common, but the fix is a modest dose, not 10,000 IU S1.
- "Vitamin D prevents cancer and heart disease." The largest trial says no for the general population S2.
- "Higher blood levels are always better." Above sufficiency, benefits plateau and toxicity risk rises S1.
- "You must get tested and optimized." Reasonable for at-risk groups; routine screening of healthy, low-risk adults is not well supported.
Sources
Every reference below is a primary source cited in this guide.