Vitamin D Supplements: Most Useful for People Who Are Deficient

The most-hyped vitamin

Vitamin D earned attention for its roles in bone, immune, and overall health, and a large share of people have low levels - especially in winter, at higher latitudes, with darker skin, or with limited sun exposure. But more is not automatically better.

In the NHANES 2005–2006 survey of 4,495 US adults, 41.6% had a blood level at or below 20 ng/mL (50 nmol/L), the usual sufficiency cut-off; among Black adults the figure was 82.1% and among Hispanic adults 69.2% (Forrest & Stuhldreher 2011). The Institute of Medicine, using the same threshold, concluded that the prevalence of inadequacy in North America has been overestimated and called for laboratory reference ranges to be reassessed (Ross et al. 2011).

How vitamin D works, in plain words

Skin makes vitamin D3 from a cholesterol relative when ultraviolet B light hits it; a smaller amount comes from oily fish, egg yolks and fortified foods. The liver converts it to 25-hydroxyvitamin D - the storage form and the thing a blood test measures - and the kidneys turn a little of that into the active hormone as needed. The hormone’s main job is to raise calcium absorption in the gut. Without enough of it, calcium absorption falls, the body borrows calcium from bone, and bone softens - which is why severe deficiency causes rickets in children and bone pain and weakness in adults. Immune cells also carry vitamin D receptors, which is the basis for the infection research below.

The key finding: deficiency vs. sufficiency

The single most important lesson from large trials: vitamin D mainly helps people who are actually deficient. In already-sufficient adults, supplementation does not reduce fractures or falls (Bouillon et al. 2022). But correcting genuine deficiency modestly supports bone health, and combined vitamin D plus calcium is linked to fewer hip fractures in at-risk groups (Bouillon et al. 2023).

The trials that settled this are large:

• VITAL (Manson et al. 2019) randomized 25,871 US adults (men 50+, women 55+) to 2,000 IU/day of vitamin D3 or placebo for a median of 5.3 years. Invasive cancer: hazard ratio 0.96. Major cardiovascular events: 0.97. Neither differed from placebo. Deaths from cancer were numerically lower (hazard ratio 0.83) but not significantly so.
• VITAL’s fracture analysis (LeBoff et al. 2022) followed the same participants: 1,991 fractures in 1,551 people, with hazard ratios of 0.98 for total fractures and 1.01 for hip fractures. Crucially, the result did not change by baseline vitamin D level - but participants were not recruited for deficiency, low bone mass or osteoporosis, so the trial says little about people who are.
• D-Health (Neale et al. 2022) gave 21,315 Australians aged 60 and over a monthly 60,000 IU dose or placebo for five years. All-cause mortality: hazard ratio 1.04. Cardiovascular mortality: 0.96. Cancer mortality: 1.15, not significant, but an exploratory analysis excluding the first two years pointed toward more cancer deaths in the vitamin D group. Blood levels in the placebo group averaged 77 nmol/L (about 31 ng/mL) - most participants were already replete.

Bouillon’s 2022 review puts VITAL together with two other large trials, ViDA and D2d, for a combined 30,000-plus participants: in adults with baseline levels above 50 nmol/L, supplementation did not prevent cancer, cardiovascular events, falls or progression to type 2 diabetes. More than 60 Mendelian randomization studies, which use genetic variants to test whether lifelong lower vitamin D causes disease, have mostly found null effects.

Immune health

The best non-bone evidence is immune: a large pooled analysis of 25 trials (Martineau et al. 2017, BMJ) found supplementation modestly reduced acute respiratory infections, with the biggest benefit in those who started out deficient. Again, the benefit tracks with correcting a shortfall, not with topping up people who are already fine.

The numbers: across 11,321 participants the adjusted odds of an infection fell by 12%. In people taking daily or weekly doses without large bolus doses, the reduction was 19%; within that daily-or-weekly group, those who started below 25 nmol/L (10 ng/mL) saw a 70% reduction, versus 25% in those who started higher. Bolus dosing showed no benefit at all.

An updated meta-analysis of 46 trials and 75,541 participants (Jolliffe et al. 2021) found the overall effect shrank as more trials were added: 61.3% of the supplemented group had at least one infection versus 62.3% on placebo - an odds ratio of 0.92 and an absolute difference of about one percentage point. Daily dosing (odds ratio 0.78) and daily doses of 400–1,000 IU (0.70) still looked best, but the earlier finding that deficient people benefit most was no longer statistically significant in this larger dataset.

Practical approach

• Test if unsure (a simple 25-OH vitamin D blood test) rather than guessing.
• Typical maintenance doses are 1,000 to 2,000 IU/day; correcting deficiency may need more, guided by a clinician.
• More is not better - very high doses can cause toxicity (dangerously high calcium), so avoid mega-dosing.

What the guidelines and trials actually used:

Two points about testing. The IOM set the sufficiency threshold at 20 ng/mL (50 nmol/L) and found higher values were not consistently associated with more benefit, with U-shaped risks at both extremes for some outcomes (Ross et al. 2011). And the Endocrine Society’s 2024 guideline suggests against routine 25(OH)D testing in healthy adults without a specific reason, and against empiric supplementation above the RDA in healthy adults under 75 (Demay et al. 2024). So “test if unsure” is best read as: test when there is an established clinical reason - malabsorption, osteoporosis, kidney disease, or a deficiency already being treated - rather than as an annual ritual; the panel found no trial evidence to support routine screening even in people with obesity or darker skin. The same guideline does suggest supplementation for children, adults 75 and over, pregnancy and high-risk prediabetes, and prefers daily dosing over intermittent large doses.

Vitamin D is often paired with K2 on the theory that the two direct calcium to bone; the mechanism is stronger than the trial data, as covered in vitamin K2 and D3. VITAL tested fish oil in the same people, with a similarly null primary result; see the omega-3 evidence.

Safety and who should talk to a clinician

The tolerable upper intake level set by the Institute of Medicine for adults is 4,000 IU/day, a limit derived from the risk of hypercalcemia (IOM 2011); higher intakes are used under supervision to correct deficiency but are not a self-care dose. Toxicity is rare but real: sustained very high intake raises blood calcium, which can cause nausea, weakness, frequent urination, kidney stones and, in severe cases, kidney damage and heart-rhythm problems.

Large intermittent doses carry their own risk. In a trial of 2,256 community-dwelling women aged 70 and over who were at high risk of fracture, a single annual dose of 500,000 IU increased falls by 15% and fractures by 26% compared with placebo, with the excess falls concentrated in the three months after each dose (Sanders et al. 2010). That result, together with the bolus-dose findings in the infection trials, is why guidelines now favour daily dosing.

Talk to a clinician before supplementing above the RDA, or before testing, if you:

• have kidney disease, kidney stones or a history of high blood calcium
• have sarcoidosis or another granulomatous condition, which can over-activate vitamin D
• take thiazide diuretics, digoxin or other drugs that interact with calcium
• are pregnant, or are managing osteoporosis, where the dose should be part of a wider plan
• have been diagnosed as deficient - correction doses and re-testing intervals are a clinical decision

See also how to read a supplement study.

Bottom line

Two trials of about 47,000 people who were mostly already replete (VITAL and D-Health) found no effect on cancer, cardiovascular events, fractures or mortality; the clearest benefits are for bone in deficient people and a small reduction in respiratory infections, largest with modest daily doses. Vitamin D is worth correcting if you are low - which many people are - but it is not a cure-all to megadose when your levels are already fine. Test, correct a real deficiency, and keep doses sensible.