What vitamin D actually is
Vitamin D is less a vitamin than a hormone your skin makes from sunlight. Its classic, undisputed job is helping your gut absorb calcium and keeping bones and muscles working — but receptors for it sit in tissues all over the body, which is why it got swept up in claims about almost everything.
When UVB light hits your skin, it kicks off the production of vitamin D3 (cholecalciferol). Your liver and kidneys then convert it into the active hormone. The reason deficiency is common: modern indoor life, sunscreen, darker skin (more melanin means less D per minute of sun), northern latitudes where winter sun is too weak, and age (older skin makes less). Roughly a quarter of people have low levels, and around a billion worldwide are estimated to be insufficient.
The number that matters is your blood 25-hydroxyvitamin D [25(OH)D] — a simple test. Most of this page comes back to one idea: vitamin D's value depends almost entirely on whether you're actually low. Fixing a deficiency is genuinely worthwhile; adding more on top of a healthy level mostly isn't.
What the big trials actually showed
For years, observational studies tied low vitamin D to just about every disease, and supplements were marketed accordingly. To settle it, researchers ran VITAL — a landmark randomized trial of over 25,000 adults taking 2,000 IU of D3 daily (or placebo) for about five years, looking at cancer and heart disease. Finland's FIND trial and others asked similar questions.
The headline result deflated the hype: in a general population, vitamin D supplements did not prevent cancer, did not prevent cardiovascular disease, and did not reduce fractures. A companion analysis of the Women's Health Initiative over 22 years was similarly sobering — and even hinted at a small increase in cardiovascular-death risk alongside a small decrease in cancer-death risk. So much for the cure-all.
There were a couple of genuine bright spots in VITAL's deeper analyses: a reduction in autoimmune disease over the trial, and signals (in subgroups) for lower cancer mortality in people who were normal-weight or took it longer. These are promising threads — not settled conclusions — and we tier them honestly below.
What vitamin D does — tier by tier
Sorted by how strong the evidence actually is, not by how loud the marketing was.
The bedrock, and not in doubt. Severe deficiency causes rickets in children and osteomalacia (soft bones) in adults; correcting low levels supports calcium absorption, bone strength, and muscle function. In deficient older adults, vitamin D (usually with calcium) reduces fracture and fall risk — the adherent women in the WHI trial had ~29% fewer hip fractures. This is the use vitamin D earns its keep on.
One of VITAL's most interesting findings: the supplement group had a lower rate of new autoimmune diseases (like rheumatoid arthritis and psoriasis) over five years. Vitamin D is a genuine immune modulator — it nudges the immune system away from over-reactive states. Promising and from a real RCT, but it needs replication before it's a recommendation.
Meta-analyses suggest modest protection against acute respiratory infections — but the benefit concentrates in people who were deficient to start with, and is small or absent in the already-replete. So it's "fix a deficiency, maybe get fewer colds," not "megadose to armor your immune system." Daily low doses beat big intermittent bolus doses here.
This is where the hype lived and where the RCTs disappointed. VITAL found no reduction in cancer incidence or cardiovascular events overall. There are intriguing subgroup signals (cancer mortality in normal-weight people; the WHI's long-term cancer-death dip) but also that small CVD-death increase in WHI. Net: not a reason to supplement for prevention in well-nourished people.
Low vitamin D tracks with low mood in observational data, but VITAL's depression sub-study found no benefit from supplementing for preventing or treating depression in the general population. Correlation, most likely — not a treatment.
Who actually benefits
Vitamin D is the rare supplement where the right answer is genuinely person-specific. The people who benefit are the ones who are (or are likely to be) low:
If none of these apply, you get reasonable sun, and you eat oily fish or fortified foods, you may well be fine without a supplement — which is exactly why several guidelines now advise against blanket supplementation of healthy younger adults. The honest move isn't "everyone should take it" or "no one should" — it's test, then decide.
Reading your level, and dosing to it
Vitamin D dosing should follow your blood test, not a one-size number. Labs report 25(OH)D in ng/mL (US) or nmol/L (most other countries; multiply ng/mL by 2.5). The thresholds themselves are debated (see below), but a widely used framework:
Vitamin D is fat-soluble, so take it with a meal containing some fat — it absorbs noticeably better than on an empty stomach. Daily dosing is gentler and, for infections, more effective than large weekly or monthly "bolus" doses.
Official baselines for the general population are more modest: 600 IU/day for most adults, 800 IU over age 70 (these are intakes to prevent deficiency, not therapeutic doses). The bigger numbers above are for correcting a measured low level.
D3 vs D2 — and the K2 question
Choose D3 (cholecalciferol). It's the form your skin makes, and head-to-head it raises and maintains blood 25(OH)D far more effectively than D2 (ergocalciferol) — a 2024 meta-analysis put D3 at roughly 87% more efficient at lifting levels. D2 is mostly seen in some prescription megadoses and plant/fortified products; vegans can get D3 from lichen-derived sources.
Do you need vitamin K2 with it?
The theory: vitamin D raises calcium absorption, and K2 helps direct that calcium into bone rather than arteries. It's mechanistically plausible and popular, but not firmly proven in hard outcome trials. A reasonable, low-risk take: at modest doses (≤2,000 IU) it's optional; at sustained higher doses (4,000+ IU) some clinicians add K2 (MK-7). It's a "reasonable maybe," not an established requirement — and anyone on blood thinners must talk to their doctor first, since K2 interacts with warfarin.
The myths, busted
The big trials don't support blanket high-dosing of healthy, replete people — it didn't prevent cancer, heart disease, or fractures in that group. Benefit tracks with being low. Test first; dose to need.
This was the central hope, and the RCTs largely said no for overall prevention. There are subtle subgroup signals worth studying, but as a reason to supplement, it's not established.
True in summer at lower latitudes with bare skin — but useless in northern winters, through windows, under sunscreen, or with darker skin in weak sun. And the sun's UV is also the main cause of skin cancer, so "more sun" isn't a free strategy.
Up to a healthy level, correcting low D matters. Beyond it, more does nothing useful — and far beyond it is harmful. Vitamin D is fat-soluble and accumulates, so unlike vitamin C you can genuinely overdo it.
The live debate: how much, and for whom
This is a topic where serious experts genuinely disagree, and it would be dishonest to pretend it's settled. The core dispute is about thresholds and who to treat — and it largely comes down to how much weight you put on randomized trials versus observational data.
🔬The conservative case
Bodies like the Endocrine Society (2024) lean on the null RCTs: don't screen or empirically supplement healthy adults under 50; a level around 20 ng/mL is adequate for most; the extra-skeletal benefits aren't proven enough to act on. Treat deficiency, don't chase high targets.
☀️The proactive case
Other researchers argue the big trials were designed to fail — they enrolled people who were already replete and used modest doses — so they couldn't show benefits that may exist at higher targets (40–60 ng/mL). They weight the large observational signals more heavily.
Where they agree — and what's safe to act on — is the useful part: genuine deficiency is worth correcting; the test is cheap; D3 is the right form; and megadosing a replete person is pointless. The disagreement is mostly about the middle ground, which is exactly where talking to your own clinician (and using your own blood level) beats any blanket rule.
The honest limits & safety
Getting it from sun and food
Sensible sun is the natural source — short, regular exposure of bare skin in months when the sun is strong enough, balanced against skin-cancer risk (never burning). Food contributes modestly, which is why fortified products exist. The richest dietary sources:
| Food | Vitamin D (approx.) |
|---|---|
| Salmon (wild), 100 g | ~600–1,000 IU |
| Sardines / mackerel, 100 g | ~200–600 IU |
| UV-exposed mushrooms, 100 g | ~200–400 IU |
| Egg yolk (each) | ~40 IU |
| Fortified milk / plant milk, 1 cup | ~100 IU |
| Cod liver oil, 1 tsp | ~450 IU |
As the table shows, hitting a corrective dose through food alone is hard — which is the honest argument for a modest supplement if you're low or at risk. For where this fits a whole-diet approach, see the Food First tool and the supplement guide.
Frequently asked
Should I just take it, or get tested first?
D3 or D2?
What time of day, and with food?
Do I need vitamin K2 with it?
Can I get enough from the sun?
Is it safe long-term?
Sources & further reading
This deep dive is educational and not medical advice. Vitamin D supports normal bone, muscle, and immune function; correcting a deficiency is a medical matter best guided by a blood test and a clinician. Talk to a qualified professional before starting higher-dose supplementation — especially if you're pregnant, have kidney disease, sarcoidosis, hyperparathyroidism, or take regular medication.