Vitamin D2 or D3: what the evidence shows after a review found D2 supplements lower the body's D3

A bottle labeled vitamin D can hold either of two substances. Most buyers never check which, and for decades the two were treated as the same thing. The evidence that they are not has been building, and a meta-analysis publicized this week adds an odd twist: taking one form appears to drive down the body’s supply of the other.

The form in question is vitamin D2. In trials, people who took it ended up with less vitamin D3, the form the body makes for itself, than people who took nothing. The analysis was published in Nutrition Reviews by a team led from the University of Surrey.

Whether this matters for health is a separate question, and the answer so far is that nobody has directly tested it. This explainer sets out what is known about the two forms.

What vitamin D2 and D3 are

Both are forms of vitamin D, which the body needs to absorb calcium and keep bones strong. They differ in origin.

Vitamin D2, also called ergocalciferol, is commonly produced from fungi or yeast exposed to ultraviolet light. Vitamin D3, or cholecalciferol, is the form produced in human skin after exposure to ultraviolet B radiation from sunlight. That difference in origin has made D2 the default for vegans, because vitamin D3 in supplements has often come from lanolin obtained from sheep’s wool.

Once swallowed, each is processed by the liver. The liver converts it into 25-hydroxyvitamin D, the form doctors commonly measure in blood tests. D2 becomes 25-hydroxyvitamin D2 and D3 becomes 25-hydroxyvitamin D3. A standard blood test adds the two together and reports the total.

Results are usually given in nanomoles per liter, written nmol/L. As a point of reference, the reviewers behind a large analysis by Cochrane, an international group that reviews medical evidence, treated 20 ng/mL, which is 50 nmol/L, as the line for adequacy.

What the review of vitamin D2 supplements found

Hints that D2 might lower D3 had appeared in individual trials. The Surrey-led team wrote that it is suspected from randomized controlled trial data that vitamin D2 supplementation increases the metabolic clearance of the D3 marker, but this effect has yet to be quantified. Metabolic clearance means the rate at which the body breaks a substance down and removes it.

They set out to put a number on it. PUBMED was searched for articles published from January 1, 1975, to February 1, 2023, PubMed being the main index of medical journals. Of the 202 articles retrieved, 20 were included in this review, and of those, 11 were suitable for meta-analysis, the statistical pooling of results. All were randomized trials in which some people received D2 and others did not, and in which the D3 marker was measured separately.

The pooled result was clear in direction. There was a reduction in serum 25(OH)D3 after vitamin D2 supplementation compared with control, serum 25(OH)D3 being the D3 marker in blood. At the end of the trials, the D3 marker in people given D2 was lower by 17.99 nmol/L, with a plausible range of 10.12 to 25.86. Counted a second way, as change from each person’s starting point, the difference was smaller, at 9.25 nmol/L, with a range of 4.10 to 14.40. The university’s account adds a caution: the result does not mean vitamin D2 simply removes vitamin D from the body or that taking it is necessarily harmful.

Lead author Emily Brown described the pooled result as a discovery, though hints of the effect had appeared in individual trials. “We discovered that vitamin D2 supplements can actually decrease levels of vitamin D3 in the body, which is a previously unknown effect of taking these supplements,” Brown said.

The result needs one piece of context to avoid misreading. The fall in D3 is not a fall in vitamin D overall. Vitamin D2 itself raises 25-hydroxyvitamin D2 and can contribute to total vitamin D status, the university’s account notes. A person taking D2 has more vitamin D in total, of which less is D3.

Why vitamin D2 might lower vitamin D3

The mechanism is not known. The authors offer a hypothesis. A regulatory mechanism that increases the disposal rate of 25(OH)D after an increase in vitamin D concentrations could explain these results, they write.

In plain terms, the body appears to keep its vitamin D within bounds. When supplies rise, it speeds up the machinery that breaks vitamin D down. That machinery may not distinguish between the two forms, so a flood of D2 could speed the removal of D3 along with it. An inverse relationship between vitamin D2 and D3 concentrations has been proposed in the literature.

If that is right, the effect should run both ways, and the release says earlier experiments suggest it may, with D3 supplements sometimes lowering D2.

Which form raises vitamin D levels more

On this question the evidence is larger and has pointed the same way for more than a decade.

In 2012 the same university published the first pooling of head-to-head trials. Its conclusion was that vitamin D3 is more efficacious at raising serum 25(OH)D concentrations than is vitamin D2. Serum 25(OH)D is the standard blood measure of vitamin D, short for 25-hydroxyvitamin D. The difference depended on how the vitamin was given. It was large when given as a bolus dose, a single big dose taken weekly or monthly, but the effect was lost with daily supplementation.

A later analysis looked harder at daily dosing, which is how most people take supplements. Overall, the results based on 20 comparative studies showed D3 to be superior to D2 in raising total 25(OH)D concentrations. Among twelve comparisons of daily doses measured by the most accurate laboratory method, the rise in total vitamin D was 10.39 nmol/l (40%) lower for the D2 group compared to D3 group.

That analysis also turned up a qualification. BMI appeared to be the strongest response modifier, meaning that body weight changed the result more than anything else examined. In people with a body mass index above 25, the overweight range, D2 and D3 performed the same.

It also reported something that fits the new review. D2 and D3 had a similar positive impact on their corresponding 25(OH)D hydroxylated forms. Each form was equally good at raising its own marker. The shortfall in the total with D2 would then come from somewhere else, and a drop in D3 is the obvious candidate.

Do vitamin D2 and D3 differ in health outcomes?

Blood levels are a stand-in. What people want from a supplement is fewer fractures, infections or early deaths. Here the evidence is thinner, and none of it compares the two forms directly.

The nearest thing comes from trials that tested one form or the other against placebo and counted deaths. A team at the University of Cambridge pooled 22 such trials. Deaths from any cause were 11% lower with D3 than with placebo, with a plausible range of 1% to 20%. With D2 they were 4% higher, a difference well within the range of chance. The effects observed for vitamin D3 supplementation remained unchanged when grouped by various characteristics, such as dose and length of trial.

A Cochrane review reached the same split by a different route. When different forms of vitamin D were assessed in separate analyses, only vitamin D3 decreased mortality. The size of the benefit was modest, corresponding to 150 people treated over five years to prevent one additional death. A hint that D2 might even raise the death rate appeared in the weaker trials, and the reviewers judged that this finding could be due to random errors.

Both analyses come with a large caveat about who was studied. In the Cochrane review, most trials included women older than 70 years. Vitamin D was administered for a weighted mean of 4.4 years. The D2 trials and the D3 trials were run in different people, in different countries, at different doses. A difference between them is suggestive. It is not a comparison.

One further line of work comes from Surrey itself. In an earlier study the team measured which genes were switched on in blood cells after each form, and found differences involving the immune system. The release includes its own warning about that result: such changes are not to be read as proof that taking D3 will prevent a particular infection.

What is not known about vitamin D2 and D3

The central gap is easy to state. What remains uncertain is whether that reduction produces meaningful differences in health outcomes, the university’s account says of the fall in D3. The authors frame the gap more narrowly. Further research is needed to establish whether vitamins D2 and D3 elicit different changes in overall vitamin D metabolism, they write.

No head-to-head trial has measured anything that patients feel. The comparisons that exist measure blood chemistry.

It is not known whether a lower D3 marker matters if the total is adequate. If the body treats the two forms as equivalent once they are activated, the split between them may be of no consequence.

Dose and duration are open questions. The pooled figure averages across trials that used different doses for different lengths of time, and it does not show whether the fall in D3 grows with dose or levels off.

Much of the evidence comes from one research group. The 2012 analysis, the gene study and the new review were all led from the University of Surrey, and one of its senior researchers is also an author of the 2024 analysis.

Brown’s own reading favors D3. “This study suggests that subject to personal considerations, vitamin D3 supplements may be more beneficial for most individuals over vitamin D2,” Brown said. Co-author Cathie Martin drew a conclusion for people who avoid animal products. “This meta-analysis highlights the importance of ensuring plant-based vitamin D3 is accessible in the UK,” Martin said. Animal-free D3 made from sources such as lichen is also available, the release notes.

Whether a person needs a vitamin D supplement at all, and in which form, depends on their diet, sun exposure and health, which is assessed by their doctor.

Vitamin D3 raises blood levels of the vitamin more than D2 does, and D2 lowers the body’s own D3 marker by 9 to 18 nmol/L depending on how it is counted, while the trials that could show whether either difference changes fractures, infections or deaths have not been done.