Taking vitamin D2 lowered blood levels of D3, and researchers point most people to D3

A meta-analysis of randomized trials found that people taking vitamin D2 ended up with blood levels of 25-hydroxyvitamin D3 roughly 9 to 18 nmol/L lower than people who took no D2. The work comes from a team led by Emily Brown, a PhD research fellow at the University of Surrey, and it has fed a push to make D3 the default for most people who take a supplement.

The finding is easy to over-read, because a lower reading for one form of the vitamin is not the same as a worse health result. What the numbers do and do not show is the more useful part of the story.

Ergocalciferol and cholecalciferol: two molecules, two sources

Vitamin D comes in two supplement forms. D2, or ergocalciferol, is typically made from fungi or yeast exposed to ultraviolet light. D3, or cholecalciferol, is the chemical match for what human skin makes when sunlight hits it, and traditional supplements often derive it from sheep’s wool lanolin, although animal-free D3 made from lichen is also sold. Both forms are converted in the body into 25-hydroxyvitamin D, the circulating compound that clinicians measure to judge vitamin D status.

The distinction matters because the blood test reports each version separately. Taking D2 raises 25-hydroxyvitamin D2, which counts toward total status, while the D3 measurement is a separate line item. The U.S. National Institutes of Health, in its vitamin D fact sheet for health professionals, says D3 tends to raise and sustain blood levels more than D2 does, while allowing that both can raise overall levels.

Eleven pooled trials and 655 participants in the Nutrition Reviews analysis

The paper, published in 2025 in Nutrition Reviews, searched PubMed for randomized trials from January 1975 to February 2023. Of 202 articles retrieved, 20 entered the systematic review and 11 supplied data for the pooled calculation, covering 655 participants: 342 given D2 and 313 controls. The participants were mostly healthy adults, and pregnant or breastfeeding women were excluded. Doses ranged widely, from 300 IU to 50,000 IU depending on how often they were given, and the pooled trials ran from about four weeks to six months, according to the published abstract.

Two statistical views produced the 9 to 18 range. At the end of the trials, 25-hydroxyvitamin D3 was 17.99 nmol/L lower in D2 groups, with a 95 percent confidence interval of 10.12 to 25.86. Measured as change from baseline, the gap was 9.25 nmol/L, with an interval of 4.10 to 14.40. Heterogeneity between trials was high, at 83 percent for the end-of-trial analysis and 69 percent for the change analysis, which means the individual studies disagreed on the size of the effect even though the direction held.

A blood level is not a health outcome

The trials in this analysis measured blood concentrations, not bone fractures, falls, infections or any other illness. The University of Surrey’s release on the work does not report a health consequence of the D3 drop, and it explicitly says the result does not mean D2 removes vitamin D from the body or is harmful. Total vitamin D status is the sum of the D2 and D3 metabolites, and the release does not say how the total behaved in each trial.

The proposed explanation is a regulatory one. As circulating levels rise, the body may speed up breakdown and removal of vitamin D metabolites, so extra D2 could increase disposal of D3 metabolites already present. The authors describe this as a possible mechanism rather than a demonstrated one, and earlier experiments suggested the effect may run in both directions. They state that more research is needed on whether the two forms differ in overall metabolism and whether that should change supplement recommendations.

D2 is also not biologically inert. A separate 2026 review in Nutrition Reviews, of 26 randomized trials, found that D2 significantly reduced parathyroid hormone and produced a small rise in blood calcium, according to its published record. That is evidence the compound does something measurable, though it says nothing by itself about whether D2 or D3 is better for health.

Why researchers favor D3, and why the label form matters

The recommendation rests on the blood-level evidence rather than on proven differences in disease. Brown said her team discovered that vitamin D2 supplements can decrease levels of vitamin D3 in the body. Professor Cathie Martin, a group leader at the John Innes Centre, said the meta-analysis points to the importance of making plant-based vitamin D3 accessible in the UK, which points to the practical tension: D2 has long been the default vegan option, so a case for D3 depends on animal-free D3 being available.

A second line of evidence came from Professor Colin Smith’s group at Surrey, which compared how the two forms change gene activity in the blood. In that work D3 appeared to stimulate type 1 interferon signaling, an antiviral pathway, while D2 did not. Smith said healthy vitamin D3 status may help prevent viruses and bacteria from gaining a foothold, but changes in gene activity do not establish fewer infections, and no journal reference for that study appears in the release.

The timing of the debate follows public-health guidance. The NHS says everyone should consider a daily supplement containing 10 micrograms of vitamin D during autumn and winter, because between October and early March sunlight is too weak for skin to make enough, according to its vitamin D page. That page does not distinguish between D2 and D3, so the form printed on a bottle is a choice the guidance leaves open, and the Surrey team’s results are an argument for reading it.

This article was produced with the assistance of AI and reviewed by Morning Overview editors prior to publication.

More from Morning Overview