Vitamin D2 is much less effective than vitamin D3 in humans

Lauka A.G. Armas, Bruce W. Hollis, Robert P. Heaney

Research output: Contribution to journalArticlepeer-review

951 Scopus citations


Vitamins D2 and D3 are generally considered to be equivalent in humans. Nevertheless, physicians commonly report equivocal responses to seemingly large doses of the only high-dose calciferol (vitamin D2) available in the U.S. market. The relative potencies of vitamins D2 and D3 were evaluated by administering single doses of 50,000 IU of the respective calciferols to 20 healthy male volunteers, following the time course of serum vitamin D and 25-hydroxyvitamin D (25OHD) over a period of 28 d and measuring the area under the curve of the rise in 25OHD above baseline. The two calciferols produced similar rises in serum concentration of the administered vitamin, indicating equivalent absorption. Both produced similar initial rises in serum 25OHD over the first 3 d, but 25OHD continued to rise in the D3-treated subjects, peaking at 14 d, whereas serum 25OHD fell rapidly in the D2-treated subjects and was not different from baseline at 14 d. Area under the curve (AUC) to d 28 was 60.2 ng·d/ml (150.5 nmol·d/liter) for vitamin D2 and 204.7 (511.8) for vitamin D3 (P < 0.002). Calculated AUC indicated an even greater differential, with the relative potencies for D 3:D2 being 9.5:1. Vitamin D2 potency is less than one third that of vitamin D3. Physicians resorting to use of vitamin D2 should be aware of its markedly lower potency and shorter duration of action relative to vitamin D3.

Original languageEnglish (US)
Pages (from-to)5387-5391
Number of pages5
JournalJournal of Clinical Endocrinology and Metabolism
Issue number11
StatePublished - Nov 2004

ASJC Scopus subject areas

  • Endocrinology, Diabetes and Metabolism
  • Biochemistry
  • Endocrinology
  • Clinical Biochemistry
  • Biochemistry, medical


Dive into the research topics of 'Vitamin D2 is much less effective than vitamin D3 in humans'. Together they form a unique fingerprint.

Cite this