Vitamin D3, not D2 (top 5 reasons)

Bottom line: Cholecalciferol (D3) and ergocalciferol (D2) were long treated as interchangeable, but human data now show D3 is the better choice on every axis that matters — how much it raises your level, how well it holds that level, and whether it interferes with the form your body makes itself. The pharmacopeia still lists them as equivalent; the evidence no longer supports that.

Quick reference

#	Reason	Strongest evidence tier	Key number
1	D3 raises 25(OH)D more than D2	Multiple RCTs + meta-analyses	D3 ~87% more potent (range 56–87%)
2	The gap is widest with large/infrequent doses	RCTs + meta-analysis subgroup	Bolus D3 clearly superior; daily gap narrows
3	D2 actually lowers your measured 25(OH)D3	Meta-analysis of RCTs	~12–18 nmol/L (≈5–7 ng/mL) drop; 18 of 20 studies
4	D2 has a shorter half-life — wrong for monthly dosing	PK/stable-isotope studies	25(OH)D2 half-life shorter; faster clearance
5	D3 is the body's native form — better at every metabolic step, more stable	Mechanistic + PK	Higher affinity for DBP, hepatic 25-hydroxylase, VDR

A note on units: 1 ng/mL = 2.5 nmol/L. Figures below give both where the source used nmol/L.

Reason 1 — D3 raises serum 25(OH)D substantially more than D2

This is the best-replicated finding in the whole D2-vs-D3 literature.

Heaney 2011 (RCT, 33 adults, 50,000 IU/week × 12 wk): D3 was roughly 87% more potent than D2 at raising and maintaining 25(OH)D, and produced 2- to 3-fold greater body stores of vitamin D. Across the various measures used, the potency advantage ranged from about 56% to 87%. [PMID 21177785]
Armas 2004 (RCT, single 50,000 IU dose, 28-day AUC): the paper's title says it plainly — D2 is "much less effective" than D3 by area-under-the-curve of the 25(OH)D rise. This is the study that first quantified the gap cleanly. [PMID 15531486]
Tripkovic 2012 (meta-analysis): pooled RCTs found D3 significantly more efficacious than D2 at raising 25(OH)D (P = 0.001). The 2017 follow-up review reached the same conclusion. [PMID 22552031]
Balachandar 2021 (meta-analysis, Nutrients) and van den Heuvel 2024 (meta-analysis, Advances in Nutrition): both confirmed D3 is more potent overall.

Evidence tier: Strong — consistent RCTs plus three independent meta-analyses.

Reason 2 — The advantage is largest exactly when people use big, infrequent doses

This is the practically important refinement of Reason 1, and it's the opposite of what most prescribers assume.

Tripkovic 2012: when dosing frequency was analyzed, D3 was clearly superior when given as a bolus (P = 0.0002), but the advantage shrank toward non-significance with daily dosing. [PMID 22552031]
van den Heuvel 2024: the D2-vs-D3 efficacy difference was significantly larger for non-daily regimens (e.g., weekly/bolus) than for daily dosing (P < 0.0001). [PMID 37865222]

In other words: the more you lean on a single large dose — the 50,000 IU prescription capsule, the monthly or "loading" dose — the more D2 underperforms. D2 looks least bad in the one scenario (small daily doses) where you'd least need a prescription form anyway.

Evidence tier: Strong (consistent subgroup effects across meta-analyses).

Reason 3 — Taking D2 lowers the 25(OH)D3 your body already had

D2 doesn't just add a weaker form — it appears to actively reduce circulating 25(OH)D3.

Brown et al. 2025 meta-analysis (Nutrition Reviews): D2 supplementation produced a statistically significant fall in 25(OH)D3 versus controls — on the order of ~18 nmol/L (~7 ng/mL) in one analysis and ~9 nmol/L (~3.6 ng/mL) in the absolute-change analysis. 18 of the 20 reviewed studies found a reduction. [PMID 40973107]
Proceedings of the Nutrition Society 2018 meta-analysis: average 25(OH)D3 decrease of ~12 nmol/L (~5 ng/mL) (P < 0.00001), and people with higher baseline levels seemed more susceptible.
Dose-response meta-analysis 2024 (Steroids, 33 study arms): D2 raised total 25(OH)D and 25(OH)D2 but decreased 25(OH)D3 by ~4.6 ng/mL (~11.6 nmol/L). [doi:10.1016/j.steroids.2024.... S0039128X24000321]

Mechanism (likely): Two non-exclusive explanations are proposed. (a) A reciprocal regulatory effect — raising total 25(OH)D ramps up the catabolic enzyme CYP24A1, which clears 25(OH)D3 along with the excess; a 2017–2019 RCT showed D3 reciprocally lowers 25(OH)D2 by a similar magnitude, so the body seems to regulate total 25(OH)D rather than either form specifically (Hammami; reciprocal-phenomenon RCT, PMC6339397). (b) D2's lower DBP affinity may free 25(OH)D for faster disposal.

Honesty flag on the "% of people" question. The literature reports this as a population mean decrease and as study-level consistency (18 of 20 studies), not as a clean per-person rate like "lowers D3 in X% of people." A megadose example (600,000 IU D2, 37 subjects) showed a mean ~4 ng/mL D3 drop, but that's still a group mean. If we want to state a per-person percentage on the wiki, we'd need to pull individual-level data from the primary RCTs (Lehmann 2013, Hammami 2017) and compute it ourselves — worth doing, but I won't assert a number we haven't derived.

Evidence tier: Moderate-to-strong for the direction and average magnitude; weak for any per-person percentage and for clinical (not just biomarker) consequences.

Reason 4 — D2 has a shorter half-life, so it's the wrong tool for monthly (or rarer) dosing

Jones/Tang 2014 (JCEM, stable-isotope study): measured directly that 25(OH)D2 half-life is shorter than 25(OH)D3 half-life, and that half-life is further modulated by DBP concentration and genotype. [PMID 24885631]
Houghton & Vieth 2006 ("The case against ergocalciferol"): the weaker binding of D2 metabolites to DBP leads to a shorter circulating half-life and faster clearance. [PMID 17023693]
Logan/pharmacokinetic study 2015: D2 and D3 were equally good at raising 25(OH)D after a loading dose, but D3 was better at sustaining it over the longer term. [PMID 25782422]

Practical implication: the shorter half-life means D2 troughs faster between doses. The widely used 50,000 IU monthly or every-other-week regimens are the worst-case match for D2 — by the time the next dose arrives, more of the previous one is gone. If a clinician insists on intermittent high-dose therapy, D3 is the form that actually holds a plateau.

Evidence tier: Strong for the half-life/clearance difference; the monthly-dosing conclusion is a sound inference from PK rather than a head-to-head monthly-dosing outcome trial.

Reason 5 — D3 is the body's native form: better at every metabolic step, and more stable on the shelf

This is the unifying "why" behind Reasons 1–4, plus two practical extras.

Native vs. foreign molecule. D3 is what human skin makes from sunlight and what's in animal-source foods; D2 comes from UV-irradiated fungal/plant ergosterol and differs by a side-chain double bond and an extra methyl group. That small structural difference is the root of everything below.
Lower affinity at each step. Compared with D3, D2 binds more weakly to vitamin D binding protein, to the hepatic 25-hydroxylase, and to the vitamin D receptor (VDR) (Houghton & Vieth 2006). Fewer molecules carried, less efficiently activated, weaker signaling.
Non-physiologic catabolism. D2 can be routed directly to 24-hydroxylated (inactive) products in a way D3 is not — a "non-physiologic" metabolism (Houghton & Vieth).
Shorter shelf life. D2 is also less chemically stable in supplements/fortified products, so label dose and delivered dose can drift apart over time.

Evidence tier: Mechanistic/biochemical (well-established in vitro and in primates), supporting — not by itself outcome-proving.

What this does NOT show (caveats kept honest)

D2 still works. It does raise total 25(OH)D and treats deficiency, rickets, and osteomalacia. The argument is that D3 does it better and cleaner, not that D2 is inert.
Daily small doses narrow the gap. With ordinary daily dosing the efficacy difference is smaller and, in some analyses, not significant — most of D2's disadvantage shows up with bolus/intermittent dosing.
Body weight matters. In people with overweight/obesity, the D2-vs-D3 difference in raising 25(OH)D may disappear; the clearest D3 advantage is in healthy-weight individuals (van den Heuvel 2024).
Biomarker ≠ outcome. Most of this is about serum 25(OH)D, the status marker. Direct head-to-head trials on hard clinical endpoints (fractures, infections, mortality) comparing D2 vs D3 are sparse.
The reciprocity caveat. Because D3 also lowers 25(OH)D2, the body appears to defend a total 25(OH)D set-point. The D2-lowers-D3 effect is real, but it's part of a regulated system, not a one-way poisoning of D3.
Vegan note. The historical reason to choose D2 was that it's plant/fungal-derived. That's now moot: lichen-derived D3 gives vegans the better form.

Key references

Heaney RP, et al. Vitamin D3 is more potent than vitamin D2 in humans. J Clin Endocrinol Metab. 2011;96(3):E447–E452. PMID 21177785. doi:10.1210/jc.2010-2230
Armas LAG, Hollis BW, Heaney RP. Vitamin D2 is much less effective than vitamin D3 in humans. J Clin Endocrinol Metab. 2004;89(11):5387–5391. PMID 15531486.
Tripkovic L, et al. Comparison of vitamin D2 and vitamin D3 supplementation in raising serum 25(OH)D: systematic review and meta-analysis. Am J Clin Nutr. 2012;95(6):1357–1364. PMID 22552031.
Tripkovic L, Wilson LR, Lanham-New SA. Vitamin D2 vs. vitamin D3: they are not one and the same. Nutr Bull. 2017;42:331–337.
Balachandar R, et al. Relative efficacy of vitamin D2 and D3 in improving vitamin D status: systematic review and meta-analysis. Nutrients. 2021;13(10):3328.
van den Heuvel EGHM, et al. Comparison of daily vitamin D2 and D3 supplementation on serum 25(OH)D (total, D2, D3) and importance of BMI: systematic review and meta-analysis. Adv Nutr. 2024;15:100133. PMID 37865222.
Brown [et al.]. Effect of vitamin D2 supplementation on 25(OH)D3 status: systematic review and meta-analysis of RCTs. Nutr Rev. 2025. PMID 40973107. doi:10.1093/nutrit/nuaf166
Effect of vitamin D2 supplementation on serum 25(OH)D3 levels: systematic review and meta-analysis. Proc Nutr Soc. 2018;77(OCE1).
Effect of vitamin D2 supplementation on vitamin D levels: time- and dose-response meta-analysis of RCTs. Steroids. 2024. (S0039128X24000321)
Vitamin-D2 treatment-associated decrease in 25(OH)D3 is a reciprocal phenomenon: a randomized controlled trial. PMC6339397.
Jones KS, et al. 25(OH)D2 half-life is shorter than 25(OH)D3 half-life and is influenced by DBP concentration and genotype. J Clin Endocrinol Metab. 2014. PMID 24885631.
Houghton LA, Vieth R. The case against ergocalciferol (vitamin D2) as a vitamin supplement. Am J Clin Nutr. 2006;84(4):694–697. PMID 17023693.
Vitamin D3 seems more appropriate than D2 to sustain adequate 25OHD: a pharmacokinetic approach. PMID 25782422.

Claude AI - June 2026

Editorial standard: evidence tiered (RCT > observational > mechanistic), gaps flagged explicitly, biomarker-vs-outcome distinction maintained, and the one place the requested framing ("% of people") outran the data is called out rather than papered over.