Often need 2X to 4X more Vitamin D to treat a disease than to prevent it
Claude AI Opus 5.5 research - Nov 2026
Prevention vs. treatment doses where both are stated
| Condition | Prevent (IU/day) | Treat (IU/day) | Treat ÷ Prevent | Refs |
|---|---|---|---|---|
| Rickets (children) | 400 | 2,000–6,000 for ≥12 weeks | 5–15X | Rickets consensus Consensus summary |
| Deficiency, adults (Endocrine Soc. 2011) | 1,500–2,000 maintenance | 6,000 (or 50,000/week) × 8 weeks | 3–4X | Endocrine Soc. table |
| Deficiency, children 1–18 | 600–1,000 maintenance | 2,000 × 6 weeks | 2–3X | Endocrine Soc. table |
| Deficiency + obesity or malabsorption | 3,000–6,000 maintenance | 6,000–10,000 | ~2X | Endocrine Soc. table |
| Autism (Cannell's proposal) | 150 IU/kg (young children) | 300 IU/kg | 2X | Cannell 2017 |
| Type 2 diabetes (VitaminDWiki) | >2,000 | >4,000 | ~2X | VitaminDWiki overview |
| Cancer (VitaminDWiki) | 2,000–5,000 | >8,000 | ~2–4X | Cancer overview |
So 2X is the low end of the range. It fits conditions where patients only need to reach a somewhat higher blood level, not to rebuild a large deficit fast.
Why 2X in blood level can mean 3–4X in IU
The dose-response curve flattens as doses go up. In a study of 17,614 adults, serum 25(OH)D rose about 12 nmol/L per 1,000 IU at doses up to 1,000 IU/day, but only about 1.1 nmol/L per 1,000 IU at 15,000–20,000 IU/day (Ekwaru 2014). That means "about 2X" may be right if you measure in ng/mL, while the IU ratio comes out larger. That likely explains the gap between your 2X impression and the 4X on your page.
Likely reasons treatment needs more
- Illness uses up or lowers 25(OH)D. There is strong evidence that 25(OH)D behaves as a negative acute-phase reactant, and surgery may cut circulating levels by about 40% (CCJM 2023). An endotoxin experiment in humans showed directly that systemic inflammation lowers circulating 25(OH)D (Endotoxemia study). In chronic low-grade inflammation, though, Mendelian randomization found no causal effect of CRP on 25(OH)D (MR study). So this reason is strongest for acute illness.
- Patients are a selected group. The people who got sick tend to include more with obesity, malabsorption, or drugs that speed vitamin D breakdown. Researchers recommend 2–3 times more vitamin D for obese people and 1.5 times more for overweight people (Ekwaru 2014). In lymphomas, sarcoidosis, TB and other granuloma-forming disorders, extra-renal conversion of 25(OH)D puts patients at high risk of deficiency (MDedge). Those same patients are also at risk of hypercalcemia, so high doses need monitoring.
- Speed. Prevention can take months to reach steady state. Treatment needs adequate levels quickly, which is why trials often use loading doses.
- A higher bar. Reversing a process that has already started is harder than stopping it from starting. Your page also notes that cofactors such as magnesium and K2 often seem necessary for treatment.
- (added by VitaminDWiki:) Some diseases protect themselves by reducing the activation of some genes, examples for Vitamin D include:
Breast Cancer,
Diabetes,
Colon Cancer and
COVID ( over 100 genes - not just vitamin D)
Related in VitaminDWiki
- Diseases, Vitamin D Receptor and other genes
- Many cancers protect themselves by deactivating Vitamin D activated genes ( Pancreatic Cancer and HNF4A in this case)
- Women with Breast Cancer were 16.9 times more likely to have a poor Vitamin D Receptor
- Cancer treatment by Vitamin D sometimes is restricted by genes
Caveats
- These aren't head-to-head comparisons. Prevention doses mostly come from population and observational data, and treatment doses from small trials. Part of the ratio just reflects which doses investigators chose to test.
- Some treatment fails at any dose. In VIOLET, a single 540,000 IU dose quickly corrected deficiency in critically ill patients but gave no benefit on mortality or other outcomes (VIOLET, NEJM 2019). Compare that with early MS. In D-Lay MS, 100,000 IU every 2 weeks (about 7,100 IU/day) cut disease activity to 60.3% vs. 74.1% on placebo and roughly doubled the median time to new activity, from 224 to 432 days (D-Lay MS, JAMA 2025). Treating early seems to matter as much as the dose.
- The autism dose has a weak footing. The RCT behind the 300 IU/kg treatment dose (Saad 2016) was retracted, citing significant discrepancies and missing-data problems (Retraction coverage). If VitaminDWiki pages cite it as an RCT, they may need a note.
A summary consistent with this evidence: treatment usually takes about 2–4X the daily IU of prevention, often starting with a loading dose. In blood level it's closer to 1.5–2X, and rickets is the outlier at 5–15X.
Sources:
- VitaminDWiki – Vitamin D also TREATS
- VitaminDWiki – Overview
- VitaminDWiki – Cancer overview
- Global rickets consensus 2016
- Rickets consensus summary
- Endocrine Society 2011 treatment table
- Ekwaru 2014 – dose response and body weight
- CCJM 2023 – negative acute-phase reactant
- Endotoxemia lowers 25(OH)D
- MR – CRP and 25(OH)D
- MDedge – granulomatous disorders
- Cannell 2017 – autism
- Saad RCT retraction
- VIOLET trial
- D-Lay MS trial