Prostate cancer survival was better if higher vitamin D measured 6 years before diagnosis - meta-analysis
Prediagnostic circulating 25-hydroxyvitamin D and prostate cancer-specific survival: a collaborative pooled analysis of 13 prospective studies
BMC Medicine, Aug 22, 2026, https://doi.org/10.1186/s12916-026-05151-w
Jiaqi Huang, Aurora Perez-Cornago, Ruth C. Travis, Paul N. Appleby, Danrui Chen, Mahboubeh Parsaeian, Urwah Noor, Stephanie J. Weinstein, Neal D. Freedman, Andreas Josefsson, Ingegerd Johansson, Elizabeth A. Platz, Corinne E. Joshu, Bu B. Yeap, Leon Flicker, Lorelei A. Mucci, Konstantinos K. Tsilidis, Paul Knekt, Pamela L. Lutsey, Alison Mondul, Kala Visvanathan, Kathy J. Helzlsouer, Graham G. Giles, Matthias B. Schulze, Mélanie Deschasaux-Tanguy, Mathilde Touvier, Jose M.ª Huerta, Loic Le Marchand, Meir Stampfer, Timothy J. Key, Demetrius Albanes, Naomi E. Allen
Summary by Claude - November 2026
Men whose vitamin D was 30+ ng/mL years before a prostate cancer diagnosis were 34% less likely to die of that cancer than men below 12 ng/mL.
This is the largest study of the question so far. It pooled individual data on 12,635 men with prostate cancer from 13 prospective cohorts in the US, Europe and Australia, including UK Biobank. Blood was drawn at least a year before diagnosis (about 6 years on average), so neither the cancer nor its treatment could have lowered vitamin D levels. Earlier meta-analyses partly used levels measured after diagnosis, which has that problem. Over a median 5.6 years after diagnosis, 1,316 men died of prostate cancer.
| Vitamin D before diagnosis | Prostate cancer deaths (fully adjusted) |
|---|---|
| Under 12 ng/mL (<30 nmol/L) | Reference group (11% of men) |
| 12–20 ng/mL (30–50 nmol/L) | 17% fewer (HR 0.83) |
| 20–30 ng/mL (50–75 nmol/L) | 22% fewer (HR 0.78) |
| 30+ ng/mL (≥75 nmol/L) | 34% fewer (HR 0.66, 95% CI 0.54–0.82) |
The link held after adjusting for tumor stage and grade, held in the first 2 years after diagnosis, and held when other causes of death were treated as competing risks. It was seen only in men diagnosed at 65 or older (HR 0.69, vs 1.05 under 65), though that age difference missed the authors' strict significance cutoff. Deaths from other causes were also slightly lower (HR 0.85), but not in the competing-risk analysis. The active hormone 1,25(OH)₂D, measured in only 1,889 men, showed no link.
The authors conclude that "higher pre-diagnostic circulating vitamin D (25(OH)D) was associated with increased prostate cancer survival."
There is a twist. In 2019 the same research group reported that men with higher vitamin D were more often diagnosed with prostate cancer, but only the non-aggressive kind (Travis 2019). Both results fit health-conscious men being screened more and diagnosed earlier.
The study was peer-reviewed and funded by the US National Cancer Institute and Cancer Research UK.
What this does NOT show
- Cause and effect. The study is observational: nobody was given vitamin D, so there is no dose (IU), and it says nothing about starting vitamin D after diagnosis.
- That earlier detection explains nothing. The link looked strongest in early-stage cancer (HR 0.63) and weakest in advanced cancer (HR 0.91). That is the pattern expected if men with higher vitamin D were caught earlier, although the difference was not statistically significant.
- That other factors were ruled out. Sun exposure, outdoor exercise, supplement use and kidney disease were not measured.
- That the result holds up fully in the best-documented men. In the 40% with complete stage and grade data, the link weakened (HR 0.81, P = 0.04), which misses the authors' own P < 0.013 cutoff.
- Anything about 40–60 ng/mL. The top group started at only 30 ng/mL.
- Long-term vitamin D status. Each man had a single blood test.
- Results for non-White men. Only 796 were included.
Sources:
Related in VitaminDWiki
- Prostate Cancer death 40 percent less likely if 40 ng level of vitamin D – Meta-analysis (Song 2018, 7 cohorts)
- 1.5X more likely to die of Prostate Cancer if low vitamin D when initially diagnosed – meta-analysis (Cui 2024, vitamin D measured at diagnosis)
- 59 percent more likely to die of Prostate Cancer if very low on vitamin D
(Fang 2011; its two cohorts are in this pool) - Better Cancer survival if higher vitamin D a decade earlier (esp. Melanoma, Kidney, Prostate) (Weinstein 2018; ATBC is in this pool)
- Prostate cancer risk increase with Vitamin D (other studies disagree) – meta-analysis
Prostate Cancer death 40 percent less likely if 40 ng level of vitamin D – Meta-analysis
- Vitamin D intake and mortality of 5 cancers: breast, colorectal, lung, pancreatic, and prostate - umbrella
- Deaths from many types of Cancer associated with low vitamin D- review of meta-analyses
- Prostate cancer risk reduced by Vitamin D (4000 IU daily)
- Prostate Cancers have 2.5 X more microplastics than non-tumor sections
- Risk of Prostate Cancer 3X higher if very low Magnesium
- Overview Cancer and vitamin D