Vitamin D is recommended for bone health almost universally. So a trial of 25,871 adults finding no fracture reduction is worth examining carefully — including what it doesn’t say.
The headline is simpler than the conclusion.
What VITAL tested
VITAL was a randomised, placebo-controlled, two-by-two factorial trial in men aged 50 and over and women aged 55 and over across the United States[1].
- 25,871 participants — 50.6% women, 20.2% Black
- 2,000 IU vitamin D3 daily versus placebo
- Median follow-up 5.3 years
- 1,991 confirmed fractures in 1,551 participants[1]
Fractures were self-reported on annual questionnaires and then adjudicated by centralised medical-record review[1]. That’s a strength — the outcome wasn’t taken on trust.
The fracture analysis was a pre-registered ancillary study of the parent trial, designed specifically to test whether vitamin D reduces total, nonvertebral and hip fractures, and whether any effect varies by baseline vitamin D level[5].
The one design point that governs everything else: participants were not recruited on the basis of vitamin D deficiency, low bone mass, or osteoporosis[1].
What it found
Nothing, and consistently.
| Outcome | Hazard ratio | 95% CI |
| Total fractures | 0.98 | 0.89–1.08 |
| Nonvertebral fractures | 0.97 | 0.87–1.07 |
| Hip fractures | 1.01 | 0.70–1.47 |
[Table 1] VITAL fracture outcomes · Source: LeBoff et al., NEJM 2022[1]
Every confidence interval crosses 1. Total fractures occurred in 769 of 12,927 on vitamin D and 782 of 12,944 on placebo[1] — a difference of thirteen fractures across nearly 26,000 people.
Secondary and exploratory endpoints matched: major osteoporotic fractures, wrist, pelvic — all null[2].
And there was no modification of the treatment effect by age, sex, race or ethnic group, body-mass index, or baseline 25(OH)D levels[1].
That last item deserves attention. Even participants who started with lower vitamin D levels didn’t benefit.
A companion analysis found the same in other measures: in a subcohort, supplemental vitamin D did not improve bone density or structure (n=771) or physical performance (n=1,054)[3].
The limitation that defines the result
Here’s the number that reframes everything.
Mean baseline 25(OH)D in VITAL was 30.7 ± 10 ng/mL[3].
Thirty is at or above the threshold most guidelines describe as sufficient. The participants were already replete.
The review summarising VITAL alongside the DO-HEALTH trial states the scope precisely: in generally healthy and vitamin D-replete older adults not preselected for vitamin D deficiency or low bone mass or osteoporosis, 2,000 IU/day had no musculoskeletal benefit[3].
That’s a narrow and accurate claim. It is not “vitamin D doesn’t prevent fractures.”
The lead author’s own summary was similarly bounded: the results do not support the use of vitamin D supplements to reduce fractures in generally healthy U.S. men and women[4].
Both statements carry qualifiers that headlines tend to drop.
What the trial could not answer
Three groups sit outside what VITAL tested.
People who are genuinely deficient. With a mean baseline of 30.7 ng/mL, the trial had few. This is the same structural issue we examined in the soy intake post — a supplementation trial measures an increment, and adding 2,000 IU to someone already at 30 tests something different from adding it to someone at 12.
People with osteoporosis or low bone mass. Explicitly not selected for[1].
Vitamin D plus calcium. VITAL tested vitamin D alone. As covered in the calcium absorption post, the vitamin-D-dependent pathway carries most weight precisely when calcium intake is low — so testing one without the other leaves a gap.
Notably, personal use of supplemental calcium and/or vitamin D did not modify the results[4] — but participants taking their own supplements is different from a trial designed to test the combination.
Reading this alongside the falls research
VITAL fits a pattern this series has now encountered several times.
In the falls post, 500,000 IU annually raised falls 15% and fractures 26%, and 60,000 IU monthly produced more fallers than 24,000 IU monthly. Meanwhile 700–800 IU daily reduced falls in populations that were deficient.
Put those beside VITAL and a consistent shape emerges.
Correcting a deficiency appears to help. Adding more to someone already sufficient does not — and pushing much higher can harm.
That’s not a hedge. It’s a specific claim with a specific implication: the useful question is whether you’re short, not how much you can add.
What this means practically
If your vitamin D status is unknown, that’s the gap worth addressing rather than the supplement dose. As covered in the baseline assessment post, vitamin D isn’t in Korea’s standard screening panel.
If you have osteoporosis or low bone mass, VITAL doesn’t apply to you. Management in that situation is a clinical matter with established treatments beyond supplementation.
If you’re taking vitamin D for bone health while replete, VITAL is the most relevant evidence available, and it found nothing over 5.3 years in nearly 26,000 people.
If you’re deficient, the trial genuinely doesn’t tell you what happens — and correcting deficiency remains standard practice for reasons unrelated to this trial.
Closing
VITAL is a well-conducted trial with an unambiguous result and a frequently misquoted meaning.
What it established: in generally healthy, vitamin-D-replete adults over 50, 2,000 IU daily produced no fracture reduction over 5.3 years, with hazard ratios sitting essentially at 1 across every endpoint.
What it did not establish: that vitamin D is irrelevant to bone health, or that deficient people won’t benefit from correction.
The distinction turns on one number — a mean baseline of 30.7 ng/mL. A trial conducted in replete people can only tell you what happens when you add more to enough.
Which is exactly the question worth asking before starting anything: not how much, but whether there’s a shortfall to correct.
Key Terms
- Hazard ratio — the rate of events in one group relative to another. 1.0 means no difference.
- Confidence interval — the range the true value likely falls within. If it crosses 1, “no difference” remains possible.
- Replete — having adequate stores of a nutrient already.
- 25(OH)D — the blood measure used to assess vitamin D status.
- Adjudicated — verified against medical records rather than accepted as reported.
- Nonvertebral fracture — a fracture other than in the spine.
At a Glance
- VITAL: 25,871 participants, 2,000 IU/day vitamin D3 vs placebo, median 3 years
- 1,991 fractures confirmed by centralised medical-record review
- Total fractures HR 0.98 (0.89–1.08), nonvertebral 97, hip 1.01 — all null
- 769 vs 782 total fractures across nearly 26,000 people
- No effect modification by age, sex, race, BMI, or baseline 25(OH)D
- Subcohort: no improvement in bone density, structure, or physical performance
- Mean baseline 25(OH)D was 30.7 ng/mL — participants were already replete
- Not recruited for deficiency, low bone mass, or osteoporosis; calcium combination not tested
※ This article discusses trial findings and is for general information only. It does not replace medical advice, diagnosis or treatment, and is not a recommendation to start or stop vitamin D. If you have osteoporosis, low bone mass, or a history of fractures, management should be directed by a clinician — VITAL’s findings do not apply to those situations. Do not begin high-dose supplementation without clinical guidance.
References
- LeBoff MS, Chou SH, Ratliff KA, et al. “Supplemental Vitamin D and Incident Fractures in Midlife and Older Adults”, New England Journal of Medicine 2022;387(4):299-309, https://pubmed.ncbi.nlm.nih.gov/35939577/
- “Supplemental Vitamin D and Incident Fractures in Midlife and Older Adults” — secondary and exploratory endpoint hazard ratios, New England Journal of Medicine, https://www.nejm.org/doi/full/10.1056/NEJMoa2202106
- “The Effects of Vitamin D Supplementation on Musculoskeletal Health: The VITAL and DO-Health Trials”, PubMed (baseline 25(OH)D levels; subcohort bone density and physical performance findings), https://pubmed.ncbi.nlm.nih.gov/37325962/
- “Supplemental vitamin D did not lower risk of fractures in healthy US adults”, EurekAlert (author summary and effect modification analyses), https://www.eurekalert.org/news-releases/959693
- “VITamin D and OmegA-3 TriaL (VITAL): Fractures, Vitamin D and Genetic Markers”, ClinicalTrials.gov NCT01704859 (ancillary study design), https://clinicaltrials.gov/study/NCT01704859