Vitamin D supplementation: VITAL, fractures, and the deficiency-versus-supplementation distinction

Key takeaways

• VITAL randomized 25,871 generally healthy adults (men 50 or older, women 55 or older) in a 2-by-2 factorial design to vitamin D3 2000 IU per day, marine n-3 1 gram per day, both, or placebo. Median follow-up was 5.3 years. The vitamin D arm was negative on the primary outcomes of total invasive cancer (HR 0.96; P 0.47) and major cardiovascular events (HR 0.97; P 0.69) .
• VITAL was not enriched for vitamin D deficiency. Mean baseline 25-hydroxyvitamin D was about 30 ng per mL, in the replete range. The trial answers what supplementation does in already-replete adults, not what treating deficiency does.
• LeBoff 2022 (VITAL fractures arm) randomized the same cohort and found no reduction in total fractures (HR 0.98), nonvertebral fractures (HR 0.97), or hip fractures (HR 1.01) .
• Hahn 2022 BMJ pre-specified secondary analysis reported a 22 percent reduction in confirmed autoimmune disease over 5.3 years (HR 0.78; 95 percent CI 0.61 to 0.99) on vitamin D versus placebo .
• Cancer mortality (versus incidence) showed a possible signal in VITAL when the first 1 to 2 years of follow-up were excluded, hypothesis-generating but not the primary outcome.
• ODS NIH guidance is that the recommended dietary allowance for adults is 600 IU per day (15 mcg) up to age 70, and 800 IU (20 mcg) over age 70; the upper limit is 4000 IU per day .

The deficiency-versus-supplementation distinction

The most important framing for vitamin D evidence is that there are two distinct clinical questions that are easily confused. Question one: in an adult with documented vitamin D deficiency (commonly defined as 25-hydroxyvitamin D less than 20 ng per mL), does supplementation correct the deficiency and reduce harm associated with the deficiency? The answer is broadly yes for bone outcomes in that population.

Question two: in a generally healthy adult who is already vitamin-D-replete (25-hydroxyvitamin D 30 ng per mL or higher), does adding routine supplementation reduce cardiovascular events, cancer, or fractures? VITAL is the largest and best-powered answer to question two and the answer is no.

Most retail vitamin D supplementation conversations are about question two. Most of the older observational literature linking low 25-hydroxyvitamin D to bad outcomes is about question one. Confusing the two has driven a lot of supplementation that the trial record does not support.

What VITAL actually tested

VITAL was a randomized, double-blind, placebo-controlled, 2-by-2 factorial trial sponsored by the National Institutes of Health. Eligibility was men 50 years or older and women 55 years or older without prior cancer or major cardiovascular disease. Mean baseline 25-hydroxyvitamin D was about 30 ng per mL, with about 12 percent of participants in the deficient range (less than 20 ng per mL).

Participants were randomized to vitamin D3 (cholecalciferol) 2000 IU per day or placebo, and independently to marine n-3 fatty acids 1 gram per day or placebo. The two primary endpoints were major cardiovascular events (cardiovascular death, MI, or stroke) and total invasive cancer.

Over a median 5.3 years, neither primary endpoint was reduced by vitamin D supplementation. The hazard ratios were close to 1.0 with confidence intervals that excluded clinically important effect sizes .

The fracture analysis

Older observational and trial-level data had supported a small reduction in fractures with vitamin D supplementation, particularly in older adults at risk. The VITAL fracture analysis (LeBoff 2022 NEJM) is the largest RCT to specifically test this question in a generally healthy midlife and older adult population not selected for deficiency, low bone mass, or osteoporosis .

Over a median 5.3 years, vitamin D3 versus placebo produced no reduction in total fractures (HR 0.98), nonvertebral fractures (HR 0.97), or hip fractures (HR 1.01). The accompanying NEJM editorial framed this as a decisive verdict on routine vitamin D supplementation for fracture prevention in the general population, distinct from treatment of confirmed deficiency or osteoporosis.

The autoimmune disease finding

Hahn et al. 2022 reported the pre-specified VITAL secondary endpoint of incident confirmed autoimmune disease (rheumatoid arthritis, polymyalgia rheumatica, autoimmune thyroid disease, psoriasis, and inflammatory bowel disease, among others). Vitamin D versus placebo produced a 22 percent relative risk reduction over 5.3 years (HR 0.78; 95 percent CI 0.61 to 0.99) .

The combination of vitamin D plus n-3 fatty acids produced a slightly larger reduction (HR 0.69) but the omega-3 alone arm did not reach significance on its own. The autoimmune finding is hypothesis-generating; it is the strongest positive signal in the VITAL program and is being pursued in mechanistic and confirmatory work.

What VITAL cannot rule out

• Benefits in adults with documented vitamin D deficiency (less than 20 ng per mL) at baseline. The roughly 12 percent of VITAL participants in this range were a subgroup, not enriched, and underpowered for definitive conclusions.
• Higher doses (4000 IU per day or higher) over longer follow-up (10 plus years) for cancer or cardiovascular outcomes.
• Effects in non-white populations, premenopausal women, or younger adults. VITAL was a midlife-and-older cohort.
• Effects in patients with pre-existing osteoporosis, malabsorption, or chronic kidney disease. These patients were excluded or rare in VITAL.
• Effects on cancer mortality versus cancer incidence. A pre-specified secondary analysis excluding the first 1 to 2 years of follow-up showed a modest reduction in cancer death; this is hypothesis-generating, not the primary outcome.
• Whether intermittent high-dose dosing (50,000 IU monthly, for example) produces different outcomes than the daily 2000 IU schedule used in VITAL.

Practical implications

• Routine 1000 to 2000 IU per day vitamin D supplementation in vitamin-D-replete generally healthy adults will not reduce cardiovascular events, total cancer, or fractures, based on the largest available RCT.
• Treating documented vitamin D deficiency (less than 20 ng per mL) is a distinct clinical question and is supported by other evidence; it is not what VITAL tested.
• Patients with osteoporosis, malabsorption, chronic kidney disease, or specific clinical risk factors for fracture often have indications for vitamin D supplementation that pre-date VITAL and are not invalidated by it.
• The autoimmune-disease finding is intriguing and worth tracking in follow-up trials.
• ODS NIH RDA is 600 IU per day for most adults and 800 IU for adults over 70. The tolerable upper intake level is 4000 IU per day .

Editorial summary

VITAL is the closest thing to a definitive answer the field has produced on routine vitamin D supplementation for primary prevention in generally healthy adults. The primary endpoints (cardiovascular events, cancer, fractures) were neutral. The autoimmune disease pre-specified secondary endpoint was modestly positive and warrants follow-up. None of this changes the case for treating documented deficiency, which remains a separate clinical question with its own evidence base.