Does vitamin D prevent disease? Decoding the supplement claim

Vitamin D has one of the widest gaps in supplement science between the claim on the bottle and what the trial record actually shows. The claim under review here is broad: that vitamin D supplements prevent disease in general, and that essentially everyone benefits from taking them. Correcting a real, measured deficiency is a different question from that broad claim, and the trial evidence treats the two very differently.

What the largest trial actually found

The best single answer comes from the VITAL trial (Manson and colleagues, New England Journal of Medicine, 2019), a randomized, placebo-controlled trial of 25,871 generally healthy US adults, men 50 and older and women 55 and older, who were not selected for having low vitamin D levels. Half received 2,000 IU of vitamin D3 daily for a median of 5.3 years, half received placebo. The result was unambiguous: vitamin D supplementation did not lower the rate of invasive cancer (hazard ratio 0.96, 95% CI 0.88 to 1.06) or of major cardiovascular events, a composite of heart attack, stroke, and cardiovascular death (hazard ratio 0.97, 95% CI 0.85 to 1.12). Neither result came close to statistical significance. This was not an underpowered study; it is one of the largest vitamin D trials ever run, and it found no protective effect on the two outcomes people most often cite as the reason to supplement.

Two later VITAL ancillary studies extended the picture. A 2022 follow-up on fractures (LeBoff and colleagues, New England Journal of Medicine) tested whether the same vitamin D3 regimen reduced the risk of broken bones in the same generally healthy, non-deficient population, and again found no significant reduction in fracture risk compared with placebo. That result runs against a common assumption that vitamin D is primarily a bone-protection supplement; in adults who are not deficient and do not have osteoporosis, this large randomized trial did not confirm it.

The autoimmune-disease result cuts the other way. A separate 2022 VITAL analysis (Hahn and colleagues, BMJ) tracked self-reported, medical-record-confirmed autoimmune disease diagnoses, an outcome not part of the trial’s original primary endpoints, and found that five years of vitamin D3 supplementation, with or without omega-3, reduced autoimmune disease incidence by 22%. This is a real signal from a large randomized trial, though it was a secondary outcome analysis rather than the trial’s primary hypothesis, which matters for how much weight it should carry relative to the null primary results above.

Where the evidence is strong versus where it is weak

The evidence is strong and consistent on one narrow point: in adults with normal, replete vitamin D levels, giving more vitamin D does not measurably reduce cancer, cardiovascular events, or fractures. That is not a fringe finding; it holds across the largest and best-designed trial in the field, and the fracture result specifically contradicts the folk assumption that more vitamin D reliably means stronger bones in people who are not deficient. The evidence is weaker, and points a different direction, on autoimmune disease, where a large trial found a real reduction, though from a secondary analysis that has not yet been replicated by an independently designed confirmatory trial.

The broader clinical picture is also less settled than either enthusiasts or skeptics tend to suggest. The US Preventive Services Task Force reviewed the evidence on screening for vitamin D deficiency in adults in 2021 and concluded, again, that the evidence is insufficient to weigh the benefits and harms of population-wide screening, a repeat of its 2014 conclusion. That is a statement about screening, not about treating a deficiency once it is found, but it reflects the same underlying gap: even after decades of observational interest in vitamin D, the randomized evidence for population-level screening or blanket supplementation has not closed.

Where deficiency correction is a different question

None of this means vitamin D is irrelevant to health. Frank vitamin D deficiency is a recognized clinical condition with established consequences, including a well-documented relationship to bone and muscle disorders, and correcting a measured deficiency in someone who actually has one is a distinct clinical question from the broad prevention claim tested in VITAL. People with limited sun exposure, darker skin, malabsorption conditions, older age with reduced skin synthesis, or a lab-confirmed low blood level fall into a different evidence category than the generally healthy, non-deficient population VITAL enrolled. The size and relevance of any benefit from correcting a deficiency in these groups is a matter between a person and a clinician who has actually measured the level and considered the individual history, not something a population-level prevention trial in replete adults can settle either way.

The honest bottom line

The claim that vitamin D supplements broadly prevent disease in the general population is not supported by the largest trial designed to test it. Cancer, cardiovascular events, and fractures all came back null in a trial built with enough size and duration to detect a real effect if one existed. Autoimmune disease is the exception, a genuine positive signal from the same trial program, but from a secondary analysis, not the primary hypothesis. That mixed pattern, real evidence for one outcome and clear null results for others, is why this claim lands as mixed rather than clearly supported or clearly refuted. Correcting an actual, measured deficiency is a separate, more defensible use of vitamin D than treating it as a general-purpose preventive supplement for people whose levels are already normal.