How a 2019 NEJM Study Misled the World on Vitamin D: A Forensic Analysis

They designed the signal out of Vitamin D's impact on cancer and cardiovascular disease

In January 2019, The New England Journal of Medicine published a study that was immediately hailed as the final verdict on vitamin D: it doesn’t work. The study, known as the VITAL trial, was large, well-funded, and led by respected researchers from Harvard. Its conclusion—that vitamin D supplementation did not reduce the risk of invasive cancer or major cardiovascular events—rapidly diffused across headlines, textbooks, and clinical guidelines.

But the VITAL study didn’t fail because vitamin D failed. It failed because it was never designed to test the right question. This article walks through the anatomy of that failure, why it matters, and what we must fix if we are to take prevention seriously in modern medicine.

The Trial That Didn’t

On the surface, VITAL looked impeccable: over 25,000 participants, randomized and placebo-controlled, testing 2000 IU of vitamin D3 daily for a median of 5.3 years. The primary endpoints were the incidence of any invasive cancer and a composite of major cardiovascular events (heart attack, stroke, or death from cardiovascular causes).

But there is a foundational problem: most participants weren’t vitamin D deficient to begin with. Only 12.7% had levels below 20 ng/mL, the threshold generally associated with increased risk. The mean baseline level was 30.8 ng/mL—already at or near sufficiency. It’s the equivalent of testing whether insulin helps people who don’t have diabetes.

Further eroding the study’s contrast, participants in the placebo arm were allowed to take up to 800 IU/day of vitamin D on their own. By year 5, more than 10% of the placebo group was exceeding that limit. The intervention, in effect, became a test of high-dose vitamin D versus medium-dose vitamin D, not against a true control.

Add to that the decision to use broad, bundled endpoints like “any invasive cancer” or “major cardiovascular events” without regard to mechanisms, latency, or stage-specific progression, and the trial becomes a precision instrument for finding nothing.

The Important Real Signal They Missed

The one glimmer of benefit appeared in cancer mortality. While incidence rates were similar between groups, the vitamin D arm showed a lower rate of cancer deaths. This effect emerged only after two years of follow-up and became statistically significant once early deaths were excluded. Even more telling, among participants whose cause of death could be adjudicated with medical records (rather than death certificate codes), the benefit was stronger.

This suggests a biologically plausible mechanism: vitamin D may not prevent cancer from starting, but it may slow its progression or reduce metastasis. That theory aligns with preclinical models showing vitamin D’s role in cellular differentiation, immune modulation, and suppression of angiogenesis.

And yet, VITAL buried this signal. The paper acknowledged a significant violation of the proportional hazards assumption in cancer mortality, a red flag that time-to-event models were inappropriate. Instead of adjusting with valid statistical models for non-proportional hazards, the authors sliced the data post hoc to generate a story and dismissed the result as exploratory. Meanwhile, they mentioned in passing that fewer advanced or metastatic cancers occurred in the vitamin D group—but offered no data.

How Design Choices Shape Public Understanding

The public interpretation of VITAL has been simple and sweeping: vitamin D doesn’t help. That perception has reshaped policy, funding, and clinical guidance. Combined with errant policy based on acknowledged errors, It is dangerous and a risk to public health.

But what the trial actually tested was much narrower: Does high-dose vitamin D provide additional benefit in a mostly vitamin D–sufficient, highly compliant, aging American cohort already permitted to take moderate doses on their own? And does it do so within 5 years?

Given those conditions, the null result was foreordained.

That’s not a failure of science. That’s a failure of trial design.

What Should Have Been Done

A rationally designed prevention trial would start with a population at risk. That means recruiting participants with confirmed vitamin D deficiency, ideally below 20 ng/mL. It would require tighter control of off-protocol supplement use. It would measure achieved serum levels in all participants, not just a 6% subsample. And it would follow participants for a decade or more to match the biological latency of cancer.

Equally critical, the endpoints would reflect mechanistic expectations. Rather than bundling all cancers or all cardiovascular events, researchers should examine site-specific incidence, grade at diagnosis, metastatic progression, and mortality—particularly among subgroups most likely to benefit, such as Black participants and those with low BMI.

Reform Is Not Optional

It is not enough to run large trials. They must be designed to answer the right questions. The failure of VITAL has less to do with vitamin D and more to do with how preventive science is conducted: Over-generalized endpoints, underpowered subgroups, and insufficient attention to biological realism.

We need new standards:

• Targeted enrollment of at-risk populations

• Serum level tracking

• Clear contrasts between intervention and control

• Biomarker tracking throughout

• Outcomes matched to mechanistic hypotheses

• Transparent reporting of all stage-specific and cause-specific outcomes

None of ts is controversial. It is merely rigorous.

This Isn’t Over

Several high-quality meta-analyses and smaller trials contradict the conclusions drawn from VITAL.

Several high-quality meta-analyses and randomized trials contradict the broad null interpretation drawn from the VITAL study. A 2014 Cochrane Review found that vitamin D supplementation, particularly with cholecalciferol (D3), was associated with a statistically significant 13% reduction in cancer mortality. The authors concluded that vitamin D likely reduces the risk of cancer death over a 5–7 year period, even though effects on incidence were not evident.

A randomized controlled trial in Nebraska by Lappe et al., involving postmenopausal women who received 2000 IU/day of vitamin D3 and 1500 mg/day of calcium, showed a non-significant 30% reduction in cancer incidence, with stronger effects emerging in secondary and stratified analyses. An earlier 2007 trial by the same group found a statistically significant reduction in cancer incidence with combined vitamin D and calcium supplementation.

Pooled data from 17 cohorts, as reported by McCullough et al., show a strong inverse association between circulating 25-hydroxyvitamin D [25(OH)D] levels and colorectal cancer risk. Individuals in the highest quintile of serum 25(OH)D had a substantially lower risk of colorectal cancer compared to those in the lowest quintile, across diverse populations.

These findings converge on the possibility that vitamin D is more likely to influence cancer progression and lethality than initial incidence, particularly in populations with low baseline serum levels or in cancers like colorectal cancer that exhibit strong biological responsiveness.

Null trials can be useful. But when designed poorly, they become weapons of inference. The VITAL trial should be reinterpreted, not repeated.

If science is to regain public trust, it must show not just what it found, but what it never really asked.

References

1. Bjelakovic G, Gluud LL, Nikolova D, et al. Vitamin D supplementation for prevention of mortality in adults. Cochrane Database Syst Rev. 2014;1:CD007470. https://www.cochrane.org/evidence/CD007470_vitamin-d-supplementation-prevention-mortality-adults

2. Lappe JM, Watson P, Travers-Gustafson D, et al. Effect of vitamin D and calcium supplementation on cancer incidence in older women: a randomized clinical trial. JAMA. 2017;317(12):1234-1243. https://jamanetwork.com/journals/jama/fullarticle/2613159

3. Lappe JM, Travers-Gustafson D, Davies KM, Recker RR, Heaney RP. Vitamin D and calcium supplementation reduces cancer risk: results of a randomized trial. Am J Clin Nutr. 2007;85(6):1586-1591. https://pubmed.ncbi.nlm.nih.gov/17556697/

4. McCullough ML, Zoltick ES, Weinstein SJ, et al. Circulating vitamin D and colorectal cancer risk: an international pooling project of 17 cohorts. J Natl Cancer Inst. 2019;111(2):158-169. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6821324/