Vitamin D: The Supplement With the Biggest Gap Between Hype and Evidence
By Akash S. Chauhan | First Principles Healthspan, Issue 05
Few supplements have generated a larger gap between the claims made about them and what the rigorous trial evidence actually shows than vitamin D. The observational literature suggested it was protective against cancer, cardiovascular disease, autoimmune conditions, cognitive decline, and more. The longevity-optimization community ran with those associations and landed on targets like 60-80 ng/mL that have essentially no RCT support. Then the RCTs arrived — including one of the largest supplement trials ever run — and the picture became considerably more complicated.
This issue is not an argument that vitamin D is useless. It is an argument for precision: deficiency is real, correction matters, and the evidence for supplementation beyond that is far weaker than the dosing advice circulating in health optimization spaces would suggest.
Why this matters
The reason the vitamin D story deserves careful analysis is not that it is an obscure edge case. Vitamin D is the most-prescribed supplement in clinical practice and among the most-purchased over the counter. Influencers and functional medicine practitioners routinely recommend maintaining levels of 60 ng/mL or higher and supplementing with 5,000-10,000 IU per day to get there. If the causal evidence for benefit at those levels is weak — and it is — then a large number of people are taking a large dose of a fat-soluble hormone precursor on the basis of associations that do not survive controlled testing.
That matters both for what people are spending and for the possibility, however modest, of harm from sustained supraphysiologic levels. Vitamin D toxicity is real above approximately 150 ng/mL, but the more pressing question is whether the 60-80 ng/mL target has any clinical evidence behind it. The answer, as we will see, is no.
The vitamin D literature is a near-perfect illustration of why observational associations, however consistent and biologically plausible, cannot substitute for randomized controlled evidence.
What the observational evidence said
The observational data on vitamin D and health outcomes is genuinely striking. Theodoratou et al. (2014) published a comprehensive umbrella review in the BMJ — a meta-analysis of meta-analyses — synthesizing the evidence across 137 systematic reviews covering nearly 200 different health outcomes (PMID: 24828554). The associations were extensive: lower vitamin D levels predicted higher risks of cardiovascular disease, multiple cancers, all-cause mortality, type 2 diabetes, multiple sclerosis, and more.
The biological plausibility was also substantial. The vitamin D receptor is expressed in most tissues, including immune cells, cardiac myocytes, and pancreatic beta cells. Vitamin D modulates gene transcription across roughly 5% of the genome. It has direct anti-inflammatory and antiproliferative effects in cell culture. The mechanistic story for why low vitamin D would harm multiple systems was coherent and supported by molecular evidence.
Autier and Gandini (2007) added important nuance in an Archives of Internal Medicine meta-analysis showing that vitamin D supplementation was associated with a statistically significant 7% reduction in all-cause mortality across 18 randomized trials (PMID: 17846391). The effect was primarily driven by studies enrolling participants with low baseline vitamin D levels. This distinction — that supplementation helps deficient people — would prove to be the load-bearing finding when larger trials arrived.
Why the observational evidence misleads
The problem with vitamin D observational research is that low vitamin D is a marker of many things simultaneously. People with lower 25(OH)D levels tend to spend less time outdoors, have more adipose tissue (which sequesters vitamin D), have darker skin (which produces less cutaneous vitamin D per sun exposure), have lower socioeconomic status, have poorer diets, and are more likely to be chronically ill — any of which could explain worse health outcomes without vitamin D being causal.
This is classic confounding by indication — but in reverse. The confounding does not make vitamin D look harmless; it makes low vitamin D look dangerous, because low vitamin D co-travels with a cluster of genuinely risky variables. Randomized controlled trials are designed to break exactly this kind of confounding. And when the trials arrived at scale, the results were instructive.
What the RCTs showed: the VITAL trial
Manson et al. (2019) published the VITAL trial in the New England Journal of Medicine — the most important single piece of evidence in this space (PMID: 30415629). VITAL randomized 25,871 adults (mean age 67, racially diverse sample) to vitamin D3 at 2,000 IU per day versus placebo, with a median follow-up of 5.3 years.
The primary outcomes were cancer incidence and major cardiovascular events. For cancer incidence: no significant difference. Hazard ratio 0.96 (95% CI 0.88-1.06). For major cardiovascular events: no significant difference. Hazard ratio 0.97 (95% CI 0.85-1.12). Subgroup analyses suggested possible benefit in participants with lower baseline vitamin D levels (<20 ng/mL) and possible reduction in cancer mortality (not incidence) — but these were pre-specified secondary analyses, and the primary findings were null.
VITAL is important to understand correctly. It was not a study of vitamin D deficiency treatment — the mean baseline 25(OH)D was approximately 30 ng/mL, which is considered sufficient by most definitions. It was a study of whether adding 2,000 IU/day on top of already-sufficient levels produces clinical benefit in a large, diverse, well-followed population. The answer was no.
The D-HEALTH, ViDA, and DO-HEALTH trials
VITAL was not alone. The D-HEALTH trial (5,000 IU/day in older Australians), the ViDA trial (100,000 IU/month in New Zealand adults), and the DO-HEALTH trial (2,000 IU/day in European adults over 70) all showed no significant benefit on their primary outcomes. Some individual analyses suggested possible benefits in specific subgroups — falls prevention, muscle function — but the overall pattern was consistent with the VITAL finding: supplementation in already-replete populations does not produce the benefits that the observational data predicted.
This is not a story of one negative trial. It is a pattern across multiple adequately-powered trials across multiple populations showing that the observational associations did not survive controlled testing. The most parsimonious explanation is that low vitamin D is a marker of poor health, not a cause of it — at least across the range of deficiency observed in most Western populations.
What the evidence actually supports
The evidence does support several claims about vitamin D, stated precisely.
First: frank vitamin D deficiency — typically defined as 25(OH)D below 20 ng/mL — is associated with real clinical consequences including rickets in children, osteomalacia in adults, secondary hyperparathyroidism, impaired muscle function, and increased fall risk. Correction of deficiency with supplementation is appropriate and supported by both mechanistic and clinical evidence.
Second: the Autier and Gandini mortality finding — that supplementation reduces all-cause mortality in trials enrolling participants with low baseline levels — is consistent with the first point. The benefit is in correcting deficiency, not in optimizing to a high target.
Third: the optimal serum level for most health outcomes is almost certainly in the range of 20-40 ng/mL, not 60-80 ng/mL. There is no high-quality RCT evidence showing incremental benefit from maintaining levels above 40 ng/mL in non-deficient adults. The higher targets circulating in longevity spaces are based on observational associations and theoretical mechanisms, not outcome data.
Fourth: the dose required to maintain 25(OH)D above 20 ng/mL for most people is in the range of 600-2,000 IU per day, depending on baseline level, skin pigmentation, sun exposure, and body composition. Doses of 5,000-10,000 IU per day are not required to achieve sufficiency in most people and push levels into a range where the evidence for benefit is weakest.
If you want to know where you actually stand, getting your 25(OH)D measured is the appropriate first step. Function Health (AFFILIATE_LINK_FUNCTIONHEALTH) includes vitamin D as part of their standard panel, which lets you make supplementation decisions based on your actual level rather than population assumptions.
This Week's One Thing to Do
Test before you supplement. If you are currently taking more than 2,000 IU of vitamin D daily without knowing your 25(OH)D level, get the test. If your level is above 40 ng/mL, consider whether the evidence supports continuing at your current dose. If it is below 20 ng/mL, correction is warranted and supported by the evidence. The key principle is the same one that applies across all of preventive health: measure first, then intervene in proportion to what the measurement shows.
The supplement is cheap. The test is cheap. Using the test to calibrate the supplement is what separates optimization from speculation.
Until next week, Akash S. Chauhan
Education only. Not medical advice. Always consult a licensed clinician for individual decisions.