Vitamin D and COVID: New Research Suggests More Is Not Always Better

Vitamin D became one of the most debated supplements during the COVID pandemic. Depending on who you listened to, vitamin D was either an overlooked tool for supporting immune health or another supplement being promoted far beyond what the evidence could actually support.

A new randomized clinical trial gives us a much better opportunity to separate those two extremes.

Researchers compared 400 IU, 4,000 IU, and 10,000 IU of vitamin D per day and followed participants for up to nine months to see whether higher doses reduced the risk of COVID infection. The headline result was technically negative. When the 4,000 and 10,000 IU groups were combined, higher dose vitamin D did not significantly reduce COVID risk.

But when you look more closely at the individual doses, things get considerably more interesting.

The 4,000 IU dose was associated with a significantly lower risk of COVID compared with 400 IU, while 10,000 IU provided no apparent benefit.

That doesn't prove vitamin D prevents COVID. It does, however, raise an interesting question that applies to supplements well beyond vitamin D:

Is there an optimal dose where enough helps, but more doesn't necessarily help more?

TLDR

This was a reasonably large, double blind randomized trial, which makes it more useful than the observational studies that drove many of the early vitamin D and COVID headlines. The overall primary outcome was negative, so we shouldn't turn this into "vitamin D prevents COVID," but the dose specific results deserve attention.

• Participants taking 4,000 IU/day had a 38% lower adjusted risk of COVID compared with those taking 400 IU/day.
• COVID occurred in 10.0% of the 4,000 IU group versus 16.0% of its 400 IU comparison group.
• 10,000 IU/day showed no benefit, with COVID occurring in 12.4% of both the high and low dose groups.
• Combining the 4,000 and 10,000 IU groups did not produce a statistically significant reduction in the study's primary analysis.
• Higher blood levels of 25(OH)D after supplementation did not independently predict lower COVID risk.
• The study found no significant difference in reported adverse effects between groups.
• The most interesting takeaway is not that everyone should take 4,000 IU of vitamin D. It is that dose matters, and more is not automatically better.

Why Vitamin D Was So Interesting During COVID

Vitamin D isn't just involved in bone health. It also plays a role in both innate and adaptive immune function, which made it an obvious nutrient to investigate when COVID appeared.

Early observational research found associations between lower vitamin D status and greater COVID risk. Other research suggested vitamin D supplementation might reduce respiratory infections more generally, particularly when vitamin D was taken daily. But observational studies have a major limitation: an association doesn't necessarily tell us which direction causation runs.

Someone with higher vitamin D levels might spend more time outside, exercise more, eat differently, have better overall health, or differ in dozens of other ways from someone with low vitamin D.

Randomized controlled trials help eliminate many of those problems.

And that's what makes this new study interesting.

What Did the Researchers Do?

The study was a double blind, randomized controlled trial conducted through the University of Chicago Medicine and Rush University Medical Center. Researchers ultimately randomized 955 participants, with 923 receiving study vitamin D and continuing into the analysis. Participants were followed for as long as nine months.

The design was slightly unusual because participants first chose which dosing comparison they were comfortable entering. They could enter a branch comparing 4,000 IU with 400 IU per day, or a branch comparing 10,000 IU with 400 IU per day. They were then randomly assigned to one of the two doses within that branch.

This distinction matters because the study was not simply a three arm trial in which everyone was randomly assigned to 400, 4,000, or 10,000 IU. The researchers accounted for study branch and other variables in their statistical analysis.

The primary outcome was straightforward: How long did it take before someone developed COVID?

First, The Primary Result Was Negative

This is important because it would be very easy to cherry pick this study.

When researchers combined everyone receiving either 4,000 or 10,000 IU and compared them with participants receiving 400 IU, COVID occurred in:

10.9% of the moderate or high dose participants

versus

14.7% of the low dose participants.

That looks like a meaningful difference. After statistical adjustment, however, the hazard ratio was 0.73 with a 95% confidence interval of 0.51 to 1.06 and a P value of 0.10.

In other words, the combined higher dose intervention did not produce a statistically significant reduction in COVID risk for the primary outcome.

That needs to be the starting point for interpreting everything else.

This study did not demonstrate that simply taking more vitamin D prevents COVID.

But 4,000 IU Told a Different Story

Things became much more interesting when researchers looked at the two doses separately in prespecified secondary analyses.

Among participants in the moderate dose branch, COVID occurred in 10.0% of those receiving 4,000 IU compared with 16.0% receiving 400 IU. After adjustment, the hazard ratio was 0.62, with a 95% confidence interval of 0.39 to 0.99 and a P value of 0.045.

A hazard ratio of 0.62 corresponds to approximately a 38% lower relative hazard during follow up.

The absolute difference is also worth looking at.

Sixteen percent of the low dose group developed COVID compared with 10 percent of the moderate dose group, an absolute difference of about 6 percentage points.

The Kaplan Meier curves shown in the paper make the difference particularly easy to see. On page 16, the 4,000 IU and 400 IU groups gradually separate over the course of follow up, while the curves for 10,000 IU and 400 IU remain almost directly on top of one another.

That's a fascinating result.

But there is a statistical caveat.

The researchers explicitly point out that the 4,000 IU result came from a secondary analysis and would no longer be statistically significant after correcting for multiple comparisons.

So I would describe this as promising rather than definitive.

Then 10,000 IU Did Basically Nothing

If vitamin D's effect were simply dose dependent, you might expect 10,000 IU to outperform 4,000 IU.

It didn't.

COVID occurred in 12.4% of participants receiving 10,000 IU and 12.4% receiving 400 IU in that branch.

The adjusted hazard ratio was 0.99.

That's about as close to "no difference" as you can get.

This is arguably the most interesting finding in the paper.

A moderate dose appeared beneficial.

A much higher dose didn't.

We shouldn't conclude from one study that 10,000 IU of vitamin D somehow increases COVID risk or that 4,000 IU is a magical immune dose. The confidence interval around the high dose estimate was also quite wide, meaning there was considerable statistical uncertainty.

But the results certainly don't support the idea that if some vitamin D is good, more must be better.

More Vitamin D Produced Higher Blood Levels, But That Didn't Predict Lower Risk

Here's where the study gets even stranger.

Researchers measured participants' blood levels of 25 hydroxyvitamin D, commonly written as 25(OH)D.

At baseline, average levels were already around 34 to 37 ng/mL, depending on the study branch. After supplementation, levels remained relatively stable in the 400 IU groups but rose to approximately 50 ng/mL with 4,000 IU and 70 ng/mL with 10,000 IU.

So the supplements clearly did what we would expect them to do.

More vitamin D increased measured vitamin D status.

But higher post supplementation 25(OH)D levels themselves did not predict lower COVID risk.

That's interesting because much of the vitamin D conversation focuses on achieving a particular blood level.

The researchers raise the possibility that the daily dose itself could matter independently of the measured total 25(OH)D concentration. They discuss what is known as the free vitamin D hypothesis, where vitamin D not bound to vitamin D binding proteins may be particularly relevant to immune cells.

That's a mechanistic hypothesis, not something this trial proved.

But it could help explain why simply pushing serum vitamin D higher doesn't necessarily translate into progressively greater immune benefits.

Why Might 10,000 IU Not Have Worked Better?

The researchers discuss another interesting possibility.

At sufficiently high vitamin D intakes, the body may increase mechanisms that degrade or regulate vitamin D metabolites. In other words, our biology isn't necessarily a passive container where twice the vitamin D produces twice the physiological effect.

The authors speculate that this could partially explain why daily moderate dosing performed differently from the very high dose in this study.

This is one of those areas where I think supplement marketing often gets ahead of physiology.

People naturally assume dose response curves are linear.

If 1,000 IU helps, 2,000 must help twice as much.

If 4,000 is better, 10,000 must be even better.

Biology rarely works that neatly.

Many nutrients have an effective range. Once you have enough, adding more may provide little additional benefit and can eventually create problems.

What About People Who Were Vitamin D Deficient?

This study becomes even more interesting when you look at the participants' starting vitamin D levels.

Average baseline 25(OH)D was already around 34 to 37 ng/mL. Only 38.8% of participants were below 30 ng/mL, and just 13.6% were below 20 ng/mL.

Compare that with an earlier blinded randomized trial discussed by the authors. In that trial, 93.8% of participants had vitamin D levels below 30 ng/mL, and 4,000 IU/day produced a substantial reduction in COVID incidence compared with placebo.

That raises an important possibility.

Vitamin D supplementation may matter more when someone doesn't have enough vitamin D in the first place.

That's not unique to vitamin D. It is one of the most important principles in nutrition.

Correcting a deficiency can have a large effect.

Taking progressively more of something after your needs are already met often doesn't.

There Were Some Interesting Subgroup Findings Too

The researchers explored several subgroups, although these analyses need to be interpreted cautiously because smaller groups create greater statistical uncertainty and multiple comparisons increase the chance of finding something simply by chance.

One notable finding was that the pooled moderate and high doses were associated with lower COVID risk among men. The 4,000 IU dose also appeared beneficial among participants who were fully immunized at baseline and among those who could have participated after Omicron arrived.

These results are interesting enough to justify additional research, but I wouldn't use them to create different vitamin D recommendations for men and women or vaccinated and unvaccinated people based on this study alone.

This is exactly where interesting science can turn into bad supplement marketing if we're not careful.

What About Safety?

Another question people will naturally have is whether 4,000 or 10,000 IU per day caused more adverse effects.

In this study, researchers found no significant differences in participant reported adverse effects between the treatment groups. Fourteen participants were withdrawn following post randomization laboratory testing, but withdrawal rates did not differ between groups.

That doesn't mean 10,000 IU per day should automatically be considered appropriate for everyone indefinitely.

The trial specifically excluded people at increased risk from vitamin D supplementation, including individuals with certain kidney, calcium, parathyroid, gastrointestinal, and other conditions.

There's a big difference between saying a dose was tolerated in a controlled study and recommending everyone take that amount forever.

The Study Had Some Important Limitations

This wasn't a perfect trial.

The researchers originally planned to enroll 2,000 participants but ended up randomizing 955. Lower than expected enrollment, lower COVID incidence, and incomplete follow up reduced the study's statistical power. The authors also acknowledge that the sample size and number of statistical comparisons limited their ability to confidently identify differences between subgroups.

The study also took place over a particularly complicated period of the pandemic. Participants were recruited between late 2020 and 2023, meaning circulating variants, previous infections, vaccination status, boosters, and baseline infection risk were changing dramatically during the trial.

Most participants were also highly immunized, and the results may not generalize to different variants, geographic regions, longer supplementation periods, or other respiratory infections.

And perhaps most importantly, the statistically significant 4,000 IU finding was a secondary outcome.

That's why the correct interpretation isn't:

"Scientists prove 4,000 IU of vitamin D prevents COVID."

They didn't.

So Does Vitamin D Prevent COVID?

Based on this study alone, I would say we still don't know.

The primary analysis did not find a statistically significant reduction in COVID when the moderate and high dose groups were pooled.

At the same time, dismissing the study as simply "vitamin D doesn't work" would also miss something interesting.

The 4,000 IU group had a lower COVID incidence than its 400 IU comparison group, with a 38% lower adjusted hazard. The 10,000 IU group showed essentially no difference at all.

Previous research has also been mixed. Some randomized trials and meta analyses have suggested vitamin D supplementation can reduce respiratory infections, while other large trials have found little or no benefit. The authors note that differences in baseline vitamin D status, dose, daily versus intermittent supplementation, vaccination, variants, and population characteristics could help explain some of those inconsistencies.

So we're left with a more nuanced answer.

Which, unfortunately, doesn't make for a very good social media headline.

What I Would Actually Take From This Study

I wouldn't start taking vitamin D specifically because you're worried about COVID.

I also wouldn't look at these results and decide everyone needs 4,000 IU per day.

What this study does reinforce is the importance of getting enough vitamin D and the possibility that dose matters.

If you're deficient, correcting that deficiency makes sense for reasons extending far beyond COVID. If your vitamin D status is already adequate, taking progressively larger doses isn't automatically going to produce progressively greater benefits.

And if you're taking substantial amounts of vitamin D long term, it makes much more sense to periodically measure your vitamin D status and discuss the appropriate dose with your healthcare provider than to blindly assume more is better.

That's a lot less exciting than calling vitamin D a COVID cure.

It's also much closer to what the research actually says.

The Bottom Line

This study is a good example of why I try not to look at supplements as simply working or not working.

Dose matters. Baseline nutritional status matters. The population matters. The outcome being measured matters.

In this randomized trial, 4,000 IU of vitamin D per day was associated with lower COVID risk compared with 400 IU, while 10,000 IU showed no apparent advantage at all. But the study's primary combined analysis was not statistically significant, and the promising 4,000 IU result came from a secondary analysis that becomes less convincing when accounting for multiple comparisons.

So no, this study doesn't prove vitamin D prevents COVID.

But it does add another piece to a much bigger body of research suggesting that vitamin D plays an important role in immune function and that getting enough may matter.

Just don't make the mistake of assuming that if enough is good, a lot more must be better.

Reference

Meltzer DO, Moy JN, Schram AW, et al. Effect of moderate or high versus low dose daily vitamin D on COVID-19 incidence: a double-blind, parallel-group randomized clinical trial. Clinical Nutrition. 2026;64:106722.