Coronavirus and Vitamin D: Correlation, Causation, and Confusion

The “lifestyle vitamin” has burned us before. Is Vitamin D supplementation smart in COVID-19?

I have a bit of a love-hate relationship with Vitamin D.

A brief search of my blog has shown me questioning the link between Vitamin D and multiple sclerosis, kidney disease, schizophrenia, falls, and childhood educational attainment.

So it’s no surprise, in the COVID-era, that Vitamin D comes up again. I was intrigued when Dr. Anthony Fauci, a sober and responsible voice if ever there was one during this pandemic reported his personal use of Vitamin D. Vitamin D was also part of the presidential cocktail that Trump received during his stay at Walter Reed.

I decided to dig into the data here, but before we do I want to tell you why I am inherently skeptical of Vitamin D studies. Two main issues:

1) Low vitamin D levels have been linked to so many things. Like, Vitamin C deficiency gets scurvy — fine, but low Vitamin D has been linked to everything from Alzheimer’s to Whooping Cough. It’s either the most important Vitamin in the world, or it’s a stand-in for some other important thing.

The second reason I’m a bit of a vitamin D skeptic is because when we’ve tested all these intriguing links via randomized trials, giving some people Vitamin D and some placebo, they almost always show no effect.

Me thinking vitamin D will help some condition

I feel a bit like Charlie Brown with the football here. Vitamin D has burned us before. A couple of examples.

Multiple observational studies found that people with low levels of Vitamin D were more likely to develop cancer and cardiovascular disease. These were good studies, adjusting for appropriate confounders, the whole deal.

Randomized trial of vitamin D supplementation looking at cancer and cardiovascular outcomes. Nothing.

Then we got a 26,000 patient randomized clinical trial. No effect on either outcome.

Randomized trial of vitamin D supplementation looking at diabetes incidence. Nada.

Observational studies suggested that low vitamin D levels were linked to the development of diabetes. Then we got a trial of 2400 patients with pre-diabetes. Vitamin D supplementation had no effect.

Observational studies have found that low vitamin D levels are linked to falls. The randomized trial found no such effect. (Exercise did prevent falls though).

Randomized trial of vitamin D supplementation looking at all cause mortality. No significant effect.

Does Vitamin D save lives in general? Not really. The Women’s Health Initiative Trial randomized 36,000 post-menopausal women to vitamin D versus placebo and saw no survival benefit.

Look, I’m not saying Vitamin D doesn’t have it’s uses, or that it may have some limited benefit. Randomized trials are rarely as dramatic as observational data promises. But I can’t think of another exposure which has SUCH a wide disconnect between observational and randomized data.

Why is that?

Because your Vitamin D level is a marker of the type of life you live, and it’s really, really hard to adjust for that. Sure, you can ask people how much sun and exercise they get, how varied their diet is, how much money they make — but these are pretty blunt tools.

OK keeping all of that in mind, what do we know about Vitamin D in Covid?

Just this week we got this study out of Spain that found that, of 216 hospitalized COVID-19 patients, 80% were vitamin D deficient. Only 40% of controls were deficient. Of course, the controls weren’t hospitalized. Or sick actually.

But ok better study. Back in September which, as of this recording, was like 10 years ago, we got this study out of Chicago showing that, of 489 patients with prior Vitamin-D testing, those who were Vitamin D deficient had a 1.77 times higher risk of contracting COVID-19.

This study in PLOS One found, among 235 hospitalized patients, those with Vitamin D levels less than 30 had a 20% death rate compared to 9.7% in those with higher levels. I should point out that this article was flagged with an “expression of concern” for some data weirdness.

Ok so we have multiple observational studies, and Anthony Fauci for crying out loud, saying we should consider Vitamin D.

And why not, really? It’s cheap. It’s got a pretty benign side effect profile (though the nephrologist in me worries about kidney stones).

But when it comes to COVID, multiple factors associated with low Vitamin D levels are also associated with COVID-19 infection and severity: older age, BMI, Black race, poverty, comorbidities. It’s just really hard to appropriately account for all of this stuff statistically.

So — I’m gun shy on Vitamin D. I’d love a randomized trial. Right now we have one, two if you squint.

The one randomized trial is this one, out of Spain.

A small study, 76 patients hospitalized with COVID-19 were randomized in a 2:1 Ratio to Vitamin D + Azithromycin + Hydroxychloroquine (don’t get me started) versus Azithromycin and Hydroxychloroquine alone. Of the 50 people in the Vitamin D group, just 1 went to the ICU. Compare that to 13 of the 26 in the non-Vitamin D group.

There were zero deaths in the vitamin D group and 2 in the non-Vitamin D group. I should note that the non-D group here was a bit higher risk at baseline, with more individuals with diabetes and hypertension. But honestly, it’s the effect size that surprises me here. I mean, reducing the risk of ICU admission from 50% to 2%? That’s just a staggering amount. So we’re in the space of — is this a miracle drug or is there a problem with the study we don’t know about. And miracles are in short supply these days.

Just to throw some cold water on that study, we have this one — still in pre-print, which used mendelian randomization to look at whether Vitamin D might have a causal link with COVID severity in over 400,000 individuals.

In this study design, you compare people who are genetically predisposed to lower Vitamin D levels to those who are not. That takes lifestyle stuff out of the equation. Interestingly, people genetically destined to have low vitamin D levels were less likely to be hospitalized with COVID-19 or to develop severe disease. Not peer-reviewed yet, grain of salt, etc.

Bottom line? We know that people with low Vitamin D are at higher risk of bad outcomes in COVID-19. We still don’t know if Vitamin D supplementation will change that risk. There is some promise here, but the lifestyle vitamin has burned us before, so proceed with 400 international units of caution.

A version of this commentary first appeared on

Writing about medicine, science, statistics, and the abuses thereof. Commentator at Medscape. Associate Professor of Medicine at Yale University.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store