We all want a miracle cure. We want a silver bullet, a single source of salvation that can save us from everything. This isn’t news, or particularly surprising, but it’s amazing how much this desire permeates everything we say and do when it comes not just to COVID-19, but to every aspect of our lives. It would be so much simpler if there was one cure that could fix everything.
Unfortunately, there almost never is. But that doesn’t stop us from wanting one.
But while this is true of every aspect of our lives, it’s so astonishingly true for COVID-19. From hydroxychloroquine to ivermectin, this year has been filled with nothing but miracle cure after miracle cure, and thus far they’ve all failed*.
Which brings us to vitamin D.
I’ve written before about the hazy promise of vitamin D for COVID-19, months ago, and how way back in early May the evidence just wasn’t there to recommend supplementation for COVID-19. Usually, updating this sort of analysis after such a short space of time would be a bit of a waste, but it turns out that during 2020 evidence moves so fast that even a week or two is enough for a deluge of new data to swamp what came before.
And yet, despite literally hundreds of new studies on the topic since my last article, the message that I have today is almost identical.
If we ask the question “Does vitamin D prevent/treat COVID-19?” the only real answer is “How could you possibly know?”
A recent study on vitamin D and COVID-19 is a brilliant example of this. Hyped in the news and on social media as definitive evidence that we should all be supplementing our vitamin D levels, the study was a simple retrospective epidemiological paper looking at a small group of people who were hospitalized with COVID-19. The authors found that people who had lower vitamin D levels were more likely to experience some negative outcomes of COVID-19, such as low blood oxygen and death, and so recommended that everyone take a vitamin D supplement to stop this from happening.
The problem is, this study had huge, glaring flaws. The sample size was very small — the original database that the authors used had more than 600 people, but they could only analyze 235 of those records because not everyone had a vitamin D test done. The reduction in risk of death for low levels of vitamin D was not, in fact, for everyone, but only when the authors looked at an even smaller subgroup of people aged 40 years and over. Moreover, while vitamin D levels were associated with some issues, they weren’t with other things — low vitamin D didn’t impact people’s risk of ICU admission, chest pain, intubation, kidney damage, and more. And even the statistically significant reduction in the risk of death for over 40’s who had higher vitamin D was a bit meaningless, as you can see in the graph below.
There were also some issues with the statistical analysis, the controls for confounding factors were not robust enough to really make any conclusions, and there appear to be some numerical inconsistencies in the data, but to a great extent that’s not important. Even without other issues, doing this sort of analysis on a tiny, uncontrolled sample can at best give us a vague association between low vitamin D and COVID-19 outcomes, it certainly doesn’t prove that supplementation will do anything.
The headlines were misleading, but that’s really not all that surprising.
Now, that’s just one study, but these sort of issues crop up time and again when you look at the vitamin D literature. A very large study that generated massive hype about vitamin D had issues with the statistical analysis — the researchers dichotomized all of their continuous variables and failed to control for a plethora of confounders — which means that it’s hard to draw much meaning from the research. Another Mendelian randomization study preprinted recently had some similar issues with confounding variables, so while the authors of that paper found no benefit for vitamin D we still don’t have a conclusive answer. The first randomized-controlled trial on the topic, published to great fanfare, in fact had numerous flaws which was not at all surprising because it was a tiny pilot study including just 76 participants.
In other words, the research is just…bad. Inconclusive. Despite months of research, and hundreds of papers out already, the most we can say about COVID-19 and vitamin D supplementation amounts to a huge shrug of the shoulders and a lengthy sigh.
To be honest, it’s all a bit depressing. We’ve had more than enough time to run at least one decent study, and so far all we’ve got are an increasing roster of terrible epidemiological research and one tiny pilot study that itself was at a high risk of bias.
Based on the research out at the moment, depending on which studies you cherry-pick to support your opinion, it’s possible to say anything from “vitamin D is a total waste of time for COVID-19” to “vitamin D supplements will save us all from imminent death!”. It’s the same situation that we face when talking about supplementation for many things, because often the research on supplements just isn’t very good, but given the stakes it’s quite impressively absurd that this is where we are.
Which brings us to the choice. Should you take a vitamin D supplement or not?
Now, this is actually a different question to the one I was answering before. I think, based on the current evidence, that the only possible way to answer the question “does vitamin D prevent/treat COVID-19?” is to say that there’s no way you could know right now, but that based on previous evidence from other respiratory diseases it seems that the benefit, if any, is likely to be pretty small.
But when you’re facing the choice to supplement vitamin D, it’s actually not quite the same calculus. People are arguing that even though the evidence for vitamin D is terrible, it doesn’t matter. If there is some evidence for benefit, well, vitamin D isn’t terribly dangerous (the side-effect risk for supplements is very low), and so there are few downsides and a very large upside if it turns out to work.
In other words, the argument goes, it probably won’t hurt, and it might help, so take the supplements.
This is, unfortunately, a choice that I can’t really advise you on. While I would argue that people are underselling the risks — no medical intervention is entirely risk-free, and spending money to take a supplement every day strikes me as requiring at least some effort — it’s also not entirely untrue. If everyone everywhere starts taking vitamin D, some people will be harmed, but there remains the possibility that some will be helped. It’s possible that a large number of people will be helped — we simply don’t know.
Which brings us to the advice-giving part of the article. The place where most people immediately scroll down to so that they can see the answer without understanding the background behind it (hello, shirkers). But what I’ve got to say in terms of advice really isn’t that interesting so you might as well read the whole piece — talk to your doctor about your health, and leave it at that. No one really knows right now if vitamin D supplementation does anything for COVID-19, despite the claims of many headlines, but an expert who knows your medical history is in the best place to tell you whether a supplement is the correct choice for you right now.
Ultimately, we don’t have all the answers. Science is never perfect, but in the case of vitamin D and COVID-19 it’s substantially less perfect than usual.
We have no idea whether vitamin D supplementation is an effective treatment or preventative for COVID-19. Hopefully that will change soon (and I can update this article), but for now the only thing I can definitively say is that anyone telling you that vitamin D supplements do or don’t work for COVID-19 with any certainty is probably wrong.
*Note: Yes, dexamethasone is the one treatment that has decidedly NOT failed, but the thing there is that no one ever painted a corticosteroid which is only useful for people with severe disease as a magical cure-all, so the hype isn’t quite the same.