A new article has just been published in the Journal of Steroid Biochemistry and Molecular Biology looking to see if an oral vitamin D supplement can be used to cure covid-19. Considering that vitamin D is cheap, widely available, and safe, it would be pretty miraculous if that turned out to be the case.
The article describes a randomized controlled trial that was carried out in Spain earlier this year. The trial was open-label (i.e. both the patients and doctors knew who was in which treatment group) and involved 76 patients, of which 50 received 25-hydroxyvitamin D (a.k.a. calcifediol, the activated form of vitamin D) plus standard treatment, while 26 patients just received standard treatment (which at this point in time appears to have consisted mainly of Hydroxychloroquine and azithromycin, two drugs which have since fallen out of favor in the medical community).
Patients needed to fulfill three criteria in order to be accepted in to the study. They had to have a nasal swab PCR test positive for covid-19 and a lung x-ray showing an image consistent with viral pneumonia, and they also had to have at least moderate severity disease according to a scale that doctors use to grade pneumonia called CURB-65.
The patients in the treatment group received an oral capsule containing 0,532 mg of 25-hydroxyvitamin D on day one, followed up by capsules containing 0,266 mg on days three and seven, and thereafter weekly until discharge or admission in to the Intensive Care Unit (ICU). This is a monstrously high dose, enough to get people pretty much instantaneously up from a state of vitamin D deficiency to normal levels in the blood stream.
The patient groups varied a bit in terms of background factors. The control group was slightly sicker to begin with, with 62% having at least one bad prognostic factor, compared with 48% in the treatment group. Obviously this could have skewed the results in favor of vitamin D, since the people in the control group were more prone to have bad outcomes from the start. The overall difference in prognostic factors was due to a higher rate of high blood pressure (58% vs 24%) and a higher rate of diabetes (19% vs 6%) in the control group. The treatment group on the other hand had a larger proportion of participants who were immune suppressed (12% vs 4%).
So, what were the results?
Among the 26 participants who did not receive vitamin D, 13 were admitted to the ICU (50%), and two died (8%). Of the 50 participants who did receive vitamin D, only one individual (2%) was admitted to the ICU, and no-one died.
On the face of it, these results are pretty damn impressive! Vitamin D supplementation resulted in a 48% absolute reduction in risk of ICU admission, and a 100% relative reduction in risk of dying! If that isn’t an effective cure for covid-19, then I don’t know what is.
If these results held up in a larger, higher quality study, that would mean that you could pretty much cure the vast majority of people with covid-19 just by giving them a bolus dose of vitamin D. That would mean that we can pretty much stop researching a vaccine and just give every patient a bolus of vitamin D when they start to develop symptoms.
Of course, that is a pretty big if, as we will now discuss.
The first problem with this study is that it was small, which means that the confidence intervals are wide. For example, although 62% in the control group had at least one bad prognostic factor, compared with only 48% in the treatment group, which to me seems like a pretty big difference, that difference was not statistically significant. How can a difference that big not be statistically significant? Because only 76 people participated in the study. Now, the difference in terms of admission to the ICU was big enough to be highly statistically significant anyway, in spite of the small sample size, but there are other problems with the study.
The second problem is that the primary end point studied was admission to the ICU. Admission to the ICU isn’t a hard outcome like, say, death. It’s a decision made by a doctor about what the most appropriate level of care for a patient is. In general, doctors will place sicker patients in the ICU, and less sick patients will end up in a regular hospital ward, so ICU admission can be a proxy for more severe disease, but that doesn’t necessarily need to be the case. Considering that there wasn’t any blinding, and that the treating doctors knew exactly who was getting what treatment, it would be perfectly possible for the doctors to preferentially put people in the control group in to the ICU, just to make the results look better. Unethical? Yes definitely, but these things have happened. Can we say that happened in this study? No, but we can’t say that it didn’t happen either.
The third big problem is the lack of placebo. Apart from the fact that the doctors knew who was in which group, and might have acted differently towards different patients in ways that affected the results, the fact that the control group knew that they were the control group could also have messed up the results. Maybe the people getting vitamin D thought they would be cured from their covid-19, so they were, whereas the people who didn’t get vitamin D thought they would do badly, and so they did. One should never underestimate the power of the placebo effect to explain differences between groups.
Although there was a 100% reduction in mortality, the small size of the study means that we can’t really say anything about whether vitamin D affects mortality. It would have been enough for one person to die in the vitamin D group, and for one less person to die in the control group, for there not to have been any difference between the groups. That is giving random chance way to much influence over the results.
So, what can we conclude? Can an oral vitamin D supplement cure covid-19?
The results of this study are promising. In fact, they are very promising, and they are in line with earlier evidence that people who are vitamin D deficient get protection from respiratory infections if they take a daily vitamin D supplement.
However, this study reminds me a little too much of the first published trial of hydroxychloroquine as a treatment for covid-19. That trial also had significant flaws, and also showed significant benefit, but the results were not borne out by the bigger, higher quality trials that followed. This trial urgently needs to be followed up by bigger, higher quality double-blind randomized controlled trials that can corroborate the results.
In spite of my reservations, I do think it is reasonable for everyone to take a daily vitamin D supplement based on the results of this study. Vitamin D is after all safe, cheap, and widely available, so the potential benefits far outweigh the potential harms. Personally, I take 3,000 IU (75 ug) of vitamin D per day. If the results of supplementation are only half as good as they appear in this study, that would still be pretty incredible. Right now, vitamin D has a stronger evidence base when it comes to protecting against covid-19 than face masks do.
You might also be interested in my article about whether vitamin D protects against depression or my interview with Sky News about covid-19.