Is vitamin D really linked to excess COVID-19 mortality?

By Dr. Liji Thomas, MDMay 27 2020

The ongoing COVID-19 pandemic has caused many deaths, numbering in the hundreds of thousands, over many different countries. However, the reasons for the differential mortality from region to region are far from clear, despite intensive research.

Some studies show that vitamin D levels are linked to excessive COVID-19 cases and case fatality rates. However, a new paper published on the preprint server medRxiv* in May 2020 shows that COVID-19 infections and mortality are related to life expectancy rather than to vitamin D levels.

Study: Spurious Correlation? A review of the relationship between Vitamin D and Covid-19 infection and mortality. Image Credit: Zerbor / Shutterstock

*Important notice: medRxiv publishes preliminary scientific reports that are not peer-reviewed and, therefore, should not be regarded as conclusive, guide clinical practice/health-related behavior, or treated as established information.

Exploring Possible Solutions

The past five months are unique in human history, in that the global research community has bent its efforts in finding out more about a single disease, its predictors and risk factors, and possible vaccines and therapies. The hint that hydroxychloroquine could be a potentially effective drug against the virus led to a massive run on the drug, more commonly used to alleviate the symptoms of rheumatoid arthritis and other inflammatory conditions of the same type.

Meanwhile, governments sparred over limited international supplies of the drug, and doctors resorted to prescribing it for themselves and their patients, despite the minimal evidence behind the initial recommendation.

Similarly, a recent study highlighting the inverse relationship between vitamin D levels and COVID-19 cases and deaths, using country-level data from 20 European nations. This was shared by 85 news outlets and shared thousands of times via social media. News reports often made it appear that vitamin D could bring down the death rate by 50%. This could lead to potentially fatal overdosing on vitamin D, just like that which happened with the drug hydroxychloroquine.

The truth is that the study reported only a significant correlation and not a cause-effect relationship between the number of cases and deaths due to COVID-19 and vitamin D levels. The problem with such correlations is that they sometimes weaken to insignificant levels or disappear altogether when confounding factors are controlled for.

A Critical Look at Vitamin D-COVID-19 Correlation

In the current study, the investigators controlled for life expectancy while re-examining the data relating to vitamin D and COVID-19 cases and deaths. The result was that while vitamin D is not a significant predictor for infections and deaths due to the pandemic virus, life expectancy is.

Controlled experiments like the randomized controlled clinical trial of recent decades are the cornerstone of scientific knowledge. The reason is that this format controls for other possible explanations, due to what are called confounding factors. This means that any observed correlations are probably the right interpretation.

In social science experiments, controlled trials are rarely possible. This means that correlations must be rigorously tested using adjustments for confounding factors, and confirmed by other types of analysis as well.

Life Expectancy A Potential Confounding Factor

In the current situation, the earlier study on vitamin D and COVID-19 cases/deaths per million population based their conclusions on the presence of a negative correlation between the two. However, there could be confounding factors such as life expectancy, related to both infections/deaths and vitamin D, which might be the true cause of this correlation.

Higher life expectancy in a nation is linked to a significant proportion of aging people, while lower life expectancy is likely to have fewer older people. Since age is a risk factor for symptomatic COVID-19, a nation with a higher life expectancy is likely to have more cases and deaths from the pandemic than one with low life expectancy.

Similarly, vitamin D levels are also linked to life expectancy, since older people are typically less likely to be outdoors for significant periods of time and therefore to receive adequate sunlight. This could lead to vitamin D deficiency.

Thus, life expectancy is linked to both factors and is likely to be a major confounding factor reducing the significance of the correlation. Only by controlling for life expectancy can the real link between vitamin D and COVID-19 infection and case fatality rates be known.

Study Shows Vitamin D Not a Predictor

When the investigators examined the correlation of COVID-19 cases and mortality vs. vitamin D alone, they found the same significant negative correlation. When the same parameter was plotted against life expectancy, they found a significant positive correlation.

Finally, plotting both vitamin D and life expectancy against COVID-19 infections and deaths, the first correlation vanishes, but the second remains significant and positive. This indicates that while vitamin D is NOT a sound predictor of the number of cases, life expectancy is.

Another conclusion could be that countries with high life expectancy could be those with a robust healthcare system, which means a higher case detection. This would also account for the earlier finding. Alternatively, since older people have a higher chance of symptomatic infection, an aged population would mean more people with severe symptoms and, therefore, more people who are likely to be tested and to have positive results as well.

The researchers conclude: “The results indicate that the correlation identified in Ilie and colleagues was most likely spurious in nature.” They offer more plausible and realistic explanations for the apparent association and show that it vanishes when life expectancy is controlled for. Randomized controlled trials must now confirm this, but the results show clearly that vitamin D should not be publicized or considered therapy or preventive measure for COVID-19.

*Important notice: medRxiv publishes preliminary scientific reports that are not peer-reviewed and, therefore, should not be regarded as conclusive, guide clinical practice/health-related behavior, or treated as established information.