*By David S. Seres, MD, ScM, PNS, director of medical nutrition and professor of medicine in the Institute of Human Nutrition, Columbia University Medical Center, New York, NY

Recent ASN blogs explored the complex bidirectional links between the COVID-19 pandemic and food insecurity , and effective nutritional strategies to alleviate the global burden of COVID-19. Some of those nutritional strategies included vitamin D. A different take on vitamin D supplementation and COVID-19 outcomes is shared below.

In April of 2020, when widespread panic about COVID-19 was taking hold, I received a call from a journalist asking: “Did you see that article?”.  He was referring to a pre-review paper that had been posted to the website of a prestigious journal.  It was an analysis of historic vitamin D levels in several countries, to which the authors had correlated one COVID outcome or another.  It made for interesting hypotheses at best.  But the authors concluded that wholesale vitamin D supplementation could improve COVID outcomes.  The story that vitamin D might prevent COVID, based on this paper, appeared on national news two nights later.  It had not been peer reviewed nor accepted for publication, and the journal subsequently rejected the paper.

Since then, the number of hits when searching for “vitamin D and COVID” on PubMed has risen to over 1400.  The vast majority of these are observational studies, systematic review/meta-analysis of observational studies, or narrative reviews.  Because of a strong association between vitamin D levels and COVID outcomes, most promote a potential role for vitamin D supplementation to decrease COVID transmission, morbidity, or mortality, and refer to low levels of vitamin D as “deficiency”.  But the few randomized trials have failed to demonstrate any benefit.

There is little controversy that vitamin D is critical for a plethora of normal immunological functions, that true deficiency is likely to worsen infectious outcomes including those from COVID, and that there are strong associations between vitamin D levels and numerous COVID-related outcomes in observational studies (e.g. retrospective, cohort, etc.). Why then do all the randomized trials to date fail in the aggregate to live up to the great promise that vitamin D supplementation will prevent or ameliorate COVID, except perhaps in those with the very lowest levels? 

The answer lies partly in the inherent risk of drawing conclusions about cause from observational studies.  Equally important, vitamin D physiology easily explains the concordance between vitamin D levels and outcomes in COVID and other disease states such as that seen in critical illness.  That is, vitamin D is a fat-soluble vitamin transported in the bloodstream on a carrier protein.  The level of vitamin D binding protein (VDBP), like albumin, drops in the face of systemic inflammation.  The lower the carrier protein, the lower the measured level of that which is carried.  But is this truly a deficiency? 

To my understanding, a deficiency state occurs when the lack of a substance causes a pathological condition, which in turn is reversed or prevented when the substance is replaced or supplemented.  Disease activity often results in low blood levels of numerous nutrients, but these low levels do not represent true deficiencies.  Take calcium, for example.  Most circulating calcium is carried by albumin. Every clinician is taught to adjust the interpretation of calcium levels based on albumin levels.  A low calcium level, which is due to a low albumin level, which in turn is due to systemic inflammation, is not a deficiency state.  If supplementation given to patients with low vitamin D levels does not improve outcomes, then that low level should not be considered a deficiency.  Thus, a low vitamin D level is not a surrogate for deficiency in someone who is ill.  Rather, it is a surrogate for severity of illness.  Low vitamin D levels are strongly predictive of outcomes, but not predictive of a response to supplementation in sick patients. 

This type of relationship between nutrients and disease acuity is very common but often ignored in our zeal to demonstrate the importance of nutrition.  In my opinion, there are two critically important reasons this should be discussed.  First, continued expenditure for observational studies linking vitamin D to COVID outcomes are unwarranted.  Second, there is continuing damage being done to the credibility of science when scientists and other professionals do not monitor how research findings are discussed with the science consumer. When benefits are touted or even theorized, and then disproven, the public’s trust is betrayed and scientific credibility lost.    

Research published in ASN Journals is providing public health professionals and policy makers with the data and tools needed to mitigate the pandemic’s effect on human health and nutrition in general. We invite you to peruse all four ASN Journals for more information on the bidirectional relationship between COVID-19 and nutrition. A quick search using the term “COVID-19” will lead you to hundreds of articles. In addition, if you are researching any aspect of the relationship between nutrition and COVID-19 disease, we invite you to submit your findings for publication in an ASN Journal.