Participants positive for COVID-19 were 50% more likely to have low vs normal 25(OH)D levels in a multivariate analysis that controlled for other confounders, for example.
The take home message for physicians is to “test patients’ vitamin D levels and keep them optimal for the overall health – as well as for a better immunoresponse to COVID-19,” senior author Milana Frenkel-Morgenstern, PhD, head of the Cancer Genomics and BioComputing of Complex Diseases Lab at Bar-Ilan University in Ramat Gan, Israel, said in an interview.
The study was published online July 23 in The FEBS Journal.
Previous and ongoing studies are evaluating a potential role for vitamin D to prevent or minimize the severity of SARS-CoV-2 infection, building on years of research addressing vitamin D for other viral respiratory infections. The evidence to date regarding COVID-19, primarily observational studies, has yielded mixed results.
Multiple experts weighed in on the controversy in a previous report. Many point out the limitations of observational data, particularly when it comes to ruling out other factors that could affect the severity of COVID-19 infection. In addition, in a video report, JoAnn E. Manson, MD, DrPH, of Harvard Medical School in Boston, cited an observational study from three South Asian hospitals that found more severe COVID-19 patients had lower vitamin D levels, as well as other “compelling evidence” suggesting an association.
Dr. Frenkel-Morgenstern and colleagues studied data for 7,807 people, of whom 10.1% were COVID-19 positive. They assessed electronic health records for demographics, potential confounders, and outcomes between February 1 and April 30.
Participants positive for COVID-19 tended to be younger and were more likely to be men and live in a lower socioeconomic area, compared with the participants who were negative for COVID-19, in a univariate analysis.
A higher proportion of COVID-19–positive patients had low plasma 25(OH)D concentrations, about 90% versus 85% of participants who were negative for COVID-19. The difference was statistically significant (P < .001). Furthermore, the increased likelihood for low vitamin D levels among those positive for COVID-19 held in a multivariate analysis that controlled for demographics and psychiatric and somatic disorders (adjusted odds ratio, 1.50). The difference remained statistically significant (P < .001).
The study also was noteworthy for what it did not find among participants with COVID-19. For example, the prevalence of dementia, cardiovascular disease, chronic lung disorders, and hypertension were significantly higher among the COVID-19 negative participants.
“Severe social contacts restrictions that were imposed on all the population and were even more emphasized in this highly vulnerable population” could explain these findings, the researchers noted.
“We assume that following the Israeli Ministry of Health instructions, patients with chronic medical conditions significantly reduced their social contacts” and thereby reduced their infection risk.
In contrast to previous reports, obesity was not a significant factor associated with increased likelihood for COVID-19 infection or hospitalization in the current study.
The researchers also linked low plasma 25(OH)D level to an increased likelihood of hospitalization for COVID-19 infection (crude OR, 2.09; P < .05).
After controlling for demographics and chronic disorders, the aOR decreased to 1.95 (P = .061) in a multivariate analysis. The only factor that remained statistically significant for hospitalization was age over 50 years (aOR, 2.71; P < .001).
Implications and future plans
The large number of participants and the “real world,” population-based design are strengths of the study. Considering potential confounders is another strength, the researchers noted. The retrospective database design was a limitation.
Going forward, Dr. Frenkel-Morgenstern and colleagues will “try to decipher the potential role of vitamin D in prevention and/or treatment of COVID-19” through three additional studies, she said. Also, they would like to conduct a meta-analysis to combine data from different countries to further explore the potential role of vitamin D in COVID-19.
“A compelling case”
“This is a strong study – large, adjusted for confounders, consistent with the biology and other clinical studies of vitamin D, infections, and COVID-19,” Wayne Jonas, MD, a practicing family physician and executive director of Samueli Integrative Health Programs, said in an interview.
Because the research was retrospective and observational, a causative link between vitamin D levels and COVID-19 risk cannot be interpreted from the findings. “That would need a prospective, randomized study,” said Dr. Jonas, who was not involved with the current study.
However, “the study makes a compelling case for possibly screening vitamin D levels for judging risk of COVID infection and hospitalization,” Dr. Jonas said, “and the compelling need for a large, randomized vitamin D supplement study to see if it can help prevent infection.”
“Given that vitamin D is largely safe, such a study could be done quickly and on healthy people with minimal risk for harm,” he added.
More confounders likely?
“I think the study is of interest,” Naveed Sattar, PhD, professor of metabolic medicine at the University of Glasgow, who also was not affiliated with the research, said in an interview.
“Whilst the authors adjusted for some confounders, there is a strong potential for residual confounding,” said Dr. Sattar, a coauthor of a UK Biobank study that did not find an association between vitamin D stages and COVID-19 infection in multivariate models.
For example, Dr. Sattar said, “Robust adjustment for social class is important since both Vitamin D levels and COVID-19 severity are both strongly associated with social class.” Further, it remains unknown when and what time of year the vitamin D concentrations were measured in the current study.
“In the end, only a robust randomized trial can tell us whether vitamin D supplementation helps lessen COVID-19 severity,” Dr. Sattar added. “I am not hopeful we will find this is the case – but I am glad some such trials are [ongoing].”
Dr. Frenkel-Morgenstern received a COVID-19 Data Sciences Institute grant to support this work. Dr. Frenkel-Morgenstern, Dr. Jonas, and Dr. Sattar have disclosed no relevant financial relationships.
A version of this article originally appeared on.