By Philip R. Fischer, MD, DTM&H

Professor of Pediatrics, Department of Pediatric and Adolescent Medicine, Mayo Clinic, Rochester, MN; Department
of Pediatrics, Sheikh Shakhbout Medical City, Abu Dhabi, United Arab Emirates

SYNOPSIS: Low levels of vitamin D are associated with in-hospital mortality in patients with COVID-19, but causality is not yet known.

SOURCE: Angelidi AM, Belanger MJ, Lorinsky MK, et al. Vitamin D status is associated with in-hospital mortality and mechanical ventilation: A cohort of COVID-19 hospitalized patients. Mayo Clin Proc 2021;96:875-886.

Vitamin D has immunomodulatory properties as well as anti-inflammatory activity. In fact, vitamin D deficiency has been associated with both an elevated risk of acute respiratory infection and worse clinical outcomes following critical illnesses. Vitamin D deficiency also is associated with problems, such as obesity, older age, and cardiac disease, that are risk factors for bad outcomes with COVID-19. Thus, investigators have wondered if vitamin D deficiency (and potential treatment) might influence the clinical course of COVID-19.

Angelidi and colleagues performed a retrospective study of adults who were hospitalized at one of two hospitals (one in Boston and one in New York) with COVID-19 from February to mid-May 2020. They reviewed records and compared patient data to 25-hydroxyvitamin D levels determined either at the time of hospital admission or within the preceding six months.

A total of 144 patients were included in the study: 79 with vitamin D levels of less than 30 ng/mL and 65 with levels of 30 ng/mL or higher. The median age was 66 years. Overall, 44% of subjects were male, and 42% were non-Hispanic Blacks. The median body mass index was 29. More than 90% of included individuals had at least one significant medical comorbidity, with hypertension (74%), hyperlipidemia (55%), and diabetes (44%) being especially common. Cough, dyspnea, fever, and/or malaise were presenting symptoms in most of the patients. Steroids were used in the management of 24% of patients, antivirals were used in 10%, an antibiotic (usually azithromycin) was used in 72%, and hydroxychloroquine was used in 44%. Treatment included oxygen in 64% of patients and mechanical ventilation in 27%; 39% of patients required intensive care. In-hospital mortality was 18%.

Mortality was higher (25% vs. 9%) in patients with low (< 30 ng/mL) vs. higher 25-hydroxyvitamin D levels. The timing (during the six months prior to admission vs. during the hospitalization) of vitamin D testing was not related to mortality. Of dozens of variables, only vitamin D level, age, malignancy, and chronic obstructive pulmonary disease were associated with an increased risk of in-hospital death with COVID-19. Doing careful statistical analysis, researchers found hypovitaminosis D was strongly associated with in-hospital mortality, even independent of medical comorbidities. The inverse association between vitamin D level and mortality was present whether 20 ng/mL or 30 ng/mL was used as the cut-off (thus, whether there was vitamin D deficiency or insufficiency).


Angelidi and colleagues carefully and convincingly showed that low vitamin D levels are associated with mortality in patients hospitalized with COVID-19. Of course, this association does not necessarily imply causality, and it does not prove that either preventive or therapeutic vitamin D administration would alter mortality.

Low vitamin D levels have previously been associated with other factors that give risk for poor outcomes with COVID-19, including obesity and diabetes. However, in Angelidi’s careful analysis, low vitamin D levels were independently associated with in-hospital mortality from COVID-19. There likely is either a causal impact of hypovitaminosis on the course of COVID-19 or there are other unmeasured variables, such as outdoor activity, that link both hypovitaminosis D and death from COVID without the vitamin D level directly affecting the course of COVID-19.

Hypovitaminosis D does seem causally related to other respiratory infections, even if a causal link has yet to be proven for COVID-19. Low vitamin D levels are seen more commonly in patients with acute respiratory infection than in healthy controls, and vitamin D does have an effect on immune functioning.1 However, studies of vitamin D supplementation to prevent respiratory infections have yielded mixed results.1 A new meta-analysis of studies of vitamin D as prevention for acute respiratory infection in children aged 1 to 15 years showed a significant (P = 0.018) but modest (odds ratio, 0.92, with 95% confidence interval, 0.86 to 0.99) effect when supplements of 400 IU/day to 1,000 IU/day were administered for up to 12 months.1

It is important to consider the timing of effects of intervention. Whether vitamin D provides protection against acquiring SARS-CoV-2 or other respiratory pathogens, different mechanisms of action could be necessary for vitamin D to be effective therapeutically. Griffin and colleagues recently summarized the various stages of COVID-19 and eloquently reviewed potential effects of various interventions at various times before, during, and after the actual infection.2

A recent placebo-controlled study of high-dose vitamin D as treatment of established COVID-19 infection included 236 hospitalized adults in multiple centers in Brazil (mean age 56 years, mean 25-hydroxyvitamin D level 21 ng/mL at entry into the study, with 20 ng/mL being the upper limit of “deficiency”).3 Vitamin D levels increased significantly with treatment, and no significant adverse events were noted.3 However, hospital length-of-stay, need for intensive care, need for mechanical ventilation, and mortality were not altered by vitamin D treatment.3

Thus, these new data remind us that hypovitaminosis D is at least associated with respiratory infections, but that preventive supplementation only modestly reduces the risk of acquiring infection (in children for non-COVID-19 infection), and therapeutic administration of vitamin D does not alter the course of adults being hospitalized with COVID-19. Vitamin D generally is safe in the preventive and therapeutic doses used, but further convincing data will be required before vitamin D is recommended to either prevent or treat COVID-19.

Two years ago, Hu and colleagues reported that patients with chronic hepatitis B had lower vitamin D levels than did healthy controls, and among hepatitis B patients, viral loads were inversely correlated with vitamin D level.4 Vitamin D also has been proposed for the prevention and treatment of a variety of other conditions, including diabetes, multiple sclerosis, and cognitive decline. Observational studies are supportive, but systematic reviews and randomized controlled trials are lacking.5

Although it seems reasonable to supplement individuals with or at risk of hypovitaminosis D to maintain a “normal” vitamin D level,6 definitive studies do not yet support widespread recommendations to use vitamin D specifically to prevent or treat these other conditions.5 


  1. Jolliffe DA, Camargo CA Jr, Sluyter JD, et al. Vitamin D supplementation to prevent acute respiratory infections: A systematic review and meta-analysis of aggregate data from randomized controlled trials. Lancet Diabetes Endocrinol 2021; Mar 30. doi 10.1016/S2213-8587(21)00051-6.
  2. Griffin DO, Brennan-Rieder D, Ngo B, et al. The importance of understanding the stages of COVID-19 in treatment and trials. AIDS Rev 2021;22:40-47.
  3. Murai IH, Fernandes AL, Sales LP, et al. Effect of a single high dose of vitamin D3 on hospital length of stay in patients with moderate to severe COVID-19: A randomized clinical trial. JAMA 2021;325:1053-1060.
  4. Hu Y-C, Wang W-W, Jiang W-Y, et al. Low vitamin D levels are associated with high viral loads in patients with chronic hepatitis B: A systematic review and meta-analysis. BMC Gastroenterol 2019;19:84.
  5. Maretzke F, Bechthold A, Egert S, et al. Role of vitamin D in preventing and treating selected extraskeletal diseases – an umbrella review. Nutrients 2020;12:969.
  6. Thacher TD. Vitamin D and COVID-19. Mayo Clin Proc 2021;96:838-840.