Bianconi · Nutrition (Burbank, Los Angeles County, Calif.) 2021 · Observational cohort study · n=250

Prevalence of vitamin D deficiency and its prognostic impact on patients hospitalized with COVID-19.

Cited 23 times in the scientific literature.

Level 3 - non-randomized controlled study

Observational cohort study assessing biomarker prognosis with a comparative control group

PubMed 34388589 · doi:10.1016/j.nut.2021.111408 · record verified 2026-08-30

What was done

Serum 25-hydroxyvitamin D (Vit-D) levels were measured in 200 patients hospitalized with COVID-19 and compared with 50 age- and sex-balanced COVID-19-negative control inpatients with sepsis. Univariable and multivariable Cox regression analyses evaluated associations between baseline Vit-D status (deficiency <20 ng/mL, severe deficiency <12 ng/mL) and a composite primary endpoint of intensive care unit (ICU) admission or in-hospital death.

What was found

Serum Vit-D levels did not differ significantly between COVID-19 patients and sepsis controls (P = 0.397), nor did they differ across COVID-19 severity levels at hospital admission (P = 0.299). Rates of the composite endpoint (ICU admission/in-hospital death) were 31% in Vit-D deficient versus 35% in non-deficient patients (P = 0.649), and 34% in severely deficient versus 30% in non-severely deficient patients (P = 0.593). Vit-D level and deficiency status showed no prospective association with the composite endpoint in any Cox regression model (P > 0.05 for all).

Why it matters

These results indicate that measuring baseline serum vitamin D does not provide useful prognostic stratification for in-hospital clinical deterioration or mortality in patients hospitalized with COVID-19.

Limits

The sample size is relatively small (200 COVID-19 patients and 50 sepsis controls). The control group comprised only inpatients with sepsis rather than healthy controls or general non-COVID admissions, and potential in-hospital vitamin D supplementation or unmeasured confounding variables were not reported in the abstract.

Cited by