Word Count: 350 Manuscript Word Count: 2967Importance: Vitamin D treatment has been found to decrease incidence of viral respiratory tract infection, especially in vitamin D deficiency. It is unknown whether COVID-19 incidence is associated with vitamin D deficiency and treatment.Objective: To examine whether vitamin D deficiency and treatment are associated with testing positive for COVID-19.
Design: Retrospective cohort studySetting: University of Chicago Medicine Participants: Patients tested for COVID-19 from 3/3/2020-4/10/2020. Vitamin D deficiency was defined by the most recent 25-hydroxycholecalciferol <20ng/ml or 1,25-dihydroxycholecalciferol <18pg/ml within 1 year before COVID-19 testing. Treatment was defined by the most recent vitamin D type and dose, and treatment changes between the time of the most recent vitamin D level and time of COVID-19 testing. Vitamin D deficiency and treatment changes were combined to categorize vitamin D status at the time of COVID-19 testing as likely deficient(lastlevel-deficient/treatment-not-increased), likely sufficient(last-level-not-deficient/treatmentnot-decreased), or uncertain deficiency(last-level-deficient/treatment-increased or last-levelnot-deficient/treatment-decreased).Main Outcomes and Measures: The main outcome was testing positive for COVID-19. Multivariable analysis tested whether the most recent vitamin D level and treatment changes after that level were associated with testing positive for COVID-19 controlling for demographic and comorbidity indicators. Bivariate analyses of associations of treatment with vitamin D deficiency and COVID-19 were performed.Results: Among 4,314 patients tested for COVID-19, 499 had a vitamin D level in the year before testing. Vitamin D status at the time of COVID-19 testing was categorized as likely deficient for 127(25%) patients, likely sufficient for 291(58%) patients, and uncertain for 81(16%) patients. In multivariate analysis, testing positive for COVID-19 was associated with increasing age(RR(age<50)=1.05,p<0.021;RR(age≥50)=1.02,p<0.064)), non-white race(RR=2.54,p<0.01) and being likely vitamin D deficient (deficient/treatment-not-increased:RR=1.77,p<0.02) as compared to likely vitamin D sufficient(not-deficient/treatment-not-decreased), with predicted All rights reserved. No reuse allowed without permission.