Genetically Predicted Serum Vitamin D and COVID-19: A Mendelian Randomization Study

OBJECTIVES: Observational studies suggest that low vitamin D levels are associated with increased risk and severity of COVID-19 infection. We used Mendelian randomization (MR) to investigate causality of the vitamin D—COVID-19 association. METHODS: We constructed a biologically plausible genetic instrument for serum vitamin D based on replicated genome-wide significant variants in genes related to vitamin D metabolism, and evaluated the validity of the genetic instrument in COVID-19 risk subgroups in the UK Biobank. We then performed two sample MR using publically available summary data for the association of the genetic instrument with serum vitamin D in the UK Biobank and with COVID-19 outcomes, including infection, severe respiratory infection, and hospitalization, in the COVID-19 Host Genetics Initiative. We used the inverse-variance weighted MR method for all analyses. RESULTS: We found little to no evidence for an effect of genetically predicted serum vitamin D on susceptibility to or severity of COVID-19 infection. The odds ratio per standard deviation increase in genetically predicted serum vitamin D were: 1.04 (95% confidence interval 0.92 to 1.18) for any COVID-19 infection vs. population controls, 1.05 (0.84–1.31) for hospitalized COVID-19 vs. population controls, 0.96 (0.64 to 1.43) for severe respiratory COVID-19 vs. population controls, 1.15 (0.99 to 1.35) for COVID-19 positive vs. COVID-19 negative and, 1.44 (0.75 to 2.78) for hospitalized COVID-19 vs. non-hospitalized COVID-19. Results were similar in analyses where the genetic instrument included all variants with genome-wide significant associations with serum vitamin D (i.e., including variants with no known relationship to vitamin D metabolism) and where the genetic instrument was for risk of vitamin D deficiency. CONCLUSIONS: Our results suggest that unconfounded, long-term differences in vitamin D status do not causally affect susceptibility to and severity of COVID-19 infection. These findings are consistent with results of a recently published MR study and suggest that associations seen in observational studies may be driven by confounding. Future directions include extending this work to non-European ancestry populations and high-risk populations, for example persons with comorbid disease. FUNDING SOURCES: NIDDK, NHLBI and NHGRI of the NIH.