Reply to: Prognostic implications of vitamin D in patients with COVID-19

Excerpt

We thank Professor Missouris and colleagues for their interest in our paper. We agree that more data are needed; however, measuring 25-hydroxyvitamin D (25(OH)D) concentration in people who are already infected with SARS-CoV-2 will suffer from reverse causality as circulating vitamin D exhibits negative acute phase reactant behavior [1, 2]. All observational data have limitations and, despite the acknowledged limitations of using UK Biobank for this research question, the work we published perhaps remains as powerful as any to date. This is given key strengths of having access to serum 25(OH)D concentration measured before infection, its large scale, and adjustment for factors known to strongly influence vitamin D (i.e., ethnicity, body mass index, age, comorbidities, and socioeconomic deprivation). Many other studies that have not considered all relevant factors may have reported positive associations, but with a high likelihood of residual confounding. …