Table 2 Association between Vitamin D and COVID-19 risk in multivariable regression models.

From: An observational and Mendelian randomisation study on vitamin D and COVID-19 risk in UK Biobank

 

COVID positive (N = 1746) vs controls (N = 415,596)

COVID hospitalization (N = 1020) vs non- hospitalization (N = 576) cases

COVID death (N = 399) vs non-death (N = 1347) cases

OR (95% CI)

p-val

OR (95% CI)

p-val

OR (95% CI)

p-val

vitD (nmol/l)a

1.00 (0.99–1.01)

0.593

1.00 (0.99–1.01)

0.506

1.00 (0.99–1.01)

0.356

vitD_May_adjusted (nmol/l)a

1.00 (0.99–1.01)

0.592

1.00 (0.99–1.01)

0.674

1.00 (0.99–1.01)

0.324

vitD_categoricala

0–25 nmol/L

Ref

–

Ref

–

Ref

–

25–50 nmol/L

1.03 (0.87–1.20)

0.743

0.94 (0.64–1.39)

0.759

0.93 (0.57–1.52)

0.774

50 nmol/L

0.97 (0.81–1.16)

0.762

0.90 (0.63–1.28)

0.551

0.92 (0.54–1.56)

0.757

vitD-wGRS134b

0.91 (0.80–1.03)

0.134

1.05 (0.81–1.36)

0.721

1.25 (0.90–1.73)

0.175

vitD-UVB

1.00 (0.99–1.01)

0.557

0.98 (0.97–0.99)

< 2 × 10–16

0.97 (0.96–0.98)

< 2 × 10–16

vitD-wGRS134 + vitD-UVBc

vitD-wGRS134

0.92 (0.81–1.04)

0.191

0.88 (0.66–1.17)

0.391

1.12 (0.80–1.57)

0.508

vitD-UVB

1.00 (0.99–1.01)

0.511

0.98 (0.97–0.99)

< 2 × 10–16

0.97 (0.96–0.98)

4.35 × 10–16

  1. aAdjusted for age, gender, body mass index (BMI), month of blood draw (adjusted for vitD and vitD-categorical only), ethnicity, physical activity, smoking and alcohol status, sunshine exposure variables (i.e., time spend outdoors in summer, time spent outdoors in winter and the use of sun/uv protection), vitamin D supplement intake, deprivation index, and comorbidities of CVDs, diabetes, asthma, and malignancy.
  2. bMultivariable model was additionally adjusted for the first 20 genetic principal components and genotype panel.
  3. cMultivariable regression was fitted by including both vitD-wGRS134 and vitD-UVB in the same model to examine the effects of genetically predicted vitamin D levels and ambient UVB jointly.
Back to article page