Table 1.

Summary statistics for sizes of standardized effects of each exposure on COVID-19 hospitalization.

Exposure	Effect size
	Europeansa					Multiple Ethnicitya
	Mean	SD	Min	Max	$\widehat{\rho }$(exposure, COVID-19)b (SE)	Mean	SD	Min	Max	$\widehat{\rho }$(exposure, COVID-19)b (SE)
Type II Diabetes	7.54e−05	0.010	− 0.021	0.020	− 0.033 (0.089)	7.54e−05	0.010	− 0.021	0.020	0.153 (0.084)
BMI	3.99e−04	0.009	− 0.022	0.043	− 0.024 (0.029)	3.79e−04	0.009	− 0.022	0.043	0.013 (0.029)
DBP	− 1.78e−04	0.009	− 0.025	0.023	− 0.035 (0.033)	− 1.59e−04	0.009	− 0.025	0.023	0.026 (0.032)
SBP	− 4.01e−05	0.009	− 0.026	0.023	− 0.027 (0.034)	− 2.63e−05	0.009	− 0.026	0.023	0.036 (0.033)
PP	− 2.47e−04	0.009	− 0.022	0.027	− 0.021 (0.037)	− 2.11e−04	0.009	− 0.022	0.027	0.089 (0.037)

Mean, standard deviation, minimum, and maximum values correspond to the sizes of the standardized effects of the selected SNPs on each exposure, calculated from original GWAS estimates using Eqs. (1) and (3) above. Correlation coefficients represent the correlations between exposure effect sizes and outcome (COVID-19 hospitalization) effect sizes, the latter of which is calculated from the original COVID-19 GWAS using Eqs. (2) and (4) above. All values are presented only for the final selected SNPs (i.e., those with a p-value less than 5e−8 and a linkage disequilibrium coefficient less than 0.1 in the original GWAS). Correlation coefficients (a) Estimates from the European set are from 2956 European individuals; estimates from the variable ethnicity set are from 10,908 individuals of either of European, African, Hispanic, or Middle Eastern descent. (b) the correlation between the effect sizes for each SNP’s estimated effect on the probability of the respective exposure or COVID-19 hospitalization after all mentioned exclusions (SNP independence, p < 5e−8, Pearson’s r < 0.1 with index SNP) were applied.