Table 1.

Log causal odds ratios (95% confidence intervals) for coronary heart disease per standard deviation increase in HDL‐C, with 2‐sided P‐values. Estimates of the intercept are given in univariable and multivariable MR‐Egger

	Causal Estimate			MR‐Egger Intercept Test
	${\widehat{\theta }}_{\text{HDL‐C}}$ (CI)	$\text{se}({\widehat{\theta }}_{\text{HDL‐C}})$	P‐value	${\widehat{\theta }}_{\text{0E}}$	$\text{se}({\widehat{\theta }}_{\text{0E}})$	P‐value
Univariable IVW
All variants	−0.130 (−0.227, −0.033)	0.049	0.009	‐	‐	‐
Reduced set of variantsa	−0.114 (−0.211, −0.017)	0.049	0.022	‐	‐	‐
Univariable MR‐Egger
All variants	−0.016 (−0.138, 0.106)	0.062	0.800	−0.007	0.002	0.004
Reduced set of variantsa	0.067 (−0.070, 0.204)	0.069	0.332	−0.012	0.004	0.001
Multivariable IVW	−0.039 (−0.123, 0.045)	0.042	0.359	‐	‐	‐
Multivariable MR‐Egger	0.036 (−0.063, 0.134)	0.050	0.477	−0.005	0.002	0.008

Abbreviations: CI, confidence interval; HDL‐C, high‐density lipoprotein cholesterol; IVW, inverse‐variance weighted; MR, Mendelian randomization; SE, standard error.

^a95 variants associated with HDL‐C at a genome‐wide level of significance (P‐value<5×10⁻⁸).