Table 2.

Two-sample Mendelian randomization estimations showing the effect of plasma lipids on aortic stenosis

Methods	Exposure	Odds ratioa	95% CI		P-value	Ph	Q-statistics
Inverse-variance weighted	HDL-cholesterol	0.86	0.69	1.06	0.17	<0.001	117.1
MR-Egger		0.98	0.70	1.37	0.91
Weighted median		0.99	0.76	1.29	0.93
MR-PRESSOb		NA	NA	NA	NA
MR-Egger interceptc		−0.009	−0.026	0.008	0.31
Inverse-variance weighted	LDL-cholesterol	1.58	1.30	1.91	<0.001	0.01	81.6
MR-Egger		1.63	1.19	2.24	<0.001
Weighted median		1.64	1.28	2.11	<0.001
MR-PRESSO		1.59	1.34	1.90	<0.001
MR-Egger interceptc		−0.002	−0.022	0.017	0.80
Inverse-variance weighted	Total cholesterol	1.60	1.33	1.92	<0.001	0.04	93.7
MR-Egger		1.82	1.32	2.53	<0.001
Weighted median		1.73	1.33	2.25	<0.001
MR-PRESSOb		NA	NA	NA	NA
MR-Egger interceptc		−0.009	−0.026	0.009	0.04
Inverse-variance weighted	Triglycerides	1.52	1.12	2.03	0.006	<0.001	77.5
MR-Egger		1.49	0.95	2.33	0.08
Weighted median		1.39	1.00	1.92	0.05
MR-PRESSO		1.55	1.20	2.00	0.002
MR-Egger interceptc		0.001	−0.024	0.026	0.91

The best causal estimation highlighted in bold .

CI, confidence interval; HDL, high density lipoprotein; LDL, low density lipoprotein; NA, not applicable; P_h, P-value for heterogeneity.

^aOdds ratio per 1 SD increase.

^bNo significant outliers.

^cRegression coefficient (95% CI).