Table 4.
Results from applied example
| Primary association | ||||
|---|---|---|---|---|
| All genetic variants | with target exposure | |||
| Analysis method | Estimate (SE) | P‐valuea | Estimate (SE) | P‐valuea |
| Low‐density lipoprotein cholesterol (LDL‐c) | ||||
| Inverse‐variance weighted | 0.482 (0.060) | *** | 0.470 (0.055) | *** |
| Simple median | 0.429 (0.070) | *** | 0.429 (0.079) | *** |
| Weighted median | 0.458 (0.065) | *** | 0.457 (0.065) | *** |
| Penalized weighted median | 0.457 (0.063) | *** | 0.457 (0.067) | *** |
| MR‐Egger regression: slope | 0.617 (0.103) | *** | 0.562 (0.094) | *** |
| intercept | −0.009 (0.005) | −0.006 (0.005) | ||
| High‐density lipoprotein cholesterol (HDL‐c) | ||||
| Inverse‐variance weighted | −0.254 (0.070) | *** | −0.137 (0.066) | * |
| Simple median | −0.267 (0.090) | ** | −0.224 (0.085) | ** |
| Weighted median | −0.069 (0.071) | −0.066 (0.065) | ||
| Penalized weighted median | −0.071 (0.068) | −0.064 (0.066) | ||
| MR‐Egger regression: slope | −0.013 (0.115) | 0.092 (0.107) | ||
| intercept | −0.014 (0.005) | * | −0.013 (0.005) | * |
| Triglycerides | ||||
| Inverse‐variance weighted | 0.416 (0.081) | *** | 0.417 (0.095) | ** |
| Simple median | 0.512 (0.101) | *** | 0.565 (0.105) | *** |
| Weighted median | 0.516 (0.084) | *** | 0.521 (0.087) | *** |
| Penalized weighted median | 0.528 (0.078) | *** | 0.539 (0.089) | *** |
| MR‐Egger regression: slope | 0.422 (0.140) | ** | 0.464 (0.155) | ** |
| intercept | −0.000 (0.008) | −0.004 (0.009) | ||
Estimates (standard errors) of causal effects of lipid fractions on coronary artery disease risk. Estimates are log odds ratios per 1 standard deviation increase in the exposure. The intercept term in MR‐Egger regression provides a test of directional pleiotropy.
a P‐values are indicated as: *, **, ***