Table 1 Relationship between genetic variants, putative confounders, and MI used to examine misclassification (Scenario A) and selection bias (Scenario B)
Exposure | Outcome | Population | Estimate (95% CI) |
|---|---|---|---|
Scenario A | |||
   rs12565497 | Smoking | MI = 0 | 0.95 (0.73, 1.25) |
   Smoking | MI | Total population | 2.08 (1.48, 2.94) |
   rs12565497 | Rx | MI = 0 | 0.60 (0.25, 1.40) |
   Rx | MI | Total population | 2.62 (1.18, 5.80) |
Scenario B | |||
   rs1466535 | Smoking | MI = 0 | 1.15 (0.88, 1.51) |
   Smoking | MI | rs1466535 AAa | 2.03 (1.60, 2.57) |
   Smoking | LDLb | Total population | 1.31 (1.03, 1.66) |
   rs1466535 | LDLb | Total population | 0.81 (0.66, 0.99) |
   rs1466535 | Smoking | Total population | 1.14 (0.91, 1.42) |
   rs1466535 | MI | Total population | 0.80 (0.64, 1.00) |