Skip to main content
. Author manuscript; available in PMC: 2021 Jun 1.
Published in final edited form as: Curr Epidemiol Rep. 2020 May 16;7(2):104–112. doi: 10.1007/s40471-020-00233-6

Figure 2:

Figure 2:

Example of collider bias. An MR study of BMI and breast cancer mortality among breast cancer cases would be subject to collider bias if there are variants associated with both breast cancer incidence and prognosis, but not BMI (in the general population). Since BMI is associated with breast cancer risk, variants associated with BMI (GBMI) will be correlated with other risk variants (GBrCa) among breast cancer cases—violating MR assumptions and potentially inducing a spurious association between BMI and breast cancer mortality.