Confounding and interaction
DOI link for Confounding and interaction
Confounding and interaction book
In the last chapter we were concerned with only two variables: the risk factor and the disease status. Often a third factor may have an important influence on the apparent relationship between these two variables. If the third factor can explain (at least partially) this relationship, then
is present. For instance, a relationship between the number of children and prevalent breast cancer for a sample of mothers may be explained by the ages of the mothers: older mothers tend to have more children and also have a greater chance of having contracted breast cancer. Age is then the third factor that explains the observed relationship between number of children and breast cancer. The effect (upon breast cancer) of multiple childbearing is confounded with the effect of age.