An association can be causal or be resulting from a confounding when a third factor (or more factors) is directly linked to the risk of developing the disease (Gordis, 2014). The stronger the association, greater is odds ratio in the association, therefore, greater is the likelihood of the association being causal (Gordis, 2014). Luiz & Struchiner (2002) argues that a strong association is more likely to be causal due to the presence of an obvious bias. Consequently, weak associations are usually explained by biases that are not detected, however, causality hypothesis cannot be completely ruled out in weak associations(Luiz & Struchiner, 2002).
