Association, Cause and Causal Association: Means, Methods and Measures

Philosophers debate the metaphysical concept of cause, but medical scientists should facilitate the more pragmatic evaluation of cause, encompassing the factors that make a difference to the occurrence of health events. The identification of causal factors rests on our ability to design, conduct, and analyze studies, ensuring that the putative factors contribute to cause. Thus, we must bring to bear design strategies and analytic methods that will help avoid spurious or non-causal associations, primarily resulting from bias, confounding or reverse causation. In this chapter we review past and current views of causal association and the biases employed to infer "cause" from scientific data. We mention briefly examples of modern methodological techniques (e.g., instrumental variables, Mendelian randomization) that help remove confusion when forming causal inferences. Determination of causal association implies an imperative for intervention, to modify disease occurrence. Those interventions may be molecular, personal or societal, to better the public health.